Upload
sara-verleden
View
215
Download
5
Tags:
Embed Size (px)
Citation preview
HIGH RISE BUILDING MOVEMENT MONITORING USING RTK-GPS
(CASE STUDY: MENARA SARAWAK ENTERPRISE)
SHU KIAN KOK
UNIVERSITI TEKNOLOGI MALAYSIA
PSZ 19:16 (Pind. 1/97)
UNIVERSITI TEKNOLOGI MALAYSIA
BORANG PENGESAHAN STATUS TESIS•
JUDUL : HIGH RISE BUILDING MOVEMENT MONITORING USING
RTK-GPS (CASE STUDY: MENARA SARAWAK ENTERPRISE)
SESI PENGAJIAN : 2005/2006
Saya : SHU KIAN KOK (HURUF BESAR)
mengaku membenarkan tesis (PSM/Sarjana/Doktor Falsafah)* ini disimpan di Perpustakaan Universiti Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut :
1. Tesis adalah hakmilik Universiti Teknologi Malaysia. 2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuk tujuan pengajian
sahaja. 3. Perpustakaan dibenarkan membuat salinan tesis in sebagai bahan pertukaran antara institusi
pengajian tinggi. 4. **Sila tandakan ( ) SULIT ( Mengandungi maklumat yang berdarjah keselamatan atau kepentingan Malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972) TERHAD ( Mengandungi maklumat TERHAD yang telah ditentukan oleh organisasi/badan di mana penyelidikan dijalankan)
TIDAK TERHAD
Disahkan oleh
____________________________________ _________________________________________ (TANDATANGAN PENULIS) (TANDATANGAN PENYELIA) Alamat Tetap :
53, KG BARAT KERAYONG ASSOC. PROF. DR. WAN ABDUL
28200 BERA, AZIZ WAN MOHD AKIB
PAHANG D.M NAMA PENYELIA Tarikh : Tarikh :
CATATAN : * Potong yang tidak berkenaan.
** Jika tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/ organisasi berkenaan dengan menyatakan sekali tempoh tesis ini perlu dikelaskan sebagai SULIT atau TERHAD.
• Tesis dimaksudkan sebagai tesis bagi Ijazah Doktor Falsafah dan Sarjana secara penyelidikan, atau disertasi bagi pengajian secara kerja kursus dan penyelidikan, atau Laporan Projek Sarjana Muda (PSM).
“I/We* hereby declare that I/we* have read this thesis and in my/our*
opinion this thesis is sufficient in terms of scope and quality for the
award of the degree of Master of Science (Geomatic Engineering)”
Signature : ....................................................
Name of Supervisor I : Assoc. Prof. Dr. Wan Abdul Aziz Wan Mohd Akib
Date : ....................................................
Signature : ....................................................
Name of Supervisor II : ....................................................
Date : ....................................................
Signature : ....................................................
Name of Supervisor III : ....................................................
Date : ....................................................
* Delete as necessary
BAHAGIAN A – Pengesahan Kerjasama* Adalah disahkan bahawa projek penyelidikan tesis ini telah dilaksanakan melalui kerjasama antara _______________________ dengan _______________________
Disahkan oleh:
Tandatangan : Tarikh :
Nama :
Jawatan : (Cop rasmi)
* Jika penyediaan tesis/projek melibatkan kerjasama.
BAHAGIAN B – Untuk Kegunaan Pejabat Sekolah Pengajian Siswazah
Tesis ini telah diperiksa dan diakui oleh:
Nama dan Alamat Pemeriksa Luar : Dr. Noordin Bin Ahmad
Geoinfo Services Sdn. Bhd,
31 Jalan Bandar 2, Taman Melawati,
53100 Kuala Lumpur.
Nama dan Alamat Pemeriksa Dalam : Prof. Madya Dr. Md Nor Bin Kamaruddin
Fakulti Kejuruteraan & Sains Geoinformasi
UTM, Skudai.
Nama Penyelia Lain (jika ada) :
Disahkan oleh Penolong Pendaftar di SPS:
Tandatangan : Tarikh :
Nama : GANESAN A/L ANDIMUTHU
HIGH RISE BUILDING MOVEMENT MONITORING USING RTK-GPS
(CASE STUDY: MENARA SARAWAK ENTERPRISE)
SHU KIAN KOK
A thesis submitted in fulfilment of the
requirements for the award of the degree of
Master of Science (Geomatic Engineering)
Faculty of Geoinformation Science and Engineering
Universiti Teknologi Malaysia
DISEMBER 2005
ii
“I declare that this thesis “High Rise Building Movement Monitoring Using RTK-
GPS (Case Study: Menara Sarawak Enterprise)” is the result of my own research
except as cited in the references. The thesis has not been accepted for any degree and
is not concurrently submitted in candidature of any other degree.”
Signature : .........................................
Name : SHU KIAN KOK
Date : .........................................
iii
To my beloved mother and father
iv
ACKNOWLEDGEMENT
In particular, I wish to express my sincere appreciation to my supervisor,
Assoc. Prof. Dr. Wan Abdul Aziz Wan Mohd Akib for his encouragement, guidance,
and friendship. Without his continue support and interest, this thesis would not have
been the same as presented here.
I would like to extend my deepest gratitude to Mr Zulkarnaini Mat Amin for
his guidance and contribution in this thesis. Besides that, i would like to express my
gratitude to Geodesy Section Department Survey and Mapping Malaysia (DSMM)
for providing the GPS data. I also wish to extend my heartiest gratitude to Assoc.
Prof. Dr. Md. Nor Kamarudin for his kindness to lend his anemometer instrument for
this research.
A special thanks to Jayalah Cemerlang Reality Sdn. Bhd. for the permission
and technical assistance to carry out this research at the Menara Sarawak Enterprise
Building, Johor Bharu.
I also wish to express my thankful to the following persons whose have
assisted me by factual help in the implementation of this research:
The technicians at the Engineering Survey Laboratory, FKSG, UTM.
Mr. Voon Min Hi
Mr. Kee Tuan Chew
Mr Wong Chee Siang Tony
Mr Tan Wee Keng
Last, but not last, I am also grateful to my family members for giving me all
the support that I needed.
v
ABSTRACT
The need for deformation surveys of large engineering structures such as long
span bridges, dams and tall structures often arises from concerns associated with
environmental protection, property damage and public safety. There are many high
buildings nowadays, therefore it is very important to monitor the buildings to ensure
they are still under stable condition. Recently, the Global Positioning System (GPS)
especially Real Time Kinematics (RTK-GPS) has emerged as a survey tool for many
deformation applications. The RTK-GPS is carrier phase observation processed in
real time, giving results such as position coordinates. This study highlights the
concept and methodology of the continuous RTK-GPS and its potential application
for high rise building monitoring surveys. The main objectives of this study are to
study the ability and efficiency of the continuous RTK-GPS method in high rise
building’ deformation detection and also to develop KFilter program for movement
monitoring using Matlab v6.1 with Kalman Filter method. The GPS instruments’
calibrations had been carried out to ensure accuracy and reliability of the continuous
RTK-GPS observation for high rise building movement monitoring. The surveys had
been carried out on Menara Sarawak Enterprise, Johore Malaysia in two different
epochs. Thus, the developed KFilter program is able to perform the movement
monitoring analysis on the observed data to classify the stability of the building. The
results of this study shows that the continuous RTK-GPS can provide 1cm and 2cm
accuracy for horizontal and vertical respectively. The effectiveness of this technique
depends on radio link communication whereby obstructions will cause the
communication signal to fail. From the KFilter program analysis, the results shows
that the Menara Sarawak Enterprise building is stable. The continuous RTK-GPS
epoch 1 and epoch 2 analyses had shown the building is stable although displacement
distance around 0.5cm and 1cm respectively are detected.
vi
ABSTRAK
Keperluan bagi melaksanakan ukur deformasi terhadap struktur kejuruteraan
besar seperti jambatan, empangan dan bangunan tinggi adalah semakin penting untuk
perjagaan alam sekitar dan melindungi keselamatan awam. Terdapat semakin banyak
bangunan tinggi pada masa kini, maka amat penting untuk memastikan bangunan
tinggi berkenaan dalam keadaan yang stabil. Untuk masa kini, Global Positioning
System (GPS) telah digunakan sebagai alat pengukuran bagi kebanyakan kerja-kerja
deformasi. RTK-GPS adalah cerapan fasa pembawa yang dijalankan dalam masa
hakiki menghasilkan koordinat kedudukan. Kajian ini membincangkan konsep dan
potensi aplikasi RTK untuk ukur pemantauan bangunan tinggi. Objektif utama kajian
ini adalah untuk mengkaji kebolehan dan keberkesanan bagi teknik continuous RTK-
GPS di pengesanan deformasi bangunan tinggi dan membina program KFilter untuk
pemantauan pergerakan dengan menggunakan Matlab v6.1 bersama dengan teknik
Kalman Filter. Kalibrasi peralatan GPS telah dijalankan untuk memastikan kejituan
dan keupayaan cerapan continuous RTK-GPS bagi pemantauan pergerakan
bangunan tinggi. Percerapan teknik ini telah dilaksanakan di Menara Sarawak
Enterprise dalam 2 epok yang berlainan. Lepas itu, program KFilter digunakan untuk
analisis pemantauan pergerakan ke atas cerapan data demi menentukan kestabilan
bangunan berkenaan. Hasil kajian ini menunjukkan bahawa continuous RTK-GPS
dapat memberi kejituan mendatar 1cm dan menegak 2cm. Keberkesanan teknik ini
amat bergantung kepada perhubungan komunikasi radio dimana halangan akan
menyebabkan isyarat komunikasi radio terputus. Daripada hasil analisis program
KFilter menunjukkan bahawa Menara Sarawak Enterprise dalam keadaan stabil.
Analisis epok 1 dan 2 bagi cerapan continuous RTK-GPS mengesahkan bangunan
tersebut stabil walaupun jarak pergerakan lebih kurang 0.5cm dan 1cm telah dikesan
dalam kedua-dua epok.
vii
TABLE OF CONTENTS CHAPTER TITLE PAGE THESIS STATUS DECLARATION
SUPERVISOR’S DECLARATION
DECLARATION ON COOPERATION WITH
OUTSIDE AGENCIES AND CERTIFICATION OF
EXAMINATION
TITLE PAGE i
DECLARATION ii
DEDICATION iii
ACKNOWLEDGEMENTS iv
ABSTRACT v
ABSTRAK vi
TABLE OF CONTENTS vii
LIST OF TABLES xi
LIST OF FIGURES xii
LIST OF ABBREATIONS xv
LIST OF APPENDICES xvi
1 INTRODUCTION 1
1.1 Introduction 1
1.2 Problem Statement 4
1.3 Research Objectives 5
1.4 Research Scopes 5
1.5 Significance of Study 5
viii
1.6 Research Methodology 6
1.6.1 Literature Review 7
1.6.2 Field Data Acquisition 7
1.6.3 Development of KFilter Program 7
1.6.4 Observation Data Processing 8
1.6.5 Analyses and Results 8
1.6.6 Conclusions and Recommendations 8
1.7 Thesis Overview 8
2 LITERATURE REVIEW 10
2.1 Literature Review 10
2.2 High Rise Buildings Structure Material 16
2.3 Deformation in Structure 17
2.3.1 Deflection of Beams 18
2.3.2 Settlement of Foundations 19
2.3.3 Wind Loading Problem 20
2.4 Review of GPS 21
2.5 GPS Positioning Techniques 22
2.5.1 Real Time Kinematics (RTK-GPS) 23
2.6 Error Sources in GPS Measurement 24
3 THE APPLICATION OF KALMAN FILTER IN
DEFORMATION STUDY 28
3.1 Introduction 28
3.1.1 The Discrete Kalman Filter Algorithm 31
3.1.2 The Extended Kalman Filter 32
ix
3.2 Advantages, Problems and Disadvantages of Kalman Filter 34
3.3 Application of Kalman Filter In Deformation Monitoring 36
4 FIELD METHODOLOGY AND DATA PROCESSING 38
4.1 Introduction 38
4.2 The Menara Sarawak Enterprise Monitoring Network 39
4.3 Instruments Used for GPS Observation 42
4.4 GPS Instruments Calibration 43
4.4.1 Test on RTK – GPS Performance 43
4.4.2 Test on Accuracy of RTK-GPS Baseline 45
4.5 GPS Observation 47
4.5.1 GPS Network of Coordinates Transfer 48
4.5.2 GPS Monitoring Network 49
4.6 Data Processing and Adjustment 50
4.6.1 Trimble Geomatics Office Data Downloading 51
4.6.2 Leica Ski Pro Data Downloading 52
4.7 KFilter Program 52
4.8 Simulation Test 57
4.8.1 ‘Movement’ Simulation Test 57
4.8.2 ‘Timing’ Simulation Test 58
4.9 Static GPS Deformation Analysis 59
4.10 Movement Monitoring Analysis 62
4.11 Study of Wind Effect (Vibration) Using RTK-GPS Data 62
x
5 ANALYSES AND RESULTS 64
5.1 Introduction 64
5.2 Results Analysis for Study on RTK-GPS Baseline 64
5.3 Results Analysis for Test on Accuracy of RTK-GPS
Baseline 65
5.4 Results Analysis on ‘Movement’ Simulation Test 67
5.5 Results Analysis on ‘Timing’ Simulation Test 69
5.6 Case Study: Menara Sarawak Enterprise 70
5.7 Results Analysis For Study of Wind Effect (Vibration)
Using RTK-GPS Data 73
5.8 Summary 79
6 CONCLUSIONS AND RECOMMENDATIONS 81
6.1 Conclusions 81
6.2 Recommendations 82
REFERENCES 84 APPENDICES 93 - 116
xi
LIST OF TABLES
TABLE NO. TITLE PAGE 4.1 Adjusted Grid Coordinates from Static Processing 44 4.2 Adjusted Geodetic Coordinates from Static Processing 44 4.3 Adjusted Grid Coordinates from Static Processing 46 4.4 Adjusted Geodetic Coordinates from Static Processing 46 4.5 GPS Observation Schedule of Menara Sarawak Enterprise
Building 49
4.6 Data processing Options 51 4.7 Schedule of ‘Timing’ Simulation Test Observation 59 5.1 Analysis on One and half hour Continuous RTK-GPS Data
For Station UTMR 64
5.2 RMS Analysis on Continuous RTK-GPS Data for T200,
T300 and TR2300 65
5.3 Explanation Analysis 65 5.4 Simulation Test for Vertical Axis 67 5.5 Simulation Test for Horizontal (Northing & Easting) 68 5.6 Results Processing From GPS DEFORMATION ANALYSIS
PROGRAM, GPSAD2000 and KFilter 72
xii
LIST OF FIGURES
FIGURE NO. TITLE PAGE
1.1 General Definition of High-rise Building 2 1.2 Menara Sarawak Enterprise 3 1.3 Flow of Research Methodology 6 2.1 Comparison of the Bulbs of Pressure under a Single
Footing on Test Load and Under a Large Building 19
2.2 GPS Segments 21 2.3 RTK-GPS Observation Configuration 24 4.1 DSMM Geodetic Control (GPS) Station, J416 39 4.2 Location of Control and Monitoring Stations 40
4.3 Base 1 (B1) 40 4.4 Base 2 (B2) 40 4.5 Rover 1 (R1) 41 4.6 Rover 2 (R2) 41 4.7 Design of Rover Monument 41 4.8 Leica GPS System 500 Receiver 42 4.9 Trimble 4800 Series GPS Receiver 43 4.10 Coordinates of UTMB and UTMR Derived from
RTKNet Stations 44
xiii
4.11 Coordinates of T200, T300 and TR2300 Derived from TRS
Station and JHJY RTKNet Stations 46 4.12: Information of Satellite Visibility on 21/12/2004 47 4.13 Information of DOP Horizontal on 21/12/2004 48 4.14 Information of DOP Vertical on 21/12/2004 48 4.15 GPS Network of Coordinates Transfer 49 4.16 GPS Monitoring Network 50 4.17 KFilter user interface 52 4.18 Flow Chart of Stage Analysis KFilter 53 4.19 Format of Input Data for Developed Program KFilter 54 4.20 The Deformation Visualization Graph 54 4.21 Flow Chart of KFilter Program 55 4.22 Example of Deformation Report 56 4.23 Preparation of ‘Movement’ Simulation Test 57 4.24 Static (Left of Figure) and ‘Vibrated’ (Right of Figure) 58 4.25 Process Methodology of Static GPS Deformation Analysis 60 4.26 Anemometer 63 5.1 No Deformation Detected 69 5.2 Deformation Detected 70 5.3 Northing and Easting Displacements Graph 73 5.4 Northing Movements Value Resulted From Winds Effects 75 5.5 Easting Movements Value Resulted From Winds Effects 76
xiv
5.6 WGS84 Ellipsoid Height Movements Value Resulted From Winds
Effects 77
5.7 The Deformation Report (KFilter) for Without Wind
Effect and With Wind Effect 78
xv
LIST OF ABBREATIONS
GPS Global Positioning System
Hz Hertz
RTK Real Time Kinematics
cm centimeter
mm millimeter
m meter
hr hour
PRN Pseudo Random Noise
ppm Part per million
DSMM Department of Survey and Mapping Malaysia
TGO Trimble Geomatics Office
DOP Dilution of Positioning
RMS Roof Mean Squares
OTF On-the-fly
B1 Base 1
B2 Base 2
R1 Rover 1
R2 Rover 2
WGS84 World Geodetic System 1984
cont. continuous
LSE Least Square Estimation
xvi
LIST OF APPENDICES
APPENDIX. TITLE PAGE A SPECIFICATIONS OF LEICA GPS SYSTEM 500 93 B SPECIFICATION OF TRIMBLE 4800 GPS SYSTEM 96 C ONE HOUR CONTINUOUS RTK-GPS
OBSERVATION DATA FOR UTMB AND UTMR 99 D HALF HOUR CONTINUOUS RTK-GPS
OBSERVATION DATA FOR UTMB AND UTMR 100 E 5 MINUTES OBSERVATION DATA FOR T200 (BASE)
AND TR2300 (ROVER) 101 F 2 MINUTES OBSERVATION DATA FOR T300 (BASE)
AND TR2300 (ROVER) 102 G NETWORK ADJUSTMENT REPORT
(TRIMBLE GEOMATIC OFFICE) 103 H TRIMBLE GEOMATICS OFFICE DATA
DOWNLOADING PROCEDURES 107 I LEICA SKI PRO DATA DOWNLOADING
PROCEDURES 108 J OBSERVATION SCHEDULE OF ‘TIMING’
SIMULATION TEST 109
xvii
K SPECIFICATION OF ANEMOMETER DAVIS 111 L DEFORMATION REPORT FOR GPS
DEFORMATION ANALYSIS PROGRAM 113 N DEFORMATION REPORT FOR GPSAD2000 114 M DEFORMATION REPORT FOR KFilter 116
CHAPTER 1
INTRODUCTION
1.1 Introduction
Deformation refers to the changes which a deformable body undergoes in its
shapes, dimension and position. Deformation survey can be used for obtaining
information about the stability of some objects like natural or man-made objects. The
man-made objects such as large engineering structures are subject to deformation due
to various factors: changes of ground water level, tidal phenomena, tectonic
phenomena, land movements, or any other natural disasters. The large engineering
structures include dams, long span bridges, high rise buildings, reservoirs, sport
domes, planetariums, Olympic stadium etc. Therefore it is important to measure this
movement for the purpose of safety assessment as well as to prevent any disaster in
the future.
A high-rise building is defined as a building 35 meters or greater in height,
which is divided at regular intervals into occupiable levels (Emporis, 2004). To be
considered a high-rise building an edifice must be based on solid ground, and
fabricated along its full height through deliberate processes (as opposed to naturally-
occurring formations). A high-rise building is distinguished from other tall man-
made structures by the following guidelines
i. It must be divided into multiple levels of at least 2 meters in height;
ii. If it has fewer than 12 such internal levels – see Figure 1.1, then the
highest undivided portion must not exceed 50% of the total height.
2
Figure 1.1: General Definition of High-rise Building (Emporis, 2004)
Nowadays, there are much more large and tall engineering structures (high
rise buildings) than the past. These structures are designed to be much more flexible
and to resist extensive damage from changes in temperature, severe wind gusts and
earthquakes. Structural engineers require precise, reliable instruments to resolve their
concerns about angular movements, displacements and structural vibrations. Hence,
some actions can be taken before the disasters strike. It can save lives, avert large
financial liabilities and avoid severe environmental damage.
In general, there are two types of technique in deformation survey, i.e.
geodetic surveys and non-geodetic survey (geotechnical and structural). Geodetic
survey using total stations, precise levels, Global Positioning System (GPS), etc can
be based on absolute and relative networks. Deformation detection via geodetic
method mainly consists of two step analysis independent least square estimation
(LSE) of each epochs followed by deformation detection between two epochs. On
the other hand, geotechnical and structural methods use special equipments to
measure changes in length (extensometer), inclination (inclinometer), strain
(strainmeter) etc.
In contrast, the GPS technology can measure directly the position coordinates
and nowadays relative displacements can be measured at the rate of 10Hz or higher.
This provides a great opportunity to monitor, in real time, the displacement or
3
deflection, behavior of engineering structures under different loading conditions,
through automated change detection’ and alarm notification procedures (Ogaja et. al.,
2001).
One of the most recent real time GPS techniques to date is RTK-GPS. Such
real-time application had been widely used in various survey applications and
navigational purposes, regardless on land, at sea or in the air (Rizos, 1999). RTK-
GPS can achieve the accuracy of ± 2 cm + 2 ppm. In RTK-GPS configuration, a
receiver is placed on the reference point with known coordinates as reference station.
This reference station will continuously transmit correction message to rover
receiver. For example, a fully automated monitoring system using RTK-GPS
technique had been implemented successfully in Dam Diamond Valley Lake,
California. This system will provide the information on the displacement of the
monitoring points weekly (Michael et.al., 2001).
High rise building research was carried out at Menara Sarawak Enterprise
which is located at Stulang Laut, Johor Bahru (see Figure 1.2). The height of the
building is almost 120m above ground. The building’s structure is consisted of 30
storey tower and 3 basements as car park level. Each storey is about 3.5 meters in
height.
Figure 1.2: Menara Sarawak Enterprise
4
1.2 Problem Statement
Since our national high rise buildings inventory are aging and they are
carrying more and more loads, the need to monitor high rise buildings’ performance
has increased significantly over the past few years. High rise buildings require
careful provisions of life-safety systems because of their height and their large
density of occupant. Therefore, both for maintenance and repair planning, high-rise
building monitoring is becoming increasingly important. What's more, structural
deformation and deterioration problems faced by the high-rise building authorities
are very similar to those faced by dam, large span bridge, and highway and railroad
authorities.
In satellite surveying, static GPS positioning technique is perhaps the most
common method used by surveyors because of the high accuracies it can obtain. In
general, one to two hours is a good observation period for Static GPS baseline up to
30 kilometers. Static GPS method can be used for deformation detection. However,
this method is not suitable for continuous deformation monitoring because Static
GPS methods cannot provide data continuously compared to Real Time Kinematics
(RTK) GPS positioning technique. A high precision, carrier phase based, RTK-GPS
has been considered to play an important role as an alternative technique to the
geotechnical methods or in addition to such a sensor (Ogaja, 2000). The notable
advantage of using RTK-GPS is that this technique can detect deformation if the
structure has drifted (a few cm) relative to some reference or baseline while
accelerometers can not detect, directly, the absolute or relative displacements of the
structure (Ogaja, 2000). Therefore, the aim of this study is to analyze the potential
application of RTK-GPS method in deformation monitoring purpose of high rise
building.
5
1.3 Research Objectives
The objectives of this study have to fulfill the following requirements:-
i. To study the ability and efficiency of the continuous RTK-GPS
method in high rise building’s deformation detection.
ii. To develop program for monitoring movement using Kalman Filter
algorithms.
1.4 Research Scopes
The research scopes of this study involve:-
i To carry out the GPS data observation in continuous RTK-GPS
technique
ii. To process and analyze the data in order to get the pattern and
magnitude of the deformation.
iii. To study the ability of RTK-GPS to be applied in high density
construction area.
1.5 Significance of Study
The significance of this study includes:-
i Develop a RTK-GPS movement monitoring system with the aid of
Kalman Filter on high rise building.
ii Determine the type of the errors caused by RTK-GPS observation in
movement monitoring.
6
1.6 Research Methodology
Research methodology is divided into a few stages in order to achieve the
objectives of this study (see Figure 1.3).
Literature Review.
Calibration of GPS Instruments
Design the Control and Monitoring Stations’ Network.
Field Data Observation.
Analysis and Results
Conclusion and
Recommendations
Data Processing
Develop KFilter Program Using Matlab Version 6.1 and
Kalman Filter Method for Movement Monitoring.
Simulation Tests.
Figure 1.3: Flow of Research Methodology
7
1.6.1 Literature Review
Literature reviews were carried out on the concepts of GPS, deformation
surveying, structural monitoring, and the understanding of the GPS instrumentation.
Calibration of GPS instruments (Trimble 4800 series and Leica System 500) had
been carried out to ensure the instruments in good condition to perform the GPS
observations. At this stage, the GPS instruments were studied to ensure the
instruments can carry out continuous Real Time Kinematics technique with one
second sampling rate. Both of them are dual-frequency (L1 and L2) and able provide
high precision results.
1.6.2 Field Data Acquisition
Before field data acquisition has been carry out, the control network and
monitoring stations should be designed and placed in suitable locations. In this study,
Trimble 4800 series and Leica System 500 observations had been used to carry out
for two epochs. First epoch had been carried out on 21 December to 23 December
2004 whereas the second epoch had carried out on 28 April 2005 to 29 April 2005.
1.6.3 Development of KFilter Program KFilter program had been developed using Matlab version 6.1 and based on
the Kalman Filter algorithm for the object movements monitoring purpose. The
program will read continuous RTK input data from GPS receiver and performs
movement monitoring analyses with the help of Kalman Filter algorithm. The
program will give some warning alarms if it detected displacements from the
observed data. Beside that, the simulation tests had been carried out to ensure the
reliability of the developed KFilter program in movement monitoring.
8
1.6.4 Observation Data Processing The observed data had been processed using certain commercial software or
self-developed program. The continuous RTK data had been downloaded to Leica
SKI-Pro and Trimble Geomatics Office. The output files with its suitable format for
the developed program will be created. The program which is developed using
Matlab v6.1 will perform its analysis based on the observation data.
1.6.5 Analyses and Results
Analyses in this study include the reliability of the observed data and the
effectiveness of the program in determining the stability of the high rise building. In
this study, the program will perform structural monitoring analyze on the GPS
observation data.
1.6.6 Conclusions and Recommendations Summarizes findings, make conclusions and recommends topics for further
investigations. The prospects and limitations of continuous RTK-GPS technique
were also presented.
1.7 Thesis Overview
Chapter 1 described the important of the deformation monitoring for high rise
building using Global Positioning System (GPS). The problem statement, research
scopes and the significant of the study had been described.
9
Literature review is an important stage of this study to ensure that the
research can be carried out successfully. It was discussed in Chapter 2. The types of
material of high buildings were stated out in this chapter. The factors that affect
concrete strength of the buildings were explained. The RTK-GPS was used in this
study for movement monitoring. Thus the introduction and literature review on the
RTK-GPS were stated out. There were included the errors of RTK-GPS observation
and its configuration.
The program for movement monitoring with the help of Kalman Filter
method had been developed. Therefore, the introduction and definition of the
Kalman Filter method including its algorithms were elaborated in Chapter 3.
The calibration of GPS instruments and field data acquisition is the most
important stage in the study and discussed in details in chapter 4. 2 epochs of
observation were carried out in the study. Setting up a deformation network which
consists of selected reference stations and the monitoring points is necessary. The
GPS observations were carried out using GPS instruments, Leica GPS System 500
and Trimble GPS 4800 System. Meanwhile, the software used for data downloading
and data processing were Trimble Geomatics Office and Leica Ski-Pro. The
simulation test was carried out to ensure that the developed program can detect the
displacement or vibration successfully.
Chapter 5 discussed the calibration and simulation tests analysis results.
Besides that, the stability analysis of Menara Sarawak Enterprise using developed
program had been carried out. The analysis was verified by other program, such as
GPS Deformation Analysis Program-Bayrak (Turkey) and GPSAD2000-Boon
(Malaysia). This increased the reliability of the analysis for Menara Sarawak
Enterprise movement monitoring.
Lastly, chapter 6 presented the conclusions of this study. Some
recommendations had been proposed and considered to improve this study.
CHAPTER 2
LITERATURE REVIEW 2.1 Literature Review
The Global Positioning System (GPS), also known as NAVSTAR
(NAVigation System using Time and Ranging) is a space-based navigation system
created and developed by the US Department of Defense (DoD) for real time
navigation since the end of the 70’s. For the past ten years, the GPS has made a
strong impact on the geodetic world. The main goal of the GPS is to provide
worldwide, all weather, continuous radio navigation support to users to determine
position, velocity and time throughout the world (Hofmann-Wellenhof et. al., 1994).
With recent full constellation of GPS satellites, available satellite signals processing
software, the differential measurement of the satellite signals using geodetic type of
GPS receivers will provide any baseline vector with high precision at millimeter
level (Leick, 2004).
In the basic approaches of geometrical analysis, the displacements at discrete
points are directly compared with specified tolerances. In more advanced analyses,
the point displacements are assessed for spatial trend, and a displacement field is
determined by the fitting of a suitable spatial function. The displacement field may
then be transformed into a strain field, which provides a unique description of the
overall change in geometric status, by the selection of a suitable deformation model
(Chrzanowski et al., 1986).
At present, instead of static deformation monitoring approaches, continuous
dynamic deformation monitoring methods have been increasingly used to understand
11
natural events such as landslides and to monitor the stability of manmade structures
such as building, bridges and dam (Bock and Bevis, 1999; Leick, 2004).
An RTK-GPS (Real Time Kinematic – Global Positioning System) has a
nominal accuracy of ±1cm +1ppm for horizontal displacements with a sampling rate
of 10Hz. It was found to be suitable for measuring building responses when the
vibration frequency is lower than 2Hz and the vibration amplitude is larger than 2cm.
According to Tamura et. al., (2004), the RTK-GPS can measure not only dynamic
components but also static components and quasistatic components. The member
stresses obtained by hybrid use of FEM analysis and RTK-GPS were close to the
member stresses measured by strain gauges. Meanwhile, accelerometers have been
used for field measurements of wind-induced responses of buildings. However,
wind-induced responses consists of a static component, i.e a mean value and a
dynamic fluctuating component. The static component is difficult to measure with
accelerometers. The uses of RTK-GPS for measurements of building responses have
been proposed. (Tamura et. al., 2004)
Ge et.al., (2000) had tested the feasibility of a “fully closed-loop design” for
large structures, two Leica CRS1000 and two Trimble MS750 GPS receivers have
been tested in the Real Time Kinematics (RTK) mode, with fast sampling rates 10
Hz and 20 Hz to determine how well they measure relative displacements, from very
low (DC) to high (10Hz) frequency. In the test involving two Leica GPS receivers
through there was much noise in the low frequency band due to the effects of
atmosphere, multipath, receiver noise, etc, FFT analysis of the fast RTK outputs
indicate that vibrations of 2.3 Hz and 4.3Hz with an amplitude of 12.7mm, applied to
the rover antenna to simulate vibration of structures, can be recognized in both the
time and frequency domain (though they are more clearly resolved in the frequency
domain), not only in the latitude and longitude components, but also in the height
component. The fast RTK results at DC, 2.3Hz and 4.3Hz were found to exhibit
similar noise patterns. The choke ring antenna is shown to yield results with higher
SNR than the standard antenna. In addition to the signals, harmonics and aliasing
were also detected. Data from a reference accelerometer confirmed that the harmonic
was generated by the shaker and is not an artifact of the GPS experiment. The fact
12
that all the supplied signals and their by-products, namely the harmonics and
aliasing, are successfully resolved, from 0.8Hz, 1.4Hz, 2.3Hz, 3.1Hz, 4.3Hz to
4.6Hz, has proved the strength of high rate GPS RTK as the technology to support
the "fully closed-loop design" for large structures. In the Trimble GPS receiver test,
RTK measurements at 20Hz for over three hours, over the same period on five
successive days, were recorded to determine whether measurements on consecutive
days can be used for noise reduction (Ge et. al., 2000).
According to Ogaja et. al. (2001), a high precision dynamic RTK-GPS
system had been installed at the Republic Plaza Building, Singapore. The purpose of
the system is to provide, to sub-centimeter accuracy, and at rate of up to 10 samples
per second, position vectors with respect to a fixed base station, of two antennas
installed on the building parapet. The system was operated in parallel with, linked to
an existing logging system that records signals from accelerometers and
anemometers. The observations data analyzes by ‘Time-Frequency’ wavelets method
to automatically detect ‘low’ and ‘high’ frequency components embedded in the
noisy time series, frequency changes and their onset times. The algorithm is
formulated through the estimation of ‘instantaneous’ using the wavelet transform and
‘change detection’ using the cumulative sum (CUSUM) scheme (Mertikas & Rizos,
1997). The wavelet transforms method gives the time locations of each frequency.
This allows the visualization of transient frequencies and the determination of the
occurrence of discontinuities in the signal. However, there was no achievable
accuracy had been mentioned in this experiment.
In time series analysis, frequency-domain signature is obtained by converting
time-domain data into its unique frequency components using a Fast Fourier
Transform (FFT). Through the study of frequency-domain vibration signatures, the
natural frequencies of structures can be detected and isolated to form the basic data
for seismic and wind response analyses. Such data are valuable as more and more
important high rise building are analyzed through structural performance and seismic
loading for improved structural design (Brownjohn et. al., 2000). Ideally, in the time-
frequency analysis, it is preferable to represent the structural signature for each of the
three directional components Northing, Easting and Height in such a way that it: (a)
13
indicates which frequencies existed for a duration, (b) shows how the frequencies
change with time and (c) shows the time-based waveform.
The feasibility of GPS for detecting and discriminating tall building
displacement and frequency signatures was investigated through the use of a joint
time frequency domain analysis. According to Ogaja et. al., (2000), the analysis of
the data collected from the UNSW-GPS-Seismometer experiment indicate that GPS
is capable of resolving high frequency vibrational signature, provided the Nyquist
sampling theorem is obeyed: that is, for a band limited signal, the signal can be
recovered from discrete sampled values if the sampling is done at a sampling rate fs
≥ 2fmax where fmax is the highest frequency in the signal. This condition was met in the
experiment for which fs was 10Hz and the highest frequency recovered in the time
series was 4.3Hz. However, results from the analysis of the Republic Plaza building
experiment data seem to suggest that low frequency vibrational signature of tall
buildings cannot be easily recognized in the time and frequency domain of the data
sampled at 1Hz under normal loading conditions. It was however interesting to note
that on 'zooming in' on the section of the data sampled during the windy period, some
suspect variations could be detected on the time frequency plane. This may be an
indicator that the effectiveness of the recovery of the low frequency vibrational
signature of high-rise buildings can be enhanced through the application of special
analysis procedures such as the time-frequency domain analysis or through the use of
a simple spectrogram (Ogaja et. al., 2000).
Ince and Sahin (2000), had developed a real time GPS monitoring system
with the aid of a Kalman Filter for use in as active tectonic region near Istanbul and
its surrounding region has been developed. In order to set up a powerful control
system, a surveying and estimation method was designed and the necessary software,
called RT-MODS2 (Real Time Monitoring of Dynamic System 2) was developed.
The observation interval was one second. However, two estimation intervals are
taken into consideration, which are 5 and 3 seconds. This means that each filtering
step takes five or three observations into account. The software reads real time input
data from GPS receivers and perform deformation analyses with help of Kalman
Filter. The deformation analyses are performed in three dimensions: north, east and
14
height. The obtained magnitudes for the deformation detection are ±3.5 and ±3.0 cm
for 3 and 5 second intervals respectively. The software reads real time input data
from GPS receivers and perform deformation analyses with help of Kalman Filter
(Ince and Sahin, 2000).
Fortan program called KINDEF for 3D deformation detection via geodetic
methods have been developed. KINDEF is a kinematic deformation analysis program
and performs 3D statistical analysis to inspect the significance of geodetic network
point displacements, velocities and acceleration of displacement coming from three
repeated surveys of the same network. For deformation detection, KINDEF uses
Kinematical Single Point Model solved by Kalman Filter. In this program, movement
parameters (displacements, velocities and acceleration) are statistically tested and
moved points, velocities and accelerations of moving points are determined. The
program was written by Microsoft Fotran Visual Workbench v1.0 editor being a
window based and using maximum memory. It has only one screen facility for
representing the results of deformation detection, numerical representation. KINDEF
has been used successfully to analyze repeated GPS surveys belonging to a geodetic
network in Trabzon province, Turkey established for landslide monitoring and
control (Bayrak and Mualla, 2004)
Real-time GPS technology is an important development to aid continuous
deformation monitoring, where the timely detection of any deformation is critical.
The kinematic/dynamic parameters of deformation are computed in order to the
predict failure events. Hence the use of the Kalman Filter for the estimation of the
state vector of a deformation object is very convenient (Grewal and Andrews, 1993).
Kalman Filter is an important tool for deformation analysis combining
information on object behaviour and measurement quantities. It is applicable to the
four well-known deformation models. Kalman filtering usually requires white
measurement and process noise. Due to electronic measurement devices with high
sampling rate used nowadays, time dependent systematic deviations arise in
neighbouring epochs in a similar way, resulting in autocorrelation. Especially in case
of GPS measurements deviations due to multipath and signal propagation are
15
changing slowly, and thus the assumption of white noise is not justified. To eliminate
this deficiency within a shaping filter the state vector in Kalman filter is augmented
and thus formulating an adequate noise process (Kuhlmann, 2003).
Kalman Filter is simply and optimal recursive data processing algorithm. The
Process Noise (Q) matrix, the Dynamics (Phi) matrix, the Partials (H) matrix, the
Measurement (Z) vector, the initial Covariance (P) matrix, and the initial State (X)
vector are the parameters will be taken into account in the calculation of Kalman
Filter. Initially, the state vector and covariance matrix adds with the process noise to
calculate the gain, and then updates the covariance matrix and state vector. The stage
by stage calculation is shown (Newcastle Scientific, 2004):
State Propagation – Propagates the state vector to the time of the current
measurement.
Covariance Propagation – Propagates the covariance matrix to the time of the current
measurement and adds process noise.
Kalman Filter Gain – Derives the Gain “weighting”matrix.
Covariance Update – Update the covariance matrix.
State Vector Update – Updates the state vector with the current measurements,
weighted by the gain matrix.
The Kalman Filter provides a method for combining in an optimum fashion
all the information available up to and including the time of the latest measurement
to provide an estimate at that time. In addition to the measurements, information
about the dynamic of the process, statistics of the disturbances involved, and a priori
16
knowledge of the quantities of interest are included in the problem formulation. If the
dynamics can be described by linear differential or difference equations and if the
disturbances have Gaussian distributions, the resulting estimate is both a maximum
likelihood and minimum variance estimate (Jansson, 1998).
There are three estimation problems can be solved using Kalman Filter
algorithms (Cross, 1983):
i. Filtering – the estimate of the state vector at time tk using the
measurements at all epochs up to and including time tk.
ii. Prediction – the estimation of the state vector at time tj after the last set
of measurements at epoch tk (tj > tk).
iii. Smoothing – the estimate of the state vector at time ti using all the
available sets of measurements from the first to last epochs at times t1
and tn respectively ( t1 ≤ ti ≤ tn).
2.2 High Rise Buildings Structure Material
Buildings utilize an extensive number of building materials but their
structural systems usually have one material (either concrete or steel) as the
predominate material to carry the structural loads. Since the 1960’s there have been
an increasing use of "composite systems" in which both steel and concrete are
utilized together in ways that neither material predominates over the other (Emporis,
2004).
Some buildings were built with a skeletal framework consisting of steel
beams. The early high buildings (built after 1920) utilized cast and wrought iron in
their framing systems. Its ability to carry heavy live loads at the expense of only a
relatively small increase in dead load is used in the structural steel frame. This light
construction of beams and columns carries the whole weight of the building.
17
The other type of famous material is concrete (as building framework
material). Concrete: a hard aggregate substance made from cement, lime, crushed
rock or sand, water, and other ingredients. Concrete performs exceptionally well
under compression, but not well under tension so in most construction, including
skyscrapers, concrete is reinforced with steel bars (rebars). Most residential
skyscrapers are built with concrete frames.
Aluminium is one of the most widespread elements. Aluminium itself is soft
and quite unsuitable for use in carrying load. When alloyed with copper, silicon or
magnesium, its properties improve and fulfill the conditions required of a structural
material. There are nearly forty aluminium alloys used commercially and they
contain in all about a dozen alloying elements in varying amounts. These alloys are
about 35 percent of the weight of steel.
The Menara Sarawak Enterprise building in this case study is mainly
comprised of the concrete material, which is a common construction material for
high buildings in Malaysia. The strength of the material is good enough to support
live loads compressions. Therefore, it is more cost-effective compared to aluminium
alloy (Davis Langdon & Everest, 2004).
2.3 Deformation in Structure
It must not be assumed that once a building is constructed it is static and
immovable. Throughout its life it is in constant movement, and some understanding
of these possible movements must be gained if unsightly cracks and disfigurements
are not to render the architectural design less effective. The movements which the
whole building, or part of it, is likely to suffer can be classified as:
i. Deflections.
ii. Settlements.
iii. Deformations due to temperature and moisture changes.
18
Of these, the third applies to building materials used in cladding the structure,
and the behaviour of these materials under changing conditions of temperature and
humidity is dealt with in a companion volume of this series. It remains, then, the
deflections and displacements of one portion of a building relative to another and the
settlements either of the whole building, or part of it, relative to foundation level
needs to study. These movements may take place from various causes, but the most
important are (W. Fisher Cassie et. al., 1966):
i. Variation of live load causing beams to deflect and recover as the live load is
applied and released.
ii. Consolidation of clay or other soft soil under the foundation, with resultant
settlement.
2.3.1 Deflection of Beams
Beams are normally designed for strength; their sizes are made adequate to
withstand the stresses imposed by the loading. In withstanding these stresses, beams
of elastic material, such as steel, deflect and recover their position as the load is
applied and released. It is quite possible for the beam to be perfectly capable of
carrying its design load and at the same time to suffer a considerable deflection. If
this deflection is too large, the repeated application of live load will result in
unsightly and unwanted cracks in ceiling and wall finishes.
Before using a beam whose size has been determined from considerations of
strength alone, should be convinced that its central deflection under dead and live
load is not excessive; something less than one-three-hundredth of the span is
acceptable. In order to deal effectively with this problem it is necessary to become
familiar with the effects of variations in the factors controlling the deflection.
There are four quantities which have an influence on deflection; modulus of
elasticity (the kind of material), second moment of area (the shape and size of the
19
beam cross section), the load carried, and the length of the span. By keeping three of
these constant and varying the fourth, the effect of such variation is unobscured by
the results of other changes.
2.3.2 Settlement of Foundations
A foundation fails to fulfill its function is by showing excessive differential
settlement. The soil whose behavior determines the amount and nature of the
settlement may be considered as the mass contained within the effective bulb of
pressure. In fig 2.1, a comparison is made between the bulbs of pressure of a small
isolated footing and that of a large raft foundation. The stiff boulder clay probably
suffers little settlement, and a relatively high bearing pressure may be allowed for the
isolated footing. This high pressure cannot, however, be used for the wider
foundation, for a highly compressible layer of soft clay is cut by the bulb of pressure,
and it is this deep-seated layer whose consolidation would result in the settlement of
the building. For a rectangular or circular foundation, as has been mentioned above,
the significant portion of the bulb of pressure extends to a depth of approximately
one-and-a-half times the breadth of the foundation. For wide raft foundations the
consequent cost of deep borings may appear to be uneconomic, but there have been
instances of damaging settlement of large buildings due to the consolidation of soft
clay strata as much as 100 ft below the surface.
Figure 2.1: Comparison of the Bulbs of Pressure under a Single Footing on Test
Load and Under a Large Building (W. Fisher Cassie, 1966)
20
There is an established belief that all danger of settlement of this kind is
avoided if the load is sufficiently widely spread. That this is a fallacy can be seen if
the bulb of pressure for a concentrated load is compared with that for the same load
applied over a larger area. If the bulbs of pressure of both are drawn to scale, it is
clear that the spread footing induces less intense stresses near the surface. At the
considerable depth at which a compressible layer may lie, however, the concentrated
and the spread loads exert a similar intensity of pressure. The settlement which
occurs because of the consolidation of a deep-seated layer is not always effectively
reduced by spreading the load.
2.3.3 Wind Loading Problem
Any structure which is built upon the earth's surface must be capable of
withstanding the loads imposed on it by the weather. The wind, in particular,
constitutes one of the major forms of structural loading and even moderate winds are
capable of imposing high forces on structures. As a result, most building codes of
practice incorporate fairly lengthy sections devoted specifically to those aspects of
the design and construction of buildings which are concerned with the resistance of
wind load.
The loads imposed on structures by the wind usually act horizontally and they
cannot normally be resisted by the main structural system, which is designed to carry
the vertically downwards acting gravitational loads. Two distinct structural systems
are therefore required in a building to ensure stability, one to resist vertical loads and
one to resist horizontal loads due to wind. They may both be present in one
component, as is, for instance, the case with a masonry pier which is stable both
horizontally and vertically, or they may be separate as in a lattice tower in which the
columns resist primarily the gravitational loads while the diagonal members provide
stability against lateral loads
21
2.4 Review of GPS
Nowadays, the Navigation Satellite Timing and Ranging Global Positioning
System (NAVSTAR GPS, or commonly know as GPS) has become one of the most
successful extraterrestrial positioning technique. A definition given by Wooden
(1985) reads: “The Navstar Global Positioning System (GPS) is an all-weather, space
based navigation system under development by the U.S. Department of Defense
(DoD) to satisfy the requirements for the military force to accurate determine their
position, velocity and time in a common reference system, anywhere on or near the
earth on a continuous basis.”
Since the DoD is the initiator of GPS, the system primary goals were military
usage only. However the U.S. Congress, with the guidance from the U.S. President,
directed DoD to promote the system to civil usage. This has given a great impact of
technology on geodetic surveying.
The GPS system component consists of three segments, space segment,
control segment and user segment (see Figure 2.2).
Figure 2.2: GPS Segments (21CEP, 2000)
The Space Segment consists of the constellation of spacecraft and the signals
broadcast that allow users to determine position, velocity and time. The fully
deployed GPS space segment consists of 24 satellites with three actives spares in six
orbital planes, four satellites in each plane. The satellite orbits repeat almost the same
22
ground track once each day. The orbit altitude is such that the satellites repeat the
same track and configuration over any point approximately each 24 hours (4 minutes
earlier each day).
The Control Segment consists of facilities necessary for satellite health
monitoring, telemetry, tracking, command and control, satellite orbit and clock data
computations and data up linking. There are five ground Monitor Stations (Hawaii,
Colorado Springs, Ascension Island, Diego Garcia and Kwajalein) - three Ground
Antennas, (Ascension Island, Diego Garcia, Kwajalein), and a Master Control
Station (MCS) located at Schriever AFB in Colorado.
The User segment consists of the GPS receivers and the user community.
GPS receivers convert satellites signals into position, velocity, and precise timing to
the user. A minimum of four satellites are required to compute the four dimensions
of X, Y, Z (position) and Time. The user community is divided into two main
categories, which are the military user and the civilian user. However the diversity of
the uses is matched by the different type of receivers available. In brief, the receiver
differences are based on the type of observables (i.e. code pseudo ranges or carrier
phases) and on the availability of codes (i.e. C/A-code or P-code).
2.5 GPS Positioning Techniques In general, GPS positioning techniques can be divided into two basic
categories, which are static and kinematics. Static denotes a stationary observation
while kinematics implies motion.
Static GPS positioning technique is perhaps the most common method used
by surveyors because of the high accuracies it can achieve. The principle is based on
the vector solution between two stationary receivers. This vector is often called the
“baseline” because of its similarity to triangulation baselines. The station coordinate
differences are calculated in terms of 3D, earth centered coordinate system that
23
utilizes X-, Y-, and Z-values based on the WGS 84 geocentric ellipsoid model.
Generally, one of the GPS receivers is positioned over a point which coordinates are
known, and the second is positioned over another point which coordinates are
desired. It can be single or multiple baseline observation, where multiple solution
concerns more than two observation points. Station occupation time is dependent on
baseline length, number of satellites observed, and the GPS equipment used. In
general, 30 min to 2 hr is a good approximation for baseline occupation time for
shorter baselines of 1-30 kilometers.
On the other hand, kinematics method involves one stationary and one
moving receiver. The two receivers perform the observation simultaneously. The
accuracies of this method are lower compared to static surveying method that can
reach up to millimeter accuracies.
Today new GPS surveying methods have been developed with the two
liberating characteristics of: (i) static antenna setups no longer having to be insisted
upon, and (ii) long observation sessions no longer essential in order to achieve
survey level accuracies. These modern GPS surveying techniques includes: (i) Rapid
static positioning technique, (ii) Reoccupation technique, (iii) Stop and Go technique,
and (iv) Kinematics positioning technique. All of these modern GPS surveying
method require the use of specialized hardware and software, as well as new field
procedures.
2.5.1 Real Time Kinematics (RTK-GPS) RTK-GPS is one of the important or new kinematics techniques in GPS
positioning technique. It helps a lot in deformation monitoring. This real time
application has been widely used in various survey applications and navigational
purposes, regardless on land, at sea or in the air (Rizos, 1999). RTK-GPS started in
the early nineties (Zhang and Robert, 2003). RTK-GPS deformation survey is more
to single epoch observation technique.
24
This deformation survey using GPS technology has higher automation
degree. RTK-GPS could achieve accuracy level of ± 2 cm + 2 ppm. In RTK-GPS
configuration, a receiver is placed on the reference point with known coordinates as
reference station. This reference station will continuously transmit correction
message to rover receiver (see Figure 2.3). Its observation can be done continuously
(24 hours) or in a short observation period with its sampling rate as small as 0.1s
(10Hz).
Figure 2.3: RTK-GPS Observation Configuration
RTK-GPS process carrier phase observation in real time to produce
coordinates. RTK-GPS differential positioning can be done as long as rover can
receive signals from 4 satellites and differential signal from reference station via
radio-link (Talbot, 1993). The precise position of rover is obtained after ambiguity is
solved. There are many ambiguity solving methods and they are applied depending
on the type of receivers, for example: Least-Square AMBiguity Decorrelation
Adjustment (LAMBDA), Fast Ambiguity Resolution Approach (FARA), Fast
Ambiguity Search Filter (FASF), On-The-Fly Ambiguity Resolution (OTFAR) and
so on. The method applied in this study is OTFAR (Talbot, 1993) in which the true
ambiguity is obtained during the observation. Therefore, topographical data
collection (3D coordinates for deformation monitoring) were carried out
continuously.
2.6 Error Sources in GPS Measurement
Field observations are not prefect, and neither are the recordings and
management of observations. The measurement process suffers from several error
25
sources. Repeated measurement does not yield identical numerical values because of
random measurement error. Therefore, there are many type of error in GPS
measurement which will be affect accuracy and precision of the GPS observation.
i. Orbit Determination
The positions of the satellites can be determined by one of the two different
ways. First,, by using the orbital information contained in the broadcast ephemeris
which is transmitted from the satellite in the navigation message (commonly termed
as broadcast orbits). The Keplerian and perturbation parameters (also referred as
correction terms) contained in the orbital information are used to compute the
positions of the satellites in ECEF reference frame. The other way of determining
satellite positions is using precise ephemeris. On the contrary, precise orbits are
based from nearly 200 globally-distributed International GPS Service (IGS) stations
and computed by different analysis centres around the world.
ii. Clock Error Satellite
The satellite clock error is due to the instabilities in the GPS satellite
oscillators. The GPS satellite clock can be determined from the satellite clock
information contained in the navigation message. The clock parameters from the
broadcast ephemeris are used to compute the correction to GPS time for each satellite
(Leick, 2004). Meanwhile, the clock correction from the broadcast ephemeris is not
accurate since the parameters are essentially predicted. Precise clocks are used in
order to achieve a better estimation of clock error which is obtainable from the same
agencies as the precise orbits and also available in three forms as with the precise
orbits.
26
iii. Clock Error of Receiver The third largest error is the receiver clock error. A user equivalent range
error (UERE) from 100 meters to 10 meters may be attributed to receiver clock error,
depending on the oscillator type. Both a receiver’s measurement of phase differences
and its generation of replica codes depend on the reliability of its internal frequency
standard, its oscillator.
iv. Ionosphere Delay
Ionosphere is a region of the atmosphere that stretches roughly from 50 km to
1000 km above the earth surface. Ionosphere is a dispersive medium characterized by
ionized gas and free electrons. The delay on the GPS signals induced by the
ionosphere is dependent upon the total electron content (TEC) along the line of sight
vector between the receiver and the satellite. Due to the dispersive nature of the
ionosphere, GPS signals delay can be estimated by forming a linear combination of
L1 and L2 measurements, commonly termed as L3 or ionosphere free combination, if
the measurements on both frequencies are available, as (Leick, 2004). In general, the
ionosphere error is eliminated through differencing and the remaining residuals are
neglected for short baselines, typically 1 to 30 km
v. Troposphere Delay
Troposphere is the non-ionized portion of the atmosphere extending from the
earth’s surface up to about 50 km. The troposphere is generally modeled in two
components, the hydrostatic and wet components. The magnitude of the troposphere
delay is a function of atmospheric pressure, temperature, relative humidity, satellite
elevation and altitude which are usually handled through modeling or differencing
for short baselines. Over the past decades, various troposphere models have been
proposed and amongst the most common ones are Hopfield, modified Hopfield and
27
Saastamoinen. These models differ principally based upon the assumptions made on
the vertical refractivity profiles and the vertical delay mapping with respect to the
elevation angle (Leick, 2004).
vi. Multipath
Multipath occurs when the same GPS signal is acquired by the receiver from
reflected path apart from the direct-single path. The reflection of the signals may
caused by a variety of surrounding objects such as buildings, vehicles, ground and
water surfaces. The phase delay of the reflected signal relative to the direct signal
results in a wrong estimate of the time of travel of the GPS signals. Multipath is
difficult to model for the general case since it naturally depends on the environment
surrounding the GPS antenna. Furthermore, even differencing cannot eliminate
multipath errors since they are not spatially correlated. Therefore, GPS antenna
design can play a role in minimizing the effect of multipath. Ground planes, usually a
metal sheet of about a square meter or so, are used with many antennas to reduce
multipath interference by eliminating signals from low elevation angles.
vii. Receiver Noise
Due to the physical limitations of the receiver, noise is generated during the
measurement process, both on carrier phase and code. The noise is essentially caused
by tracking loop jitter and therefore the code and the carrier phase noise is not
correlated since both employ a different tracking loop. The receiver noise is
considered random in nature and typically smaller in magnitude compared to other
error sources. The most common way to effectively estimate receiver noise is
through a zero-baseline test.
CHAPTER 3
THE APPLICATION OF KALMAN FILTER IN DEFORMATION STUDY
3.1 Introduction The Kalman Filter is consider a vector of parameters (the state vector)
changing with time and sets of measurements observed at different epochs, ti, that are
related to the parameters by a linear model (the primary model). If the parameters are
changing with time can model both for deterministic and stochastic effects (the
secondary model), then the Kalman Filter provides a set of algorithms for the
estimation of the state vector at any point in essence. This process shows that
estimates of the state vector are unbiased and have minimum variance.
The Kalman Filter provides a method for combining in an optimum fashion
all the information available including the time of the latest measurement to provide
an estimate at that time. In addition to the measurements, information about the
dynamics of the process, statistics of the disturbances involved, and a priori
knowledge of the quantities of interest are included in the problem formulation. If the
dynamics can be described by linear differential or difference equations and if the
disturbances have Gaussian distributions, the resulting estimate is both a maximum
likelihood and minimum variance estimate. As the name suggest, a maximum
likelihood estimate has a higher probability of being correct than any other.
In order that the future state of the system is determinable from its current
state and future inputs, the dynamical behaviors of each state variable of the system
must be a known function of the instantaneous values of other state variables and the
system inputs. The state space model for a dynamic system represents the functional
29
dependencies in terms of difference equations (in discrete time), which are called its
state equations.
A simple way to think of the Kalman Filter may be the following: over a
series of time epochs, certain time-dependent parameters are changing. Without any
additional information (i.e. observations) about these parameters, they are expected
to follow a path that is dictated by specific differential equations. The Kalman Filter
is a way to take these differential equations into account and combine them with
time-dependent observations of, or related to, the parameters and estimates the values
of the parameters at specific time epochs.
The dynamic model is used to describe the motion and the dynamic noise
whose instantaneous position and velocity are sought, which is the key difference
from static application. A general form of the dynamic model for a system can be
represented by the state space model, in which a set of first order linear differential
equations express deviations from a reference trajectory, i.e. (Gelb et. al., 1974):
x’(t) = F x(t) + G w(t) (3.1)
where
x - state vector of the process (n x 1)
x’ - time derivative of the state vector (n x l)
F - is the system dynamic matrix (n x n)
G - is the coefficient matrix of the random forcing function (n x n)
w - is the system noise vector (n x l) which is usually assumed to be white
noise
t – is time
The solution of the first order homogeneous differential equation, x’(t) =Fx(t), can be write as:
x(t) = Φ(t, to) x(to) (3.2) where Φ is called transition matrix and satisfies the equations:
30
Φ (to, to) = I (3.3)
(3.4) Φ (t, to) = F Φ (t, to) The particular solution of Eq. (3.1) with random forcing function w(t) can be written as:
hich is often called the matrix superposition integral (Gelb et. al., 1974). The
here it is assumed that x(to) and w(t) are uncorrelated, and Cx(to) is the variance
here Q is called the spectral density matrix. For a time invariant system (i.e F is a
here ∆t = t - to. The above equation can be expanded into a Taylor series:
(3.5)
w
corresponding variance covariance matrix of the state is given by:
w
covariance matrix of the initial state x(to). When w(t) is a white noise random forcing
function, Eq (3.6) further reduces to:
w
constant matrix), the transition matrix Φ (t, to) in only a function of the time difference (t-
to). In this case, the transition matrix can thus be expressed as the matrix exponential (Gelb
et. al., 1974):
w
(3.6)
(3.7)
(3.8)
(3.9)
31
where I is the identity matrix. For a small time interval ∆t, it may be
sufficie
ra s, we have:
here:
.1.1 The Discrete Kalman Filter Algorithm
The discrete Kalman Filter is an optimal estimator for a linear dynamic
system
ensional state vector at time k, Φk is the known (n x n)
transiti
ntly approximated by:
nsforming Eq. (3.5) and (3.7) into discrete form
(3.10)
(3.11)
(3.12)
(3.13)
T
w
where ∆t = tk – tk-1
(3.14)
3
which can be described in state space by the following equations (Gelb et. al.,
1974).
(3.15)
(3.16)
Where xk is the n-dim
on matrix describing the state transition from time (k-1) to k, zk is the
measurement vector, and Hk is the (m x n) design matrix. It is also assumed that wk
and vk are white noise processes which have the following properties:
32
(3.17)
Q is the process noise matrix and accounts for the uncertainty in the state
model, and is usually included in the terms that drive the state model. Note that a
white noise sequence is a sequence of zero mean random variables that are
uncorrelated timewise. However, the variables of the sequence may have a mutual
nontrivial correlation at any point in time tk (Brown and Hwang, 1992).
The optimality of the Kalman Filter is achieved by meeting the criteria of
linearity, unbiasedness and minimum variance. Firstly, the filtered estimate of xk at
epoch k based on all the measurements up to and including epoch k is a linear
combination of the a priori estimate of xk and the measurement at epoch km zk
(linearity). Secondly, the estimated vector x’k has a mathematical expectation which
is equal to the true value xk (unbiasedness). Thirdly, the filtered vector x’k should
have a minimum variance (minimum variance).
Under the assumptions of the initial state xo and its covariance matrix Co
being previously given, wk and vk being white noise processes, uncorrelated with xk
and with each other, the Kalman Filter estimates for model Eq(3.15) and Eq(3.16)
are given by the time update equations (Brown and Hwang, 1992):
(3.19)
(3.18)
and the measurements update equations:
(3.20)
33
where:
(3.21)
and:
(3.22)
where xk- is the updated estimate and Kk is the blending factor. The symbols – and +
denote the best estimate priori to and after the measurement update at time tk
respectively and ^ denotes an estimate. For Ck-, another version can be used which is
valid for any gain, suboptimal or otherwise:
(3.23)
The innovation sequence is the primary source of information about the
performance of a Kalman Filter and is defined as
(3.24)
where vk- is actually the predicted residual vector of observations. The reason it is
called the innovation sequence is that it represents the new information brought in by
the latest observation vector zk.
3.1.2 The Extended Kalman Filter
In practical applications of estimation theory one often encounters models in
which nonlinearities are present. The theory of nonlinear estimation has been
developed for special cases, but its practical application usually requires that the
underlying probability density functions explicitly be included in the estimation
algorithm (Gelb et. al., 1974). This latter constraint constitutes a requirement that
often is impossible to satisfy or is computationally expensive to implement. The
difficulties in the nonlinear case lead one to look for linear models which provide a
suitable approximation to the nonlinear equations. It is an accepted practice to
approximate the nonlinear models by suitable chosen linear ones. The linearization is
34
performed with respect to a computed state vector estimate. The Kalman equations
for propagating and updating the state vector and its covariance matrix, although
derived under assumptions of linearity are typically applied as an approximation to
the nonlinear case (Gelb et. al., 1974). The resulting filter equations are usually
referred to as an extended Kalman Filter (EKF).
The idea of the extended Kalman Filter is to use the ideas of Kalman Filtering
for a nonlinear problem. The filter gain is computed by linearization of the nonlinear
model. The extended Kalman Filter, in contrast to the Kalman Filter for linear
systems, is not an optimal filter, since its derivation is based on approximations.
The extended Kalman Filter is the most widely used approximate filter. It is
based on the linearization of the system dynamics and the measurement model
around the estimated state. That is the partial derivates in the design matrix
(Eq(3.11)-(3.16)) are evaluated along the trajectory that has been updated with the
filter’s estimates. These, in turn, depend on the measurements, so the filter gain
sequence will depend on the sample measurement sequence realized on a particular
run of the experiment. Hence, the gain sequence is not predetermined by the process
model assumptions as in the usual Kalman Filter.
3.2 Advantages, Problems and Disadvantages of Kalman Filter
Real time GPS technology is an important development to aid continuous
deformation monitoring, where the timely detection of any deformation is critical.
The kinematic parameters of deformation are computed in order to the predict failure
events. Hence, the use of the Kalman Filter for the estimation of the state vector of a
deformation object is very convenient (Grewal and Andrews, 1993). The elements of
the state vectors in the Kalman Filter are the unknown but normally these are
position of the object and are important for studying the behavior of deformations.
35
Least square treats each epochs independently which means that it does not
use the knowledge of the motion of the system. Thus, Kalman Filter has some
advantages over least squares. Often it is possible to make a very accurate prediction
of where the point will be at any epoch using just the previous position and the
estimated motion. Least square is a safe option but it does not have the potential
accuracy of Kalman Filtering (Jansson, 1998).
Kalman Filter is a more powerful tool than least squares to predict the
unknown’s parameters for quality control, especially in real time applications
(Jansson, 1998). Much smaller outliers and error can be detected by Kalman Filtering
if compared with least squares. However, it is recommended that least squares also
be carried out at every epoch in order to identify large outliers. This is because
Kalman Filtering can be rather time consuming from computational point of view
and any initial cleaning that can be done by other methods will increase its
efficiency. Kalman Filter can accept data as and when it is measured. With simple
least squares, data has to be reduced to a specific epoch. Subsequently, a Kalman
Filter can cope well with data arriving as a more or less continuous stream.
Kalman Filter is based on some fundamental assumptions. If all the
assumptions are met it can offer optimal estimation and prediction. These
assumptions are that unmodelled measurement errors are white (i.e uncorrelated)
noise and that unmodelled error dynamic white (i.e uncorrelated) too. A successful
application of the Kalman Filter is strongly dependent on whether the system
dynamic is perfectly modeled in the mean sense. In practice, disturbances such as
wind and current may have random variations superimposed on changing mean
values (trend). This suggests that it is very difficult, if not impossible to model the
system dynamics of a kinematics positioning system perfectly because the
environment always changes. As far as measurement noise is concerned the system
will require knowledge of the random errors, which will be assumed Gaussian.
36
3.3 Application of Kalman Filter in Deformation Monitoring
The problem of deformation monitoring is essentially that of detecting the
movement (or lack of movement) of a set of object points compared with a number
of fixed points. Observations, which connect the fixed and object points in a strong
network, are made at separate epochs, usually at regular intervals of time if possible.
In classical deformation analysis, the network is adjusted for individual epochs. From
the results of two such adjustments at consecutive epochs, the displacement vector
for each object point is computed and if appropriate, significance testing applied.
Pelzer (1986) proposes an advanced deformation analysis using Kalman
filtering. The basic idea is to model the trajectory of each object point with a special
kinematics model:
Yi = Yo + Y’(to)∆t + Y’’(to) ∆t2/2 (3.25)
The state vector is given by
Y(t) position
X(t) = Y’(t) = velocity (3.26)
Y’’(t) acceleration
This is similar to the simple model of a constant velocity vessel except that
acceleration is now included as a parameter in the state vector and the model is
applied to each and every point in the network. The fixed points are easily catered for
by setting
Y’i = 0 and Y’’i = 0
The Kalman Filter can be used to analyze for changes in the steady state
conditions (eg changes in velocity) in addition to changes in position. The analysis is
carried out in the following steps:
i. Prediction of the state vector to an actual epoch (of observation).
ii. Comparison with actual observations (“innovation”).
iii. Updating of the state vector from the observations.
37
iv. Testing the change in state vector resulting from the update for any
significant change in state.
This analysis is repeated for all epochs, using the increased system noise
determined in step(ii) if the innovation was significant at the previous epoch. If there
are only two epochs to analyze, this analysis is identical to the classical approach.
Therefore the Kalman Filter is only worth applying if three or more epochs are
observed. In which case, the advantages of this method of analysis over the classical
method are:
i. The deformation model includes position, velocity and acceleration.
ii. A least squares solution at any epoch uses historical information as well
as the actual observation at that epoch.
iii. The calculation of the actual state of the point field includes position,
velocity and acceleration.
This method of deformation analysis is applicable to many problems such as
crystal movement, landslides, ice flows and engineering structures. However, as
Pelzer suggests a dynamic model that takes into account all the forces acting on each
objects point (or the land mass around each object point) would be far superior. But,
at this point in time our knowledge of the forces involved is limited, therefore the
Kalman Filter provides a practical alternative to the classical method of deformation
analysis.
CHAPTER 4
FIELD METHODOLOGY AND DATA PROCESSING
4.1 Introduction
A geodetic network is defined as a geometric configuration of three or more
control survey points that are connected either by geodetic measurements or by
astronomical observations or space-based techniques. The network design will
answer the essential questions of where the network points should be placed, how
many control points should be established, and how a network should be measured in
order to achieve the required accuracy in a cost-effective way. At least two epochs
will be carried out in the deformation study. Before such a campaign could begin to
be implemented, it is necessary to set up a deformation network which consists of
selected reference stations (datum) and the monitoring (object) points with respect to
the corresponding engineering structural design, i.e. size and shape of structures.
The purpose of field data acquisition is to obtain the needed and relevant data
for the study. Besides that, it enables us to experience the fieldwork procedure and
learnt to operate the equipments involved.
The survey data had been downloaded into the computer. All survey data
were processed and adjusted by least square adjustment program. During the data
processing, various statistical tests were carried out to check the data quality and
generate the adjusted coordinates along with covariance matrices information. The
adjusted coordinates and its covariance information were then combined with similar
data from the second epoch to complete a deformation analysis. The deformation
39
analysis is to identify stable points in the monitoring network by performing single
point test on them.
4.2 The Menara Sarawak Enterprise Monitoring Network
The points’ location for the monitoring network was chosen according to the
following criteria:
i. Easily accessible by transportation (vehicle)
ii. Safe and free from any public or other disturbances.
iii. Development is particularly nil/minimum.
After reconnaissance, the location of control and monitoring stations were
selected. For GPS observations, the choices were made based on clear sky view of
the stations, free of any disturbance and its stability. The preparation includes
gaining the information about DSMM GPS reference station (J416) at Stulang Laut,
Johor Bahru which located near the research area (see Figure 4.1).
F
igure 4.1: DSMM Geodetic Control
(GPS) Station, J416
40
The next stage of the study is to establish the control and monitoring stations.
The location of some control stations are inescapable because the surroundings of
Menara Sarawak Enterprise building comprised of houses, shops and construction
area – see Figure 4.2.
Figure 4.2: Location of Control and Monitoring Stations
After that, each station is given its own ID to be recognizable. The control
station monuments then were left alone for almost one month (after the stations
established) before the GPS observation has been carried out -see Figure 4.3 and 4.4.
The purpose is to allow any possible settlement until the stations became stable.
Figure 4.3: Base 1 (B1)
Figure 4.4: Base 2 (B2)
41
In this study, there are two monitoring stations: Rover 1 (R1) and Rover 2
(R2) – see Figure 4.5 and 4.6. For the monitoring stations, special gadgets with
bracket shape are made so that it can be easily mounted on the building concrete
guardrail – see Figure 4.7. The GPS receiver was placed on top of this light PVC
bracket.
Figure 4.5: Rover 1 (R1)
Figure 4.6: Rover 2 (R2)
Figure 4.7: Design of Rover Monument
42
4.3 Instruments Used for GPS Observation As mentioned above, the GPS receiver instruments were used in this study for
three dimensional positioning. The instruments used are Leica GPS System 500 and
Trimble 4800 GPS Series.
Figure 4.8: Leica GPS System 500 Receiver
Leica GPS System 500 (see Figure 4.8) is a dual-frequency GPS system that
able to operate in Real-Time Kinematics (RTK) mode with accuracy up to 10mm +
1ppm (RMS). It has twelve channels of both L1 and L2 frequency, which means it
can receive up to 12 satellites signals at once. The system can be easily operated
using the user-friendly interaction menu provided. The system also comes with a
software name SKI-Pro that runs all the downloading and processing applications.
The specification of Leica GPS System 500 is shown in Appendix A.
Trimble 4800 GPS Series (Figure 4.9) is a dual-frequency GPS system
capable of receiving 18 channels of satellites signals. Like Leica GPS System 500,
Trimble 4800 GPS System can also perform kinematics survey either in real-time
(RTK) or post-processed using OTF technique. The processing application is run by
a software name Trimble Geomatics Office (TGO). The specification of Trimble
4800 GPS System is shown in Appendix B.
43
Figure 4.9: Trimble 4800 Series GPS Receiver.
4.4 GPS Instruments Calibration
RTK-GPS is one of the latest methods that have been used in structural
deformation monitoring application. Therefore, effective research should be carried
out to test the feasibility of this technique in structural monitoring. To perform this
test, RTK-GPS baseline is to study, especially its accuracy. The distance and
accuracy for base and rover RTK-GPS communication also had been studied to
verify its effectiveness in the observations.
4.4.1 Test on RTK – GPS Performance
The hardware has its own specifications, such as Trimble 4800 System is able
to give 1 cm accuracy horizontally (northing and easting) and 2 cm vertically. This
statement is defined by Trimble Manufacture Company in the hardware’s manual. To
verify this statement, a test had been carried out. The test was carried out in
Universiti Teknologi Malaysia and the instruments used were two Trimble 4800
Series. The working procedures for this test are accordance to Field Test of Trimble
4000 Real Time Kinematics GPS Survey System (Jay and Ralph, 1998).
In this test, there were 2 stations, namely UTMB and UTMR and their
coordinates were derived from 4 RTKNet stations (JHJY, KLUG, KUKP, TGPG)
44
which are located in Johore State (see Figure 4.10). The observation period was 1
hour (in the morning) using Static-GPS Positioning technique. Later, the observation
data was processed using Trimble Geomatics Office software to obtain the
coordinates for UTMB and UTMR station (see Table 4.1 and 4.2).
Distance:- KLUG – KUKP : 78024.807m KUKP – TGPG : 72972.326m TGPG – KLUG : 114230.042m KLUG – JHJY : 75944.795m KUKP – JUJY : 4432.143m TGPG – JHJY : 39424.493m KLUG – UTMR : 62192.202m KUKP – UTMR : 33727.689m TGPG – UTMR : 56236.068m JHJY – UTMR : 17287.818m KLUG – UTMB : 62232.947m JHJY – UTMB : 17245.411m TGPG – UTMB : 56191.770m KUKP – UTMB : 33734.472m UTMB – UTMR : 45.296m
Figure 4.10: Coordinates of UTMB and UTMR Derived from RTKNet Stations
Table 4.1: Adjusted Grid Coordinates from Static Processing
Point Name Northing N
error Easting E error Elevation e error
JHJY 169946.585m 0.000m 644467.239m 0.000m 31.779m 2.293m
KLUG 223992.242m 0.000m 591129.278m 0.000m 68.656m 2.293m
KUKP 147462.343m 0.000m 606279.468m 0.000m 9.248m 2.293m
TGPG 151206.592m 0.000m 679145.901m 0.000m 9.303m 2.293m
UTMR 173610.342m 0.003m 627575.341m 0.005m 144.176m 2.294m
UTMB 173587.812m 0.003m 627613.781m 0.005m 136.057m 2.294m
Table 4.2: Adjusted Geodetic Coordinates From Static Processing
Point Name Latitude N error Longitude E error Height h error
JHJY 1°32'12.51698"N 0.000m 103°47'47.51425"E 0.000m 39.189m 0.000m
KLUG 2°01'31.36066"N 0.000m 103°19'00.52466"E 0.000m 73.588m 0.000m
KUKP 1°19'59.78966"N 0.000m 103°27'12.35915"E 0.000m 15.422m 0.000m
TGPG 1°22'02.67821"N 0.000m 104°06'29.73440"E 0.000m 18.093m 0.000m
45
UTMR 1°34'11.56659"N 0.003m 103°38'40.90764"E 0.005m 150.960m 0.012m
UTMB 1°34'10.83360"N 0.003m 103°38'42.15176"E 0.005m 142.843m 0.012m
After that, the coordinates of UTMB station was defined as base station for
continuous RTK-GPS observations meanwhile the UTMR station as rover. The
coordinates of UTMB station (refer Table 4.2) was input into Trimble controller GPS
during observations. The coordinates of UTMR will be obtained from GPS
observation data. The observation period for this technique was one hour and had the
same configurations and conditions with the Static-GPS Positioning technique (see
Appendix C). Another set of observations data was obtained through observation
with the same configuration as the continuous RTK-GPS Positioning technique. The
only difference for this observation was that the observation period was shorter,
about half hour only – refer Appendix D.
4.4.2 Test on Accuracy of RTK-GPS Baseline
It has been mentioned by the Trimble Manufacture Company that range
between base and rover RTK varies depends on the radio-link used in local terrain
and operating conditions. Therefore distance and accuracy of RTK-GPS using
Pacific Crest radio link as communication instrument between base and rover RTK
should be studied to ensure its effectiveness.
There were 3 stations named T200, T300, and TR2300 had been established.
The coordinates derived from 2 stations (TRS Station in UTM and RTKNet Station –
JHJY in Johor Jaya) – see Figure 4.11. Fast Static Positioning technique is used to
determine the coordinates of these 3 stations. The observation period was around 20
minutes. After that, the observation data was processed using Trimble Geomatics
Office software to obtain the coordinates (see Table 4.3 and 4.4).
46
Figure 4.11: Coordinates of T200, T300 and TR2300 Derived from TRS
Station and JHJY RTKNet Stations
Table 4.3: Adjusted Grid Coordinates from Fast Static Processing
Point Name Northing N error Easting E error Elevation e error TR2300 173621.086m 0.003m 627580.843m 0.005m 144.176m 0.459m
T200 173473.388m 0.004m 627708.013m 0.006m 119.699m 0.459m
JHJY 169946.585m 0.000m 644467.239m 0.000m 31.779m 0.459m
TRS 172558.835m 0.000m 626589.385m 0.000m 51.785m 0.459m
T300 173354.433m 0.009m 627717.586m 0.010m 110.610m 0.460m
Table 4.4: Adjusted Geodetic Coordinates from Fast Static Processing
Point Name Latitude N
error Longitude E error Height H
error
TR2300 1°34'11.91650"N 0.003m 103°38'41.08547"E 0.005m 150.960m 0.010m
T200 1°34'07.10927"N 0.004m 103°38'45.20255"E 0.006m 126.489m 0.013m
JHJY 1°32'12.51698"N 0.000m 103°47'47.51425"E 0.000m 39.189m 0.000m
TRS 1°33'37.31182"N 0.000m 103°38'09.02300"E 0.000m 58.544m 0.000m
T300 1°34'03.23604"N 0.009m 103°38'45.51421"E 0.010m 117.402m 0.025m After that, the station T200 acted as base station for continuous RTK-GPS
observation meanwhile the TR2300 station as rover. The coordinates of T200 station
had been input into GPS controller during observations based on the coordinates
from the processing (see Table 4.4). The coordinates for TR2300 station will be
obtained from GPS observation data. The observation for this technique went on
47
about 5 minutes (refer Appendix E). Then, GPS instrument of base T200 was
transferred to T300 in which the latter became base station for the next continuous
RTK-GPS observation and the measurement was carried out as explained before (see
Appendix F). Then, the observation was carried out in canopy area which formed
obstruction between base and rover for RTK-GPS observation.
4.5 GPS Observation
TGO software version 1.6 offers a software interface for planning purpose.
Observer will know about the DOP value, sky plot data, numbers of satellite (see
Figure 4.12), visibility rate etc. The smaller the Position Dilution of Precision, PDOP
(value <4), the better the results and higher satellites visibility will helps observer for
data collection in suitable time – refer Figure 4.13 and 4.14 (Trimble, 2001).
Figure 4.12: Information of Satellite Visibility on 21/12/2004.
48
Figure 4.13: Information of DOP Horizontal on 21/12/2004.
Figure 4.14: Information of DOP Vertical on 21/12/2004.
4.5.1 GPS Network of Coordinates Transfer Conventional static GPS method is applied to determine the locations
(coordinates) of control stations. The data observation was carried out on 21/12/2004
for half an hour, which involved three Trimble 4800 series GPS receivers and
RTKNet station. Station occupation time for static baseline is dependent on baseline
length, number of satellites observed, and observation period of 30 min to 2 hr is a
good approximation for baseline occupation time for shorter baselines of 1-30
kilometers (South Dakota Department of Transportation, 2005). Therefore, the
timeline of this observation is shorter (30 minutes) because the number satellite more
49
than 7 during the GPS observation and all the stations including control and
monitoring stations are nearby. The Figure 4.15 shows the GPS network of
coordinates transfer from DSMM geodetic control station (J416) and RTKNet
Station (JHJY) to control stations of this study, B1 and B2.
100.331m (B1 – J416)
8750.543m (JHJY-B2)
554.809m (B2 – B1)
8685.248m (JHJY – J416)
493.497m (B2 – J416)
8590.964m (JHJY – B2)
Figure 4.15: GPS Network of Coordinates Transfer
4.5.2 GPS Monitoring Network
The field survey had been carried out in two different epochs, in the months
of December 2004 and April 2005. The first epoch data observation was done on
22/12/2004 – 23/12/2004, for a period of 2 days whereby the second epoch was done
on 28/04/2005 – 29/04/2005, for 2 days (see Table 4.5).
Table 4.5: GPS Observation Schedule of Menara Sarawak Enterprise Building
Days Tasks Remarks
22/12/2004 ½ hour static GPS data
Deformation Monitoring Network: B1, B2, R1, R2
23/12/2004 1 hour continuous RTK-GPS data
Deformation Monitoring Network: B1 and B2 as Base; R1 and R2 as Rover
50
28/04/2005 ½ hour static GPS data
Deformation Monitoring Network: B1, B2, R1, R2
29/04/2005 ) 1 hour continuous RTK-GPS data
Deformation Monitoring Network: B1 and B2 as Base; R1 and R2 as Rover
Figure 4.16: GPS Monitoring Network
The GPS observation was carried out using two units of Leica GPS System
500 and two units of Trimble GPS 4800 System. All the stations included B1, B2,
R1, and R2 stations had been observed for at leas half an hour using static survey
method (see Figure 4.16).
The GPS instruments were used to carry out Real Time Kinematics (RTK)
GPS observation for the next day. The instruments were set up at monitoring stations
(R1 and R2) and control stations (B1 and B2) respectively with one hour period for
RTK-GPS observation procedure. The study used 2 monitoring stations (2 Rovers)
which is located on both side of the building’s rooftop to improve the strength of
monitoring network. In other words, if R1 detect the vibration and it can be proven
by another rover, R2.
4.6 Data Processing and Adjustment
All GPS data for both epochs were downloaded into computer and then
processed using the Leica SkiTM Pro and Trimble Geomatics Office software. The
option adopted for the GPS data processing and adjustment is summarized in Table
4.6. All observations were referred to the WGS84 coordinate system. All GPS
51
baselines vector were processed and adjusted in 3 dimensional with minimal
constrain network adjustment. The SkiTM Pro and Trimble Geomatics Office
software are designed for the GPS receiver and it can be used to process a large
number of data with minimum step in Microsoft Windows Environment. The
software can also process all types of GPS survey methods such as static, stop and
go, rapid static and kinematics survey. Network adjustment report of Trimble
Geomatic Office is shown in Appendix G.
Table 4.6: Data processing Options
Items Assigned Options
Cut-off angle 15 degrees
Tropospheric model Hopfield
Ionospheric model Computed model
Ephemeris Broadcast
4.6.1 Trimble Geomatics Office Data Downloading
During the RTK-GPS surveys, observation data was collected using Trimble
4800 Series in RTK mode. When GPS observations are in RTK mode, post process
is not necessary.
First, a new project file is created using Trimble Geomatics Office (TGO)
software. Then, observation data was downloaded from the controller using Survey
Device (TGO Data Transfer) by selecting the file name which was created during
observations. Then, data will be displayed in graphic. Finally, coordinates data was
exported into a work sheet according to the format required. The graphical of
Trimble Geomatics Office data downloading is shown in Appendix H.
52
4.6.2 Leica Ski Pro Data Downloading
As aforementioned in 4.6.1 – TGO Data Downloading, post-processing will
not be necessary for observation data collected in RTK mode. The processing
procedures in Leica Ski Pro are quite similar as the procedures in Trimble Geomatic
Office.
First, the observed GPS data was downloaded from the PC memory Card to a
laptop. Then, Ski Pro was used to import the GPS data by created a new project and
assign the data into the project. After that, data will be displayed in graphic and
numerical format. Finally, the coordinate’s data was exported according to the format
required. The graphical of Leica Ski Pro data downloading is shown in Appendix I.
4.7 KFilter Program One of the main objective in this study is to develop a program that will be
used in high rise building movement monitoring for GPS using the algorithm as
aforementioned in Chapter 3, section 3.1.1.
The program was developed using Matlab version 6.1 with the aid of Kalman
Filter, thus named KFilter. KFilter is a window based program developed specially in
this study. Figure 4.17 shows the user interface of KFilter.
Figure 4.17: KFilter user interface
53
Movement Monitoring Result Analysis
Graph Visualization for Data Presentation Every
Minute with 5 Seconds Interval.
Developed Program named KFitler (Kalman
Filter) With Matlab.
Continuous Real Time Kinematics
(RTK) Data Observation
Figure 4.18: Flow Chart of Stage Analysis KFilter
Continuous RTK-GPS (Northing, Easting and Ellipsoid Height) observation
data was obtained with 1 second sampling rate is illustrated in Figure 4.18. The TGO
version 1.6 and Leica SKI-Pro software will create output data based on the
continuous RTK-GPS data with a format compatible to the format of input data for
developed program KFilter as shown in Figure 4.19).
54
Time Northing(N),m Easting(E),m Ellipsoid height(h), m
Figure 4.19: Format of Input Data for Developed Program KFilter
There is graph visualization for observed data presentation for every minute
with 5 seconds interval. If the waveforms in the graph are steady and consistent, then
it can be assumed that the point is stable. However, if the waveforms in the graphs
jump from its original consistent path and become consistent in the new path, then
there is some detectable deformation (see Figure 4.20).
Detectable Deformation
No Deformation
Figure 4.20: The Deformation Visualization Graph
55
Finally, the simulation tests had been carried out to ensure the reliability of the
developed KFilter program in movement monitoring works.
Input Data
Global Test (<3 Sigma)
Active So
Local Test
T
O
The pred
filt
equations o
Filter (C
Figure 4.21: Flow
The flow chart in Figure 4.21 sh
program KFilter. A file contained conti
input for program KFilter was created.
carried out to ensure that there is no any
test fails, the program will discard the ir
continue the processing with other dat
input data will be processed with the p
the Kalman Filter.
Yes
iction and
ering
f the Kalman
ross, 1983)
Deformation Dete
und System
/ Single Point
est
Chart of KFil
ows the step by
nuous RTK data
After that, the
gross error insi
relevant data (da
a. After the dat
rediction algorit
N
Yes
cted
No Deformation Detected
ter Program
step process of the developed
with suitable format as data
global test (3-sigma) will be
de the data processing. If the
ta contained gross error) and
a passed the global test, the
hm and filtering equations of
56
Then, single-point test is carried out at a significance level α (typically =
0.05). The Single Point Test known as final testing of deformation in the form of a
local test. The test is based on the null and alternative hypothesis:-
Ho : Edih = 0. No deformation for each point.
Ha : Edih ≠ 0 Existence of deformation for each point.
The single point test statistic is (Ince and Sahin, 2000):-
where,
Tih = test value
dih = the difference vector
σdi = the variance of the difference vector.
KFilter performs the single point test analysis for the northing, easting and
ellipsoid height in the processing. If Tih (t-calculate) ≤ F(1-a,t,fo) (t-table) – see Figure
4.22, it is considered that the point is stable. Otherwise, the rejection of the test
indicates that the point is deformed and the difference vector, dih is indeed a
significant deformation. Moreover, the sound system will be active of there are some
displacements about the point.
------Single Point Test ( 13:18:49 )------- Difference,cm t-calculate t-table Result 0.299 0.32 1.96 Stable 0.079 0.08 1.96 Stable -0.466 0.26 1.96 Stable ------Single Point Test ( 13:18:54 )------- Difference,cm t-calculate t-table Result 0.280 0.30 1.96 Stable 0.192 0.20 1.96 Stable -0.696 0.39 1.96 Stable
Figure 4.22: Example of Deformation Report
57
4.8 Simulation Test
Two different simulation tests had been carried out to ensure the reliability of
developed program. The simulation tests are ‘Movement’ Simulation Test and
‘Timing’ Simulation Test. The purpose of ‘Movement’ Simulation Test is to test the
limitation of displacement detection level for horizontal and vertical that can be
detected by the developed program. Meanwhile, the ‘Timing’ Simulation Test is to
test the developed program so that it can detect the displacements accurately and
successfully with the timing of ‘vibration’.
4.8.1 ‘Movement’ Simulation Test
The ‘Movement’ Simulation Test is divided into two types. The first test is
vertical simulation test and another test is horizontal simulation test. The tests had
been carried out in Universiti Teknologi Malaysia. A specially design paper which
has Northing and Easting axis with centimeter separations (intervals) was created.
The ‘origin’ of such paper is to be overlapped with the center point of GPS
monument. Then, the Northing direction is to aligned and pointed to North with the
aid of compass (see Figure 4.23). After that, the adjusted pole was put on the origin
of the paper to start the observation.
Figure 4.23: Preparation of ‘Movement’ Simulation Test
58
The pole was seated on the origin for first 5 minutes and then moved up the
vertical axis 1 centimeter by 1 centimeter (1 minute observation for every centimeter
of displacement) until the observation reached the height displacement of 8 cm. After
that, the pole was returned to the original height for another 5 minutes observation.
From here, the horizontal simulation test will be started later on. Then, the pole was
moved every 1 cm to Northing axis for every minute observation until 5 cm Northing
movements. The pole was returned to original position (origin on the paper) for 4
minutes, after that the same procedures were carried out for Easting axis. Finally, the
pole was put back to original position (origin of the paper) for 1 minute.
4.8.2 ‘Timing’ Simulation Test
The purpose of this study is to study the timing detection of the
displacements. Therefore, the adjusted pole was moved manually (see Figure 4.24)
using hands during the RTK-GPS observations to generate ‘deformable’ data inside
the observed data. Otherwise the pole was seated motionlessly on the GPS point to
create the no deformable continuous RTK-GPS observed dataset. In this study, the
height of pole was maintained during observations.
Static ‘Vibrated’
Figure 4.24: Static (Left of Figure) and ‘Vibrated’ (Right of Figure)
The schedules of the data observation are shown in Table 4.7. The further
information about the KFilter program deformation report – single point test for this
dataset test is shown in Appendix J.
59
Table 4.7: Schedule of ‘Timing’ Simulation Test Observation
Time Remarks
13:52:43 - 13:55:29 Static
13:55:34 - 13:56:44 ‘Vibrated’
13:56:49 - 13:58:40 Static
13:58:45 - 14:00:05 ‘Vibrated’
14:00:10 - 14:01:01 Static
4.9 Static GPS Deformation Analysis
GPS Deformation Analysis Program – Bayrak (2003) and GPSAD2000 –
Boon (2000) are the static GPS deformation analysis programs. Both programs are
specially developed for GPS baseline adjustment via LSE, deformation detection and
visualization analysis. A priori coordinates of the stations and baseline information
from observations data such as ∆X, ∆Y, ∆Z with its variance covariance are the input
data for the programs.
The processing methodology of the programs consisted of least squares
adjustment, global congruency test, localization of deformation, S-transformation,
and single point test (see Figure 4.25).
60
Figure 4.25: Process Methodology The least square adjustment/leas
been performed on the observed data. Th
with respect to elimination of errors an
precisions. Any deformation survey mu
survey so that gross or systematic err
movements and produce false results. Sur
at certain time intervals (measurement e
expected movement / settlement of the
stations and targets are put in place and
established procedures are repeated at ea
errors. Each of these repeated netwo
measurement, so the comparison and anal
commonly known as epoch testing.
No
of Static GPS Deformation Analysis
t square estimation (LSE) procedures had
e adjustment requires all the considerations
d the correct estimation of measurement
st pay particular attention to errors in the
ors do not contaminate the detection of
veys for deformation are generally repeated
pochs). The time interval depends on the
structure and the risk to life. Generally,
suitable field procedures established. The
ch epoch to minimize systematic and gross
rk surveys is known as an epoch of
ysis of the results of the repeated surveys is
61
The essence of epoch testing is to determine whether the differences between
the coordinates from two different epochs are statistically significant. Epoch testing
must take into account the precisions of the coordinates, as well as the correlations
between both the coordinates of individual stations and coordinates of different
stations hence the full weight coefficient matrices from each epoch contribute as
follows:
Epoch 1:
x1 station coordinates vector
Q1 weight coefficient matrix
Epoch 2:
x2 station coordinates vector
Q2 weight coefficient matrix
and the differences in the coordinates and the associated weight coefficients
are:
d = x2 - x1
Qd = Q1 + Q2
The first test conducted should always be the global congruency test:
analogous to the global test for a single network. The quantity Ω = dt Qd-1 d is tested
against a Fisher statistic at an appropriate confidence level, if Ω passed then there has
been no (statistically significant) movement and the networks are congruent. If
Ω fails, the global congruency test then each point must be assessed by a local test
which compares the contribution to Ω of the point against a Fisher critical value:
analogous to the local testing of residuals for a single network (localization of
deformation). This test is done by recalculating Ω without each point in turn. The
worst point is rejected (S-transformation) and the entire testing process repeated,
including the transformation for free networks (without the rejected points, which are
now considered to be moved) and the global congruency test. Once the global
congruency test passes, all those points which have been rejected are considered to
have moved whilst those that are still contributing to Ω are considered to be stable
(single point test).
62
Graphical representations of deformation analyses are often shown as error
ellipses (normally 95% confidence interval) with vectors of movement - differences
between measurement epochs. Example, the visual representation is useful for
empirical checking and for the identification of the characteristics of any movement.
4.10 Movement Monitoring Analysis
One of the important applications of statistical methods in surveying
engineering is the deformation analysis / structural monitoring (see Figure 4.35). It
can detect and analyze movements of individual station or group of stations. There
are two major deformation analysis methods which can be used for monitoring
strategies. One method uses observation differences while the other method uses
coordinates differences.
This study selected the coordinate differences method because it allows more
flexibility with regards to any types of horizontal survey control measurements. The
only requirement for the coordinate differences method is that the station positions
should be observed in both epochs regardless of which measurements were made to
determine that position. Another advantage of the coordinates difference method is
that each survey network is independently adjusted, and this will ensure data quality.
4.11 Study of Wind Effect (Vibration) Using RTK-GPS Data The RTK-GPS (Real Time Kinematics – Global Positioning System) has a
nominal accuracy of 1cm + 1ppm for horizontal displacements with sampling rate of
10Hz. Celebi (1998) proposed the use of RTK-GPS for measurement of building
responses, the responses with amplitudes (vibrations) larger than 2cm can be
detected by RTK-GPS (Tamura, 2004).
63
The RTK-GPS is used to study wind effects on the building especially its
vibration direction. In order to verify the outcome, an instrument named anemometer
had been used (see Figure4.26). The anemometer used is manufactured by DAVIS. It
can display wind direction and wind speed. The specification of the instruments is
shown in Appendix K.
Figure 4.26: Anemometer (Meteorologica Ltd, 2004)
CHAPTER 5
ANALYSES AND RESULTS
5.1 Introduction
Accuracy analysis (Root Means Square) is very important in the study. This
is because the RMS value represents the accuracy of the positioning technique. The
RMS can be calculated using the following algorithm:
RMS =√ [∑(Xi – X)2/N] (5.1)
Where, Xi = Continuous RTK-GPS Observation Data
X = Coordinates from Static-GPS
N = Sum of the observation data.
5.2 Results Analysis for Study on RTK-GPS Baseline The analysis focused on UTMR station, which has both Static-GPS
coordinates and continuous RTK-GPS observation data, using the algorithm (5.1).
Table 5.1: Analysis on One and half an hour Continuous RTK-GPS Data
For Station UTMR
Northing, mm Easting, mm WGS84 Ellipsoid Height, mm
Periods 1 hour 1/2 hour 1 hour 1/2 hour 1 hour 1/2 hour Max 16 13 16 15 21 12 Min 4 1 -2 2 -26 -33 RMS 10 7 7 9 9 12
65
Table 5.1 shows the analysis of RMS RTK-GPS for the two different sets of
observation data, which are the one hour data and the half-hour data. The table shows
RMS Northing continuous RTK-GPS is 10mm, RMS Easting is 7mm and RMS
WGS84 Ellipsoid Height is 9mm for one hour data. Meanwhile RMS Northing
continuous RTK-GPS is 7mm, RMS Easting is 9mm and RMS WGS84 Ellipsoid
Height is 12mm for half an hour data. In conclusion, both data gave same outcomes
whereby the horizontal accuracy <1cm and vertical accuracy < 2cm. This result
fulfilled the accuracy RTK-GPS of Specification Trimble 4800 Series.
5.3 Results Analysis for Test on Accuracy of RTK-GPS Baseline The analysis will focus on TR2300 station, which the processing utilize both
Fast Static-GPS coordinates and continuous RTK-GPS observation data altogether
using algorithm as in equation (5.1).
Table 5.2: RMS Analysis on Continuous RTK-GPS Data for T200, T300 and
TR2300
Northing, mm Easting, mm WGS84 Ellipsoid Height, mm
Station T200 T300 T200 T300 T200 T300 Max 20 27 11 -12 15 81 Min -6 19 -20 -22 -45 19 RMS 7 23 10 17 13 49
Table 5.3: Details Explanation Analysis
RMS (mm)
Northing Easting WGS84 Ellipsoid Height
Station Remarks
7 10 13 T200 200m away from TR2300 station.
23 17 49 T300 300m away from TR2300 station and near some thin
canopy.
- - - >300m >300m away from
TR2300 station and canopy acted as
66
obstructions. Thus, radio link communication down
and procedure no corrections signal.
According to Table 5.2 and 5.3, T200 is a base station situated 200m away
from rover (TR2300). The RMS of Northing is 7mm, RMS of Easting is 10mm and
RMS of WGS84 Ellipsoid Height is 13mm. This outcome fulfilled the specification
of RTK-GPS Trimble 4800 Series whereby horizontal accuracy is less than 10mm
and vertical is less than 20mm.
T300 is a station which is 300m away from rover (TR2300) and located near
some thin canopy. 10 set of data were obtained in 2 minutes/120 seconds (see
Appendix F). That means sometime the correction signal transmitted from base
station to rover station was lost during the observation. The RMS for horizontal and
vertical are poor. RMS of Northing, Easting and WGS84 Ellipsoid Height do not
fulfill the specifications of Trimble 4800 Series.
Radio link communication between base and rover RTK-GPS was down for
areas that are more that 300 meter where the canopy acted as obstructions. It means
that radio signal strength can not transmit and pass through the obstructions, and the
factor of distance did not fail the radio link communication. Bad radio Link
communication became the main problem for RTK-GPS observation because there
will be no transmission signal/corrections from base station to rover station.
In conclusion, maintaining the line of sight is very important for radio link
communication between base and rover. This had been proven by the observation in
which the radio link communication was down in canopy area (obstructions),
between base and rover during RTK-GPS observations. Some recommendations for
the usage of continuous RTK-GPS are pointed out as below: -
i. Line of sight is important between RTK-GPS base and rover.
ii. Distance between base and rover <200m with clear line of sight will produce
horizontal accuracy 1cm and vertical accuracy 2cm.
67
iii. According to Lee Kong Fah (2003), the distance between base and rover
stations within 1 – 2 km (>200m with clear line of sight) will produce
horizontal and vertical accuracy 2cm respectively.
5.4 Results Analysis on ‘Movement’ Simulation Test
Table 5.4: Simulation Test for Vertical Axis
Time Value Displacement,
cm (min – max)
Descriptions Remarks
12:51:30 - 12:56:28 0.006 – 0.792 Default Stable 12:56:58 - 12:57:55 0.899 – 1.732 Move 1cm Stable 12:58:43 - 12:59:39 1.909 – 2.672 Move 2cm Stable 13:01:05 - 13:02:00 2.211 – 3.079 Move 3cm Stable 13:03:13 - 13:04:03 3.062 – 4.019 Move 4cm Stable /
Moved 13:04:55 - 13:05:52 4.657 – 5.704 Move 5cm Moved 13:06:24 - 13:07:16 5.757 – 6.342 Move 6cm Moved 13:07:53 - 13:08:55 6.023 – 6.927 Move 7cm Moved 13:09:47 - 13:10:43 6.661 – 7.792 Move 8cm Moved 13:11:46 - 13:13:36 0.023 – 0.508 Default Stable
Table 5.4 shows the results of height/vertical simulation test. The selected
starting 2 minutes of the observed RTK-GPS data is defined as coordinate’s
reference. The displacement value can be obtained by comparing the RTK-GPS data
with the coordinate reference. In the table, the developed program will detect the
vibrations or displacements if the object moved larger than 4 ± 1cm – see column 5
Table 5.4. 4cm is defined as limitation of vertical deformation detection of the
developed KFilter program, means if the point moved >4cm in vertical axis and the
program can detect it and defined as deformable point. Meanwhile ±1cm value
represents the RTK-GPS error during observations – see column 1 and 10 Table 5.4.
68
Table 5.5: Simulation Test for Horizontal (Northing & Easting) Time Value
Displacement, cm (min – max)
Descriptions Remarks
13:16:07 - 13:21:02
0.159 – 0.572(N) 0.006 – 0.965 (E)
Default Stable
13:25:18 - 13:26:22
0.593 - 0.800 (N) 0.495 - 0.833(E)
Moved 1cm N Stable
13:26:51 - 13:27:46
1.683 – 1.965 (N) 0.984 – 1.397 (E)
Moved 2cm N Stable / Moved
13:28:25 - 13:29:26
2.022 – 2.680 (N) 1.115 – 1.830 (E)
Moved 3cm N Moved
13:30:06 - 13:31:11
3.563 – 3.789 (N) 1.848 – 2.111 (E)
Moved 4cm N Moved
13:31:58 - 13:32:56
3.789 – 4.183 (N) 1.642 – 2.036 (E)
Moved 5cm N Moved
13:33:47 - 13:37:37
0.011 – 0.591 (N) 0.006 – 0.495 (E)
Default Stable
13:38:18 - 13:39:20
0.629 – 1.080 (N) 0.927 – 1.059 (E)
Moved 1cm E Stable
13:40:01 - 13:41:02
0.873 – 1.211 (N) 1.341– 1.961 (E)
Moved 2cm E Stable / Moved
13:41:34 - 13:42:39
1.343 -1.775 (N) 2.187 – 2.675 (E)
Moved 3cm E Moved
13:46:39 - 13:48:01
1.343 – 1.888 (N) 3.258 – 3.747 (E)
Moved 4cm E Moved
13:49:01 - 13:50:07
1.944 – 2.245 (N) 3.897 – 4.160 (E)
Moved 5cm E Moved
13:50:48 - 13:51:52
0.008 – 0.572 (N) 0.044 – 0.570 (E)
Default Stable
Table 5.5 illustrates the results of horizontal simulation test. The RTK-GPS
data was compared the coordinates reference (means value for selected starting 2
minutes of observed RTK-GPS data) to obtain displacement values. In the table, the
developed program will detect the vibrations or displacements of horizontal
(Northing and Easting) if the object moved larger than 2 ± 1cm either in Northing
(see column 3 Table 5.5) or Easting (see column 9 Table 5.5) dimension
respectively. ±1cm value is the RTK-GPS error during observations – see column 1,
7 and 13 Table 5.5. This study fulfilled the statements claimed by Tamura (2004)
that amplitude (vibration) larger than 2cm can be detected by RTK-GPS.
69
5.5 Results Analysis on ‘Timing’ Simulation Test
The graph presentation shows a ‘linear’ line (see Figure 5.1) during
observations because the object was stationary in the original position from 13:52:43
until 13:55:29 – see Appendix H. This means there is no deformation detected in the
RTK-GPS data during the observations. Meanwhile, if the object was moved
purposely from 13:55:34 until 13:56:44 (see Appendix H) to generate deformation
data in the RTK-GPS observations data and the KFilter program can detect it on the
spot (see Figure 5.2). Thus, the developed KFilter program shows it has good graph
presentation to represent the deformation detection vs. the observation times.
Figure 5.1: No Deformation Detected
70
Figure 5.2: Deformation Detected
5.6 Case Study: Menara Sarawak Enterprise
In order to verify (benchmark) the results of structural analysis using
developed KFilter program, the Static GPS was processed using GPS Deformation
Analysis Program (Bayrak, 2003) and GPSAD2000 (Boon, 2000).
From Table 5.6, all the programs show that there isn’t any deformation
displacement detected at the Menara Sarawak Enterprise building. GPS Deformation
Analysis Program (Bayrak, 2003) did not give further information in the deformation
report (see Appendix L) if the points are stable by mentioning ‘There is NOT
deformations/Moving Points in the Networks’.
GPSAD2000 (Boon, 2000) also mentioned same statements in this case study
by showing that R1 and R2 on the rooftop of Menara Sarawak Enterprise Building
are displaced around 1cm (Cartesians Coordinate System) but still under stable
71
conditions. Similarly, GPSAD2000 program shows that the B1 and B2 base stations
considered stable although both stations having 0.39 cm displacement.
The stability of base stations for RTK-GPS (B1 and B2) should been
identified before carrying out further analysis for R1 and R2 stations. Therefore,
stability of B1 and B2 stations had been proven using GPS Deformation Analysis
Program and GPSAD2000. It is because RTK-GPS method is relative positioning
technique. Any unstable base stations will cause the calculated displacements of the
rover points to be wrongly interpreted. Continuous RTK-GPS epoch 1 and 2 analyses
show that R1 and R2 stations are in stable condition. . The cont. RTK-GPS epoch 1
analysis declared that the R1 and R2 points displaced around 0.46cm 0.39cm
respectively. However, such changes did not lead to any deformation because the
displacement values were within 0.5cm. Moreover, the epoch 2 analysis declared that
the R1 and R2 points displaced around 1.36cm and 1.14cm respectively (bigger
much compared with displacement distance of epoch 1) but the points are still under
stable and safety conditions.
72
Table 5.6: Results of Processing From GPS Deformation Analysis Program, GPSAD2000 and KFilter
Points ∆X ∆Y ∆Z ∆N ∆E ∆h Displacement Distance
Results Program Observations
B1 B2 R1 R2
- - - - - - - Stable
GPS Deformation
Analysis Program
Static
B1 0.21 0.17 0.27 - - - 0.39 Stable GPSAD2000 StaticB2 0.21 0.17 0.27 - - - 0.39 Stable GPSAD2000 StaticR1 -0.36 0.66 -0.27 - - - 0.80 Stable GPSAD2000 StaticR2 -0.07 -1.02 -0.28 - - - 1.05 Stable GPSAD2000 Static
R1 - - - -0.3420 0.2999 (mean) (mean)
0.0713 (mean)
0.46 Stable KFilter Cont. RTK(Epoch 1)
R2 - - - 0.3040 0.2240 (mean) (mean)
0.1114 (mean)
0.39 Stable KFilter Cont. RTK(Epoch 1)
R1 - - - 0.0118 -0.4586 (mean) (mean)
-1.2793 (mean)
1.36 Stable KFilter Cont. RTK(Epoch 2)
R2 - - - 0.7824 0.7896 (mean) (mean)
0.2638 (mean)
1.14 Stable KFilter Cont. RTK(Epoch 2)
** Mean value for ∆N, ∆E and ∆h can obtain from KFilter Program by getting the mean for difference values from deformation report. ** Units in centimeters, cm.
73
5.7 Results Analysis For Study of Wind Effect (Vibration) Using RTK-GPS Data
Figure 5.3: Northing and Easting Displacements Graph
From the Figure 5.3, there are many points located at North-West Side of the
starting point. The points are RTK-GPS observation data which were affected by
wind effects on the building during field measurements. Meanwhile the starting point
is the mean coordinates of selected starting 2 minutes of observation without any
wind effects during the measurements. In this case, the direction of movements for
the coordinates of RTK-GPS observation data had been proven by outcome from
anemometer. Both of them show the same movement direction from time 14:30:00
until 14:37:30 for wind effect on the building.
The Figures 5.4, 5.5 and 5.6 shows the RTK-GPS observed data without wind
effects – precision of RTK-GPS is shown in the ‘box’ in three figures respectively.
Meanwhile the coordinate’s differences in Northing, Easting and Ellipsoid height of
74
RTK-GPS observed data compared to the mean coordinates respectively whereby
there is critical time of wind effects on the building, the time is 14:30:00 until
14:37:30. From the three figures, the wind effects gave critical impacts on horizontal
axes. On the other hand, wind effect has no significant effect on the height
component.
75
Figure5.4: Northing Movements Value Resulted From Winds Effects
76
Figure 5.5: Easting Movements Value Resulted From Winds Effects
77
Figure 5.6: WGS84 Ellipsoid Height Movements Value Resulted From Winds Effects
78
------Single Point Test ( 14:36:05 )------- Difference,cm t-calculate t-table Result -0.827 0.88 1.96 Stable 0.335 0.36 1.96 Stable -0.095 0.05 1.96 Stable ------Single Point Test ( 14:36:10 )------- Difference,cm t-calculate t-table Result -0.921 0.98 1.96 Stable 0.260 0.28 1.96 Stable 0.366 0.21 1.96 Stable ------Single Point Test ( 14:36:15 )------- Difference,cm t-calculate t-table Result -1.015 1.08 1.96 Stable 0.110 0.12 1.96 Stable -0.077 0.04 1.96 Stable ------Single Point Test ( 14:36:20 )------- Difference,cm t-calculate t-table Result -1.015 1.08 1.96 Stable 0.091 0.10 1.96 Stable 0.100 0.06 1.96 Stable ------Single Point Test ( 14:36:25 )------- Difference,cm t-calculate t-table Result -0.827 0.88 1.96 Stable 0.335 0.36 1.96 Stable 0.313 0.18 1.96 Stable ------Single Point Test ( 14:36:30 )------- Difference,cm t-calculate t-table Result -0.958 1.02 1.96 Stable 0.448 0.48 1.96 Stable 0.118 0.07 1.96 Stable ------Single Point Test ( 14:36:35 )------- Difference,cm t-calculate t-table Result -0.940 1.00 1.96 Stable 0.298 0.32 1.96 Stable 0.189 0.11 1.96 Stable ------Single Point Test ( 14:36:40 )------- Difference,cm t-calculate t-table Result -0.733 0.78 1.96 Stable 0.486 0.52 1.96 Stable -0.006 0.00 1.96 Stable ------Single Point Test ( 14:36:45 )------- Difference,cm t-calculate t-table Result -0.846 0.90 1.96 Stable 0.448 0.48 1.96 Stable 0.047 0.03 1.96 Stable
------Single Point Test ( 14:17:45 )------- Difference,cm t-calculate t-table Result -0.094 0.10 1.96 Stable 0.016 0.02 1.96 Stable 0.260 0.15 1.96 Stable ------Single Point Test ( 14:17:50 )------- Difference,cm t-calculate t-table Result 0.094 0.10 1.96 Stable -0.229 0.24 1.96 Stable 0.224 0.13 1.96 Stable ------Single Point Test ( 14:17:55 )------- Difference,cm t-calculate t-table Result 0.038 0.04 1.96 Stable -0.003 0.00 1.96 Stable 0.331 0.19 1.96 Stable ------Single Point Test ( 14:18:00 )------- Difference,cm t-calculate t-table Result -0.150 0.16 1.96 Stable -0.059 0.06 1.96 Stable 0.224 0.13 1.96 Stable ------Single Point Test ( 14:18:05 )------- Difference,cm t-calculate t-table Result 0.132 0.14 1.96 Stable -0.097 0.10 1.96 Stable 0.082 0.05 1.96 Stable ------Single Point Test ( 14:18:10 )------- Difference,cm t-calculate t-table Result 0.245 0.26 1.96 Stable 0.128 0.14 1.96 Stable 0.029 0.02 1.96 Stable ------Single Point Test ( 14:18:15 )------- Difference,cm t-calculate t-table Result 0.113 0.12 1.96 Stable -0.078 0.08 1.96 Stable -0.219 0.12 1.96 Stable ------Single Point Test ( 14:18:20 )------- Difference,cm t-calculate t-table Result 0.000 0.00 1.96 Stable 0.166 0.18 1.96 Stable -0.130 0.07 1.96 Stable ------Single Point Test ( 14:18:25 )------- Difference,cm t-calculate t-table Result 0.075 0.08 1.96 Stable -0.153 0.16 1.96 Stable -0.042 0.02 1.96 Stable
Figure 5.7: The Deformation Report (KFilter) for Without Wind Effect Situation and With Wind Effect Situation
79
The Figure 5.7 is the deformation report generated by developed program
named KFilter. The output on the left hand side of the figure is the observation data
without wind effect. There are some small coordinate’s differences of Northing,
Easting and Ellipsoid height. On the other hand, the output on the right hand figure is
the observation data with wind effect. The coordinate’s differences are around 1cm
in Northing, and 0.5cm in Easting. But the ellipsoid height did not change too much
(small difference) either in wind effect situation or not during the observations.
Therefore, data analysis for the observation done from time 14:30:00 until 14:37:30
with wind effect on the building and the wind speed was 1.9m/s shown there was no
any deformation detected during that period. Thus, the Menara Sarawak Enterprise is
stable with 1.9m/s wind speed affects.
In conclusion, The RTK-GPS method can be used in movement monitoring.
The RTK-GPS has a nominal accuracy of ±1cm +1ppm for horizontal displacements
with a sampling rate of 10Hz. It is suitable to measures building responses when the
vibration larger than 2cm.
5.8 Summary
Before implementing the RTK-GPS as a survey method, the RTK-GPS
method must be studied first, especially its hardware specifications and operating
procedures. Therefore, the RTK-GPS baselines had been studied in this study. The
study verified the accuracy of RTK-GPS which defined by manufacturers and the
reliability of Pacific Crest as radio link for RTK-GPS base and rover
communications.
A program for movement monitoring had been developed using Matlab
version 6.1 with helps of Kalman Filtering and RTK-GPS. The developed program
had been verified by the simulation tests to ensure its reliability and accuracy in
deformation detections.
80
The high rise building case study in this study was carried out in Menara
Sarawak Enterprise which has 33 floors in total. The field measurements had been
carried out in two epochs, on Dec 2004 and Apr 2005. Then, the developed program
performed the movement monitoring analysis on the RTK-GPS observation data.
Besides that, other programs had been used to perform the movement monitoring
analysis on Menara Sarawak Enterprise building to increase the reliability of
analysis. The programs are GPS Deformation Analysis Program and GPSAD2000.
Both of them used static observation data to do further analysis.
Celebi (1998) proposed the use of RTK-GPS to measure building responses
caused by wind effects using RTK-GPS observation. Therefore, the RTK-GPS
observation data had been used to study the wind effects on the Menara Sarawak
Enterprise building especially wind vibration directions. The instruments named
anemometer had been used as the verification of such results.
CHAPTER 6
CONCLUSIONS AND RECOMMENDATIONS 6.1 Conclusions
This study focus on the way to apply continuous RTK-GPS observation for
the Menara Sarawak Enterprise building movement monitoring and observations
analysis methods and lastly to determine the high rise building’s deformation status.
The reconnaissance is very important in selecting the locations of reference
and rover stations to guarantee the effectiveness of radio link communication that
yields good RTK-GPS observations. The calibration tests proved that the line of sight
should be taken care of in order to avoid blockage (obstructions) of communication
between reference station and rover.
Analysis for RTK-GPS observation data in this study is done by the
developed KFilter program which is developed using Matlab v6.1 integrated with
Kalman Filter method. The program can detect the displacements if the object moved
>4cm in vertical axis (Ellipsoid Height) or >2cm in horizontal axis (Northing and
Easting) respectively.
From the KFilter program analysis, the results shows that the Menara
Sarawak Enterprise building is stable. The continuous RTK-GPS epoch 1 analysis
declared that the R1 and R2 points – which are placed on the rooftop using special
gadget displaced around 0.46cm and 0.39cm respectively but the points are still
stable. Moreover, the continuous RTK-GPS epoch 2 analyses had shown the points
82
(R1 and R2) are stable with displacement of 1.36cm and 1.14cm respectively. Beside
the RTK-GPS observation there is other GPS method that had been carried out, the
Static GPS. The purpose of this method is to identify the stability reference station of
RTK-GPS to avoid the calculated displacement been wrongly interpreted. Therefore,
GPS Deformation Analysis Program had been proven the B1 and B2 reference
stations are stable and GPSAD2000 program shows that although the reference
stations have 0.39cm displacement but still under stable condition.
In this study, RTK-GPS observation data had been used to study the wind
effects on the Menara Sarawak Enterprise building especially wind vibration
directions. The anemometer instrument had been used and gave same movement
direction with the movement for the RTK-GPS observed data which are affected by
wind effects. Moreover, data analysis shows that the Menara Sarawak Enterprise is
stable for observation period from time 14:30:00 until 14:37:30 with 1.9m/s wind
speed affect on the building and there was no any deformation detected.
As a conclusion, continuous Real-Time Kinematics GPS observation
technique is applicable for movement monitoring of Menara Sarawak Enterprise
building. Besides the high rise building, the method should also apply in landslide etc
for monitoring purpose.
6.2 Recommendations
The satellite-based method (continuous RTK-GPS) has the potential to be
employed for high rise buildings monitoring. However, the following
recommendations can be considered to improve this study:
i. Nowadays, recent advances in GPS technology have made it a cost effective
tool for monitoring safety and performance of high rise buildings. Here, the
system so called ‘on-line GPS integrity monitoring’ which can provide
continuous real time measurements or RTK-GPS could be experimented in
83
near future, which can, in turn, be used to indicate ‘instantaneous’
displacements and vibrations caused by wind loading, distant earthquakes,
landslides etc (Wan, 2003).
ii. The developed program could be combined with other spectral analysis
methods, likes Fast Fourier Transform and Wavelet etc to perform structural
monitoring analysis for GPS observations data. If the structure moves or
vibrates, then all the methods could have been able to detect it respectively.
Moreover, the geotechnical instruments such as accelerometer can link and
combine with RTK-GPS technique during observations will be increase the
reliability of the system
iii. The GPS instruments especially dual-frequency receivers are expensive
nowadays. Therefore, the low cost GPS receivers can be developed to replace
these instruments. The developed program with low cost receivers will
became a fully on-line GPS integrity monitoring system which can be
installed at every high rise buildings. Thus, the safety and stability of the
buildings can be monitored continuously, 24 hours a day.
84
REFERENCES 21CEP (2000), ‘Measuring the height of Mt. Everest’,
http://www.100gogo.com/ever2.htm
Adrian B and Moshe B. (2002), ‘Matlab 6 For Engineers’, Third Edition, Harlow:
Prentice Hall
Angus J. M. (1975), ‘Wind Loading on Buildings’, First Edition, London: Applied
Science Publishers Ltd.
Bayrak T. and Mualla Y. (2003), ‘A Kinematic Analysis Program for Deformation
Monitoring’, 11th International Symposium on Deformation Measurements,
Proceedings, 439-444, May 25-28, 2003, Santorini (Thera), Greece.
Bertrand M. (1989), ‘The Use of Kalman Filters in GPS Navigation’, Reports From
School of Surveying, The University of New South Wales.
Bock, Y. and M. Bevis (1999), ‘Regional Permanent GPS Arrays, Trends and
Prospects’, International Symposium on GPS, Tsukuba, Japan, October 18–22,
1999.
Bong Chin Nyet (2000), ‘GPSAD2000 - Sistem Perisian Untuk Pelarasan Vektor
GPS, Pengesanan Deformasi dan Analisis Visualisasi’, Bachelor of Degree
Report, Faculty of Geoinformation Science and Engineering, Universiti
Teknologi Malaysia.
Brown, R.G. and Y.C. Hwang (1992), ‘Introduction to Random Signals and
Applied Kalman Filtering’, Second edition, New York: John Wiley & Sons
85
Brownjohn, J. M. W, T.C. Pan and Deng X. Y. (2000), ‘Correlating Dynamic
Characteristics from Field Measurements and Numerical Analysis of a high rise
Building’, Earthquake Engineering & Structural Dynamics Journal, 29(4), 523-
543,
Brownjohn, J. M. W, and T.C. Pan (2001), ‘Response of a tall building to long
distance earthquakes’, Earthquake Engineering & Structural Dynamics Journal,
30, 709-729.
Caspary, W.F. (1987), ‘Concepts of Network and Deformation Analysis. (Ed. by
Rueger, J.M,)’, Monograf 11. School of Surveying The University of New
South Wales, Kensington, New South Wales, Australia.
Celebi M. (1998), ‘GPS and/or strong and weak motion structural response
measurements – Case Studies’, Structural Engineering World Congress ’98,
San Francisco, Conference Proceedings on CD-ROM, T193-1 pp8.
Chalermchon S., Ogaja C., J. Wang and C. Rizos(2002), ‘GPS Analysis With the Aid
of Wavelets’, School of Geomatic Engineering, The University of New South
Wales, Sydney, Australia,
http://www.gmat.unsw.edu.au/snap/publications/satirapod_etal2001c.pdf
Charles J. A. (2005), ‘Earthquake Effects’, http://eqseis.geosc.psu.edu
Chris Rizos (1999), ‘Quality Issues in Real-Time GPS Positioning’, International
Association of Geodesy SSG 1.154, IUGG Congress, Birmingham U.K. 18-29
July 1999.
Chris Rizos (1999), ‘The GPS System’,
http://www.gmat.unsw.edu.au/snap/gps/gps_survey/chap2/223.htm
Chris Rizos (1999), ‘Chapter 1: The GPS System’, University of New
South Wales, http://www.gmat.unsw.edu.au/snap/gps/gps_notes1.pdf
86
Chrzanowski, A., Chen, Y.Q. and Secord, J.M (1986), ‘Geometrical analysis
of deformation surveys’, Deformation Measurement Workshop
(Massachusetts Institute of Technology). (Ed.) Bock, Yehuda. Modern
methodology in precise engineering and deformation surveys-II.
Cambridge : Massachusetts Institute of Technology. 170–206.
Chui C. K. and Chen G. (1991), ‘Kalman Filtering With Real Time Applications’,
Second Edition, Berlin: Springer-Verlag.
Cross, P. A. (1983), ‘Advanced Least Squares Applied to Position Fixing’, Working
Paper No. 6, North East London Polytechnic.
Ctbuh (2004), ‘Criteria’, http://www.ctbuh.org
Cook N. J. (1990), ‘The Designer’s Guide to Wind Loading of Building Structures,
Part 2: Static Structures’, Building Research Establishment report, London:
Butterworths.
David Rutledge, Jack Gnipp and John Kramer (2001), ‘ Advances in Real-Time GPS
Deformation Monitoring For Landslides, Volcanoes, and Structures’, The 10th
FIG International Symposium on Deformation Measurements, 19-22 march
2001, Orange, California, USA, pp 110-121.
Davis Langdon & Everest (2004), ‘Comparing Costs’,
http://www.cda.org.uk/megab2/build/Pub%20155_costs.pdf.
Emporis (2004), General Definition of High-rise Buildings. http://www.emporis.com
Garmin Corporation (2004), ‘What is GPS?’, http://www.garmin.com
Gelb A., Kasper Jr. J. F., Nash Jr. R. A. Price, C. F. Surtherland Jr. A. A. (1974),
‘Applied Optimal Estimation’, Cambridge, MA: The MIT Press.
87
GPS WORLD (2001), ‘The Concrete and the Clay: Monitoring Large-Structure
Deformation’, http://www.gpsworld.com
Grewal, M. S. and A. P. Andrews (1993), ‘Kalman Filtering Theory and Practice’,
Englewood Cliffs, New Jersey: Prentice Hall.
GuruNet (2005), ‘skyscraper: Definition and Much More from Answers.com’,
http://www.gpsworld.com
Halim, S. and Ranjit, S. (1999), ‘Deformation Analysis of a Geodetic Monitoring
Network’, Geomatica, Vol. 55, No. 3, pp. 333-346.
Hofmann-Wellenhof, B, Lichtenegger and Collins, J. (1994), ‘GPS – Theory and
Practice’, Third Edition, Berlin: Springer-Verlag.
Hore A.V., J. G. Kehoe, R. McMullan, M. R. Penton (1997), ‘Constructions 2:
Environment, Science, Materials and Technology’, London: Macmillan Press
Ltd.
HowStuffWorks Inc (2005), ‘How Skyscrapers Work’,
http://science.howstuffworks.com
Ince C. D. and Sahin M. (2000), ‘Real Time Deformation Monitoring With GPS
and Kalman Filter’, Istanbul Technical University, Faculty of Civil
Engineering, Department of Geodesy and Photogrammetry, Turkey.
http://www.terrapub.co.jp/journals/EPS/pdf/5210/52100837.pdf.
Jansson, P. (1998), ‘Precise Kinematic GPS Positioning with Kalman Filter and
Smoothing’, Doctoral Dissertation, Royal Institute of Technology, Department
of Geodesy and Photogrammetry Stockholm, Sweden
Jay S. and Ralph R. (1998), ‘Field Test Of Trimble 4000 Real Time Kinematic GPS
Survey System’, Journal of Surveying Engineering, Vol 124 (1), pp40-48.
88
Kalman, R. E. and R. S. Bucy. (1960), ‘New Results in Linear Filtering and
Prediction Theory’, Journal of Basic Engineering, 82D, 35-45.
Kuhlmann H. (2003), ‘Kalman Filtering With Coloured Measurement Noise For
Deformation Analysis’, Institute for Applications of Geodesy to Engineering,
University of Stuttgart, Germany.
http://gib.geod.uni-bonn.de/deutsch/mitarbeiter/kuhlmann/h3.pdf.
Lee Keng Fah (2003), ‘Kajian Kesan Garis Dasar Di Dalam Penentududukan
Dengan Kaedah RTK-GPS’, Bachelor of Degree Report, Faculty of
Geoinformation Science and Engineering, Universiti Teknologi Malaysia.
Lee S. L., Yong K.Y., Swaddiwudhipong S., Sitichaikasem S., Ratanaprichavej R.
(1992), ‘Structural Design and Construction of 70-storey Concrete Building’,
Special Lecture, Proc. JCI Annual Convention Fukuoka, Japan Concrete
Institute.
Lee S. L., Swaddiwudhipong S. (1996), ‘Up-Down Construction of Tall Buildings
in City Centre’, Proc. Int. Conf. on Urban Engineering in Asian Cities in the
21st Century, Bangkok, Thailand.
Leick, A. (2004), ‘GPS Satellite Surveying. John Wiley & Sons. Third Edition.
USA ISBN 0-471-05930-7
Linlin Ge. Shaowei Han and Chris Rizos (1999), ‘ GPS-RTK Applications For Assisting The
Engineering Design of Large Structures’, The Metrology Society of Australia 1999
Conference, 22-24 September, Sydney, Australia, 287-291.
Leica Geosystems (2000) , ‘ User Manual / Getting Started with Ski-Pro Version
2.0’, Leica Geosystems AG, Heerbrugg, Switzerland.
Lovse J.W., Tesky W.F., Lachapelle G., etc (1995), ‘Dynamic deformation
Monitoring of Tall Structure using GPS Technology’, Journal of Surveying
Engineering, Vol. 121 (1), pg 16-22.
89
Lynn S. B. (1983), ‘Developments in Tall Buildings, 1983’, New York: Van
Nostrand Reinhold Company.
Nicholas J. C. (1999), ‘Wind Loading: A Practical Guide to Wind Loads on
Buildings’, Thomas Telford Publishing, Thomas Telford Limited, 1 Heron
Quay.
NISEE (1997), ‘Earthquake Damage Distribution’,
http://nisee.berkeley.edu/kobe/damage.html
Marc E. H. (2001), ‘Programming in Matlab’, Pacific Grove, California:
Brooks/COLE.
Maybeck, P. S. (1979), ‘Stochastic Models, Estimation and Control Volume I’,
423 pp., New York: Academic Press, Inc.
Mertikas, S. and C. Rizos. (1997), ‘On-line Detection of Abrupt Changes In The
Carrier Phase Measurements of GPS’, Journal of Geodesy, 71, 469-482.
Meteorologica Ltd (2004), ‘Meteorogica Weather Superstore’,
http://www.meteorologica.co.uk/
Michael A. Duffy, Chris Hill, Cecilia Whitaker, Adam Chrzanowski, James Lutes,
and Geoffrey Bastin (2001), ‘An Automated and Integrated Monitoring
Program For Diamond Valley Lake In California’, FIG Int. Symp. On
Deformation Measurements, Orange, CA.
Mualla Y. and T. Bayrak (2003), ‘GPS In Landslides Monitoring: A Case Study
From North Eastern Turkey’, GIS 2002 International Symposium,
Istanbul/Turkey, September 23-26 2002.
90
Mualla Y. and T. Bayrak (2003), ‘A Dynamic Analysis Method Regarding
Groundwater Level Changes As Causative Force For Landslides’, 11th
International Symposium on Deformation Measurements, Proceedings, 303-
311, May 25-28,2003, Santorini (Thera), Greece.
Mualla Y. and T. Bayrak (2003), ‘Comparison of Static, Kinematic and Dynamic
Geodetic Deformation Models For a Kutlugun Landslide in North Eastern
Turkey’, Journal of Surveying Engineering, 129 (3), 115-124, 2003.
Mualla Y. and T. Bayrak (2003), ‘A Dynamic Model For Monitoring Landslides
With Emphasis On Underground Water In Trabzon Province, North Eastern
Turkey’, Journal of Surveying Engineering, 129 (3), 115-124, 2003.
Newcastle Scientific TM (2004), ‘Libraries for Numerical Analysis’,
http://www.mathfunctions.com/
Ogaja, C. (2001), ‘On-line GPS Integrity Monitoring and Deformation Analysis
for Structural Monitoring Applications’, To be pres. 14th Int. Tech. Meeting of
the Satellite Division of the U.S. Inst. of Navigation, Salt Lake City, Utah, 11-
14 September, 989-999.
Ogaja, C., C. Rizos and S. Han (2000), ‘Is GPS Good Enough For Monitoring the
Dynamics Of High Rise Buildings’, School of Geomatic Engineering, The
University of New South Wales, Sydney, Australia.
http://www.gmat.unsw.edu.au/snap/publications/rpb_paper.pdf.
Ogaja, C., C. Rizos, and J. Wang (2001), ‘Toward The Implementation Of On-line
Structural Monitoring Using RTK-GPS and Analysis Of Results Using The
Wavelet Transform’, The 10th FIG International Symposium in Deformation
Measurements, California, USA.
Ogaja, C., C. Rizos, J. Wang, and J. Brownjohn. (2001), ‘High Precision Dynamic
GPS System for On-line Structural Monitoring’, 5th Int. Symp. on Satellite
Navigation Technology & Applications, Canberra, Australia, paper 35.
91
Pelzer H. (1986), ‘Deformation Analysis’, Course on Applied Geodesy for Particle
Accelerators - Report, CERN, April 1986.
Salzmann M.A. (1988), ‘Some Aspects of Kalman Filtering’, Technical Report,
Surveying Engineering University of New Brunswick.
South Dakota Department of Transportation (2005), ‘CHAPTER 4: GPS FIELD
SURVEYS’,
http://www.sddot.com/pe/roaddesign/docs/surveymanual/smchap4.pdf.
Stephen R.Deloach, Members, ASCE (1989), ‘Continuous Deformation Monitoring
with GPS’, Journal of Engineering, Vol.115, No.1.
Takenaka Corporation (2001), ‘Japan's First "Flat Slab Base-Isolated High-Rise
Condominium Building" Under Construction’, http://www.takenaka.co.jp
Talbot N. (1993), ‘Real-Time High Precision GPS Positioning Concepts: Modeling,
Processing and Results’, GPS Redbooks, Volume IV, ION Publications
Department.
Tamura Y., Yoshida A., M. Matsui, S. Ishibashi, L. C. Pagnini (2004),
‘Measurement of Wind-induced Response of Buildings Using RTK-GPS and
Integrity Monitoring’. Tokyo Institute of Polytechnics, Atsugi, Kanagawa,
Japan.
http://www.arch.t-kougei.ac.jp/tamura/Staff/yoshida/AWAS2002_GPS.pdf.
Trimble Navigation Limited (2005), ‘All about GPS’, http://www.trimble.com
Volker J. (2002), ‘GPS on the Web: GPS Volcano Deformation Monitoring’, School of
Surveying and Spatial Information Systems, The University of New South Wales,
Sydney, Australia. http://www.gmat.unsw.edu.au/snap/publications/janssen_2002.pdf.
92
Volker J. and C. Rizos (2002), ‘Mixed-Mode GPS Network Processing For Deformation
Monitoring Applications in the Equatorial Region’, School of Surveying and Spatial
Information Systems, The University of New South Wales, Sydney, Australia.
http://www.gmat.unsw.edu.au/snap/publications/janssen_etal2002b.pdf
W. Fisher Cassie and J.H. Napper (1966), ‘Structure In Building’, London: The
Architectural Press.
Wan Aziz W. A. (2003), ‘High Rise Engineering Structures Stability Monitoring
Using Integration of Geodetic and Satellite Methods’, Research Vote No
72308 Report, Faculty of Geoinformation Science and Engineering,
Universiti Teknologi Malaysia.
Yehuda B., Paul J. de Jonge, David Honcil, M. Bevis, L. Bock and S. Wilson
(2000), ‘Epoch By Epoch Positioning Applied To Dam Deformation
Monitoring at Diamond Valley Lake, Southern California’, Geodetics Inc.
http://rincon.gps.caltech.edu/FIG10sym/toc.htm
Zhang, K. and Roberts.C. (2003), ‘Network-based Real Time Kinematic Positioning
System: Current Development in Australia’, Geoinformation and Surveying
Conference 2003, the Institute of surveyor, Malaysian.
93
APPENDIX A
SPECIFICATIONS OF LEICA GPS SYSTEM 500
Receiver specifications Receiver technology
ClearTrak - patented. Multibit, SAW filters. Fast acquisition. Strong signal. Low noise.Excellent tracking, even to low satellites and in adverse conditions. Interference resistant.Multipath mitigation.
No. of channels 12 L1 + 12 L2 L1 measurements Carrier phase full wave length
C/A narrow code Precision code
L2 measurements Carrier phase full wave length with AS off or on P2 code / P-code aided under AS Equal performance with AS off or on
Independent measurements
Fully independent L1 and L2 code and phase measurements
Time to first phase measurement after switching ON
Typically 30 secs
Ports
3 RS232/power ports 1 RS232 only port 1 Power only port 1 TNC port for antenna 1 PPS, 2 Event port optional
Supply voltage Power consumption
Nominal 12V DC 7W, receiver with terminal
Dimensions: L x B x Ht Weight, receiver only
205mm x 165mm x 72mm 1.25kg
Measurement precision with AS off or on
Carrier phase on L1 Carrier phase on L2 Code (pseudorange) on L1 Code (pseudorange) on L2
0.2mm rms 0.2mm rms 5cm rms 5cm rms
Accuracy, baseline rms
Accuracy in position = baseline rms. Accuracy in height = 2 x accuracy in position
94
Baseline rms with post processing
With SKI-Pro L1/L2 software
Static (phase), long lines, long observations, choke-ring antenna
3mm + 0.5ppm
Static and rapid static (phase) with standard antenna)
5mm + 0.5ppm (rms)
Kinematic (phase), in moving mode after initialization
10mm + 1ppm (rms)
Code only Typically 25cm (rms) Baseline rms with real-time /RTK
Real-time/RTK standard
Rapid static (phase), static mode after initialization
5mm + 0.5ppm (rms)
Kinematic (phase), moving mode after initialization
10mm + 1ppm (rms)
Code only Typically 25cm (rms) On-the-fly initialization
Real time and post processing
Reliability of OTF initialization Better than 99.99% Time for OTF (on-the fly) initialization
Real time: Typically 10secs with 5 or more satellites on L1 and L2 Post processing: Typically 40 seconds with 5 or more satellites on L1 and L2
Range for OTF (on-the fly) initialization.
Typically up to 20km in normal conditions with standard radio. Up to 30km in favourable conditions with powerful radio.
Note on accuracies and times
Baseline rms, accuracy in position and accuracy in height are dependent upon various factors including number of satellites, geometry, observation time, ephemeris accuracy, ionospheric conditions, multipath etc. Figures quoted assume normal to favourable conditions. Times can also not be quoted exactly. Times required are dependent upon various factors including number of satellites, geometry, ionospheric conditions, multipath etc
Position update and latency
RTK and DGPS standard
Position update rate: Position latency:
Selectable: 0.1 sec (10Hz) to 60 secs 0.03 sec or less
Real-time RTK and GPS/RTCM
Real-time RTK standard DGPS/RTCM standard
RTK and RTCM formats for transmission and reception.
Leica proprietary format CMR, RTCM V2.1 and V2.2 formats messages 1,2,3,9,18, 19,20,21,22 (Message 9, input only).
95
Number of radio modems Real-time RTK standard DGPS/RTCM standard
No. of radio modems that can be connected
2 for RTK/RTCM transmission 1 for reception 1 for NMEA transmission
Environmental specifications
Receiver Terminal Antenna PCMCIA Flash RAM cards Optional internal memory Humidity Weather Transport Usage
Operation Storage -20°C to +55°C -40°C to +70°C -20°C to +55°C -40°C to +70°C -40°C to +75°C -40°C to +75°C -20°C to +75°C -40°C to +75°C -20°C to +55°C -40°C to +70°C up to 95%; not condensing Withstand rain, snow, dust, sand, cold, heat Withstand rough field transport, shocks, jolts, vibrations etc when packed in instrument container Built for field use
96
APPENDIX B
SPECIFICATION OF TRIMBLE 4800 GPS SYSTEM
Physical Size: 23 cm (9″) D x 17.8 cm (7″) H Receiver weight: 2.1 kg (4.6 lb) with internal radio 3.9 kg (8.5 lb) as full RTK rover Electrical Receiver power: 6 Watts, receiver only 7 Watts as full RTK rover 10.5 to 20 VDC Battery life (typical): >4 hours as full RTK rover including internal radio and TSC1, with 1 Li-Ion battery. Certification: FCC & CE mark approved Environmental Operating temp: –40°C to +55°C (-40°F to +131°F) Storage temp: –20°C to +75°C (-40°F to +167°F) Humidity: 100%, fully sealed. Buoyant Shock: 2m (6ft) accidental pole drop Static Survey Performance Modes: Quick-start, Static survey, FastStatic survey Accuracy: Horizontal: ±5mm+0.5ppm Vertical: ±5mm+1ppm Azimuth: ±1arc second + 5/baseline length in kilometers Kinematic Survey Performance (Postprocessed) (Requires TSC1™ data collector with Trimble Survey Controller™ software at rover.) Modes: Continuous, Stop & go Accuracy: Horizontal: ±1cm + 1ppm Vertical: ±2cm + 1ppm Occupation: Continuous: 1 measurement Stop & go: 2 epochs (min) with 5 satellites Fastest datalogging rate: 1 Hz
97
Real-time Survey Performance Modes: Real-time Kinematic (RTK), Real-time Differential (DGPS) Real-time DGPS accuracy: 0.2m + 1ppm RMS RTK accuracy: Mode Latency Accuracy Horizontal: 1Hz fine 0.4 second ±1cm+1ppm 5Hz fine 0.1 second ±3cm+2ppm Vertical: 1Hz fine 0.4 second ±2cm+1ppm 5Hz fine 0.1 second ±5cm+2ppm Range: Range varies depending on radios used, local terrain and operating conditions. Multiple radio repeaters may be used to extend range, depending on type used. Initialization Mode: Automatic while stationary Automatic while moving on the fly (OTF) Time: <1 minute (typical) < 10 seconds (typical for known points or RTK initializer) Reliability: >99.9% Performance criteria are a function of the number of satellites visible, occupation time, observation conditions, obstructions, baseline length and environmental effects, and are based on favorable atmospheric conditions. Assumes five satellites (minimum) tracked continuously with the recommended antenna using the recommonded static surveying procedures utilizing L1 and L2 signals at all sites; precise ephemerides and meteorological data may be required. Performance specifications are RMS and ppm values are times baseline length. General Performance Start-up: <30 seconds from power-on to start survey with recent ephemeris Measurements: L1 C/A code, L1/L2 full cycle carrier Fully operational during P-code encryption Number of channels: 18 Datalogging: In internal memory; in TSC1 data collector; or on TSC1 optional removable PC card Receiver data storage: 50 hours internal memory of L1/L2 data, 6 satellites, 15 second interval Unlimited data storage using optional TSC1 and PC data card
98
Internal Radio Modem and Antenna Performance (Requires internal radio modem and internal radio antenna option.) Modes: High gain UHF Range: Base Radio Modem TRIMTALK™ 450S TRIMMARK™ IIe Optimal: 10km 15km Typical: 3–5km 10–12km Varies with terrain & operating conditions. Repeaters may be used to extend range depending on type of radios used. Radio Modem: Freq. range: 410-420 MHz, 430–440MHz, 440–450MHz, 450–460 MHz or 460–470 MHz (only one per model) Channels: Up to 20 (factory pre-set) Channel spacing: 12.5 KHz or 25KHz (only one per system) Wireless data rates: 4800 and 9600bps Modulation: GMSK Broadcast frequency, transmit power, channel spacing and antenna gain are regulated by country-of-use. These are unique on a per country basis. The broadcast frequencies, channel spacing and country-ofuse for the radio modem must be specified at time of order. Contact your Trimble representative for further information.
99
APPENDIX C
ONE HOUR CONTINUOUS RTK-GPS OBSERVATION DATA FOR UTMB AND UTMR
100
APPENDIX D
HALF HOUR CONTINUOUS RTK-GPS OBSERVATION DATA FOR UTMB AND UTMR
101
APPENDIX E
5 MINUTES OBSERVATION DATA FOR T200 (BASE) AND TR2300 (ROVER)
102
APPENDIX F
2 MINUTES OBSERVATION DATA FOR T300 (BASE) AND TR2300 (ROVER)
103
APPENDIX G
NETWORK ADJUSTMENT REPORT (TRIMBLE GEOMATIC OFFICE)
Project : First_Static
User name Administrator Date & Time 19:31:51 28/12/2004
Coordinate System Malaysian RSO Grid Zone Malaysia
Project Datum MRT 1994 Vertical Datum Geoid Model Not selected Coordinate Units Meters Distance Units Meters Height Units Meters
Adjustment Style Settings - 95% Confidence Limits
Residual Tolerances
To End Iterations : 0,000010m Final Convergence Cutoff : 0,005000m
Covariance Display
Horizontal Propogated Linear Error [E] : U.S. Constant Term [C] : 0,00000000m Scale on Linear Error [S] : 1,96
Three-Dimensional Propogated Linear Error [E] : U.S. Constant Term [C] : 0,00000000m Scale on Linear Error [S] : 1,96
Elevation Errors were used in the calculations.
104
Adjustment Controls
Compute Correlations for Geoid : False
Horizontal and Vertical adjustment performed
Set-up Errors
GPS Error in Height of Antenna : 0,000m Centering Error : 0,000m
Back to top
Statistical Summary
Successful Adjustment in 1 iteration(s)
Network Reference Factor : 1,10 Chi Square Test (a=95%) : PASS Degrees of Freedom : 10,00
GPS Observation Statistics
Reference Factor : 1,10 Redundancy Number (r) : 10,00
Individual GPS Observation Statistics
Observation ID Reference Factor Redundancy Number B1 1,65 1,94
B2 0,78 2,43
B3 1,19 1,15
B4 0,93 1,57
B5 0,81 1,62
B6 0,98 1,30
Weighting Strategies
GPS Observations Default Scalar Applied to All Observations
105
Scalar : 1,00
Adjusted Coordinates
Adjustment performed in WGS-84
Number of Points : 4 Number of Constrained Points : 2 Horizontal and Height Only : 2
Adjusted Grid Coordinates
Errors are reported using 1,96s.
Point Name Northing N
error Easting E error Elevation e
error Fix
B1 161729,655m 0,000m 642138,997m 0,000m N/A N/A N E h
R2 161798,357m 0,001m 641889,211m 0,001m N/A N/A
B2 161746,439m 0,000m 641594,525m 0,000m N/A N/A N E h
R1 161783,576m 0,001m 641936,381m 0,001m N/A N/A
Adjusted Geodetic Coordinates
Errors are reported using 1,96s.
Point Name Latitude N
error Longitude E error Height h
error Fix
B1 1°27'45,14692"N 0,000m 103°46'26,01444"E 0,000m 11,261m 0,000m Lat
Long h
R2 1°27'47,38060"N 0,001m 103°46'17,93167"E 0,001m 137,441m 0,002m
B2 1°27'45,68590"N 0,000m 103°46'08,39805"E 0,000m 26,604m 0,000m Lat
Long h
R1 1°27'46,89996"N 0,001m 103°46'19,45804"E 0,001m 137,433m 0,002m
106
Adjusted Observations
Adjustment performed in WGS-84 GPS Observations GPS Transformation Group: <GPS Default> Azimuth Rotation : -0°00'04,9568" (1,96s) : 0°00'00,2674" Network Scale : 0,99998457 (1,96s) : 0,00000194 Number of Observations : 6 Number of Outliers : 0
Observation Adjustment (Critical Tau = 2,50). Any outliers are in red.
Obs. ID
From Pt.
To Pt. Observation
A-posteriori Error (1,96s)
Residual Stand. Residual
B1 B1 R2 Az. 285°21'11,3338" 0°00'00,6086" 0°00'00,4690" 1,28
DHt. 126,182m 0,002m 0,000m -0,23
Dist. 259,100m 0,001m -0,002m -2,44
B3 B2 R2 Az. 79°58'54,5194" 0°00'00,4128" -0°00'00,0841" -0,53
DHt. 110,839m 0,002m 0,001m 0,78
Dist. 299,261m 0,001m 0,000m -1,30
B6 B2 R1 Az. 83°46'25,8764" 0°00'00,3794" -0°00'00,0179" -0,10
DHt. 110,831m 0,002m -0,001m -1,11
Dist. 343,909m 0,001m 0,000m 0,02
B4 B1 R1 Az. 284°52'37,3966" 0°00'00,7287" -0°00'00,3231" -0,88
DHt. 126,174m 0,002m 0,001m 0,45
Dist. 209,699m 0,001m 0,000m 0,90
B2 B2 B1 Az. 91°44'23,4599" 0°00'00,2674" -0°00'00,0980" -0,35
DHt. -15,343m 0,000m -0,002m -0,73
Dist. 544,802m 0,001m 0,000m 0,48
B5 R2 R1 Az. 107°22'22,4697" 0°00'02,7857" -0°00'01,0705" -0,70
DHt. -0,006m 0,002m 0,001m 0,70
Dist. 49,438m 0,001m 0,000m 0,70
107
APPENDIX H TRIMBLE GEOMATICS OFFICE DATA DOWNLOADING PROCEDURES
108
APPENDIX I
LEICA SKI PRO DATA DOWNLOADING PROCEDURES
109
APPENDIX J
OBSERVATION SCHEDULE OF ‘TIMING’ SIMULATION TEST
The pole was seated on the GPS point during observation from 13:52:43 until 13:55:29. The below shows the examples of deformation report (single point test) which is performed by KFilter program for that period of time. ------Single Point Test ( 13:52:43 )------- Difference,cm t-calculate t-table Result 0.304 0.32 1.96 Stable -0.107 0.11 1.96 Stable 0.303 0.17 1.96 Stable … … ------Single Point Test ( 13:55:29 )------- Difference,cm t-calculate t-table Result 0.360 0.38 1.96 Stable -0.145 0.15 1.96 Stable 0.214 0.12 1.96 Stable Then, the pole was ‘vibrated’ from 13:55:34 until 13:56:44. The below shows the examples of KFilter program deformation report for that time span. ------Single Point Test ( 13:55:34 )------- Difference,cm t-calculate t-table Result -3.136 3.34 1.96 Moved -0.258 0.27 1.96 Stable 0.143 0.08 1.96 Stable … … ------Single Point Test ( 13:56:44 )------- Difference,cm t-calculate t-table Result 3.650 3.88 1.96 Moved -3.622 3.85 1.96 Moved 0.817 0.46 1.96 Stable Then, the pole was return and seated on the GPS point again from 13:56:49 until 13:58:40. The KFilter program deformation report is shown as below.
110
------Single Point Test ( 13:56:49 )------- Difference,cm t-calculate t-table Result 0.511 0.54 1.96 Stable -0.258 0.27 1.96 Stable 0.870 0.49 1.96 Stable … … ------Single Point Test ( 13:58:40 )------- Difference,cm t-calculate t-table Result 0.229 0.24 1.96 Stable -0.164 0.17 1.96 Stable 0.427 0.24 1.96 Stable Then, the pole was ‘vibrated’ again from 13:58:45 to 14:00:05. The deformation report KFilter program processing is shown in the followings. ------Single Point Test ( 13:58:45 )------- Difference,cm t-calculate t-table Result -2.290 2.44 1.96 Moved -0.013 0.01 1.96 Stable 0.480 0.27 1.96 Stable … … ------Single Point Test ( 14:00:05 )------- Difference,cm t-calculate t-table Result 3.443 3.66 1.96 Moved 2.938 3.13 1.96 Moved 1.278 0.72 1.96 Stable At last, the pole was seated on the GPS point again from 14:00:10 until 14:01:01. The examples of deformation report (single point test) which is performed by KFilter program is shown as below. ------Single Point Test ( 14:00:10 )------- Difference,cm t-calculate t-table Result 0.379 0.40 1.96 Stable 0.137 0.15 1.96 Stable 0.977 0.55 1.96 Stable … … ------Single Point Test ( 14:01:01 )------- Difference,cm t-calculate t-table Result 0.435 0.46 1.96 Stable 0.344 0.37 1.96 Stable 1.154 0.65 1.96 Stable
111
APPENDIX K
SPECIFICATION OF ANEMOMETER DAVIS
General Sensor Type Wind Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wind cups and magnetic switch Wind Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wind vane and potentiometer Attached Cable Length. . . . . . . . . . . . . . . . . . . . . . 40’ (12 m) Cable Type . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . 4-conductor, 26 AWG Connector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modular connector (RJ-11) Recommended Maximum Cable Length Wizard and Monitor . . . . . . . . . . . . . . . . . .. . . . . . . . 140’ (42 m) Sensor to Console GroWeather and EnviroMonitor . . . . . . . . . . . . . . . . 250’ (75 m) from Sensor to SIM
+ 250’ (75 m) from SIM to Console
Material Wind Vane and Control Head . . . . . . . . . . . . . . . . . . UV-resistant ABS Wind Cups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polycarbonate Anemometer Arm . . . . . . . . . . . . . . . . . . . . . . . . . . . Black-anodized aluminum Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.5" long x 7.5" high x 4.75"
wide (470 mm x 191 mm x 121 mm)
Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 lbs. 15 oz. (1.332 kg) Console Data (These specifications apply to sensor output as converted by Davis Instruments weather station consoles.) Range Wind Speed (See Note 1) . . . . . . . . . . . . . . . . . . . . . 0 to 175 mph (150 knots, 78 m/s, 280 km/hr) Wind Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0° to 360° or 16 compass points Wind Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 to 1999.9 miles (1999.9 km) Accuracy Wind Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±5% Wind Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±7° Wind Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±5% Resolution Wind Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 mph (1 knot, 0.1 m/s, 1 km/hr) Wind Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1° (0° to 355°), 22.5° between compass points Wind Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.1 m (0.1 km) Measurement Timing Wind Speed Sample Period . . . . . . . . . . . . . . . . . . . . 2.25 seconds
112
Wind Speed Sample and Display Interval . . . . . . . . . 2.25 seconds (Monitor & Wizard), 3 seconds (GroWeather & EnviroMonitors)
Wind Direction Sample Interval. . . . . . . . . . . . . . . . . 1 second (Monitor & Wizard), 1.5 seconds (GroWeather & EnviroMonitors)
Wind Direction Filter Time Constant (typical). . . . . . 8 seconds (Monitor & Wizard), 6-9 seconds (GroWeather & EnviroMonitors)
Wind Direction Display Update Interval . . . . . . . . . . .2 seconds (Monitor & Wizard), 1 second (GroWeather & EnviroMonitors)
Wind Run Sample and Display Interval . . . . . . . . . . . 3 seconds (GroWeather & Energy EnviroMonitor)
WeatherLink¨ Data (These specifications apply to sensor output as logged and displayed by the WeatherLink.) Wind Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Average during archive
interval High Wind Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Maximum during archive
interval Wind Direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dominant wind direction
during archive interval Wind Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sum over archive interval
(GroWeatherLink & Energy WeatherLink)
Input/Output Connections Black . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wind speed contact closure to ground Green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wind direction pot wiper
(360° = 20 kOhm) Yellow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pot supply voltage Red. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground
113
APPENDIX L
DEFORMATION REPORT FOR GPS DEFORMATION ANALYSIS PROGRAM
GPS DEFORMATION ANALYSIS PROGRAM Asst. Prof. Dr. Temel BAYRAK, 2004 [email protected] Deformation Analysis Report 06-May-2005 01:07:41 Number of Points : 4 ----------------------------------------------------------------------------- TetaSquare R : 6.049 H : 9 s02=(vTpv1+vTpv2)/(df1+df2) s02 : 2.143 Test Value T=R/(s02*H) T : 0.314 F_table 9 24 q : 2.703 Significance level for F_table value : 0.975 Good News T < q !!! There is NoT deformations/Moving Points in the Networks :)
114
APPENDIX N
DEFORMATION REPORT FOR GPSAD2000
2 Epoch Deformation Analysis Summary ===============================
Date analysis = 5/6/05 Total Station = 4 ================================================================ VARIANS RATIO TESTS ==================== Testing On Variance Ratio At Significance level 0.05 Degree Of Freedom Epoch 1 = 10 Degree Of Freedom Epoch 2 = 10 Pool Variance Factor = 1.05 Varians Ratio Test Pass Where (Ratio Test)1.1 < 2.984(Passing Level) ================================================================= GLOBAL CONGRUENCY TEST ========================= Significance Level For Congrunsy testing =0.05 Total Degree Of Freedom = 20 H = 9 [Total datum stations x 3 - datum defect] Global Congruency Test Pass Where (Congruency Test)0.000 < 2.393(Passing Level) No Deformation Detected ================================================================= SINGLE POINT TEST ================= Significance Level For Single Point Test =0.01 Total Degree Of Freedom = 20
115
H = 3 Stn Dx Dy Dz Dis. Vector fcom ftab info 1 0.0021 0.0017 0.0027 0.0039 0.00 4.94 stable [1] 2 0.0021 0.0017 0.0027 0.0039 0.00 4.94 stable [1] 3 -0.0036 0.0066 -0.0027 0.0080 0.00 4.94 stable [1] 4 -0.0007 -0.0102 -0.0028 0.0105 0.00 4.94 stable [1] [codes:1 = datum pts, 0 = non-datum pts] Total Station = 4 Number of Datum Stesens/stable = 4 Number of Datum Stesens/moved = 0 Number of non-datum stns/stable = 0 Number of non-datum stns/moved = 0 End Of File
116
APPENDIX M
DEFORMATION REPORT FOR KFilter ------Single Point Test ( 14:15:40 )------- Difference,cm t-calculate t-table Result 0.395 0.42 1.96 Stable 0.110 0.12 1.96 Stable -0.130 0.07 1.96 Stable ------Single Point Test ( 14:15:45 )------- Difference,cm t-calculate t-table Result 0.151 0.16 1.96 Stable 0.354 0.38 1.96 Stable -0.237 0.13 1.96 Stable ------Single Point Test ( 14:15:50 )------- Difference,cm t-calculate t-table Result 0.207 0.22 1.96 Stable 0.166 0.18 1.96 Stable -0.255 0.14 1.96 Stable ------Single Point Test ( 14:15:55 )------- Difference,cm t-calculate t-table Result 0.094 0.10 1.96 Stable 0.373 0.40 1.96 Stable -0.343 0.19 1.96 Stable ------Single Point Test ( 14:16:00 )------- Difference,cm t-calculate t-table Result -0.207 0.22 1.96 Stable -0.360 0.38 1.96 Stable 0.136 0.08 1.96 Stable ------Single Point Test ( 14:16:05 )------- Difference,cm t-calculate t-table Result -0.301 0.32 1.96 Stable 0.298 0.32 1.96 Stable -0.166 0.09 1.96 Stable ------Single Point Test ( 14:16:10 )------- Difference,cm t-calculate t-table Result -0.056 0.06 1.96 Stable 0.279 0.30 1.96 Stable 0.029 0.02 1.96 Stable