Smart Information for a Sustainable World

True Adaptive Signal Control A Comparison of Alternatives

Technical Paper #1154

1

Presentation to the 18th World Congress on Intelligent Transport Systems

Technical Session 95 (TS95)Wednesday 19 October 2011

19 Oct. 2011

Authors

Cary Vick, PE – Telvent Transportation (presenting)

Farhad Pooran, PE, PhD - Telvent Transportation

José Carlos Riveira Martinez - Telvent Trafico y Transporte

2

Purpose / Agenda

Clarify definitions and capabilities of true adaptive signal control

Urge selection of some type of advanced signal control

We can do better with what we haveIt’s not just for research any more

Summarize approach of available technologies / products

Not a comprehensive review and comparison of results

3

Deployment of Adaptive Signal Control Technologies

SCATS

SCOOT

ACSLite

InSync

RHODES

OPAC

LA ATCS

State with Active ASCT

State with Pending ASCT

4

Categories of Signal Control

Definitions (USDOT / FHWA Traffic Signal Control Handbook (Gordon and Tighe)):

Isolated Time Based CoordinationInterconnected Control Traffic Adjusted Control Traffic Responsive Control Traffic Adaptive Control

The primary differentiator for Traffic Adaptive Control is the use of a traffic flow model to predict demand and adapt timings to meet expected traffic flows

5

What is NOT Adaptive ControlTraffic Responsive:

Trigger plan or timing changesVolume Density:

Modify parameters based on demandOff-line Optimization:

Real time measurements trigger re-optimization of timing plans

Central + Local Optimization:

Real time measurements trigger re-optimization of selected timing parameters (off set) then local control applies and adjusts

All approaches have value, but they are NOT adaptive signal control

6

Summary of ASCT Alternatives Function

Control System

  

ACS Lite InSync LA ATCS OPAC RHODES SCATS SCOOT UTOPIA ITACA NAZTEC QuicNet

Action P + R P + R P + R P + R P (only)

R(only)

P + R Pro-active P + R R(only)

R(only)

Architecture Dist. Dist. Central Dist. Central Central Central Central Central Central Central

Simulation Model?

Yes Yes Yes Yes Yes No Yes Yes Yes No No

Time Frame 5–10 min Phase/ Cycle / 15 min.

Cycle Phase/ Cycle / 5 min.

Sec by sec

Cycle Cycle/ 5 min.

3 sec & Cycle

5 sec & Cycle

5–10 min 5–10 min

Optimization S / O S / C / O / PS

S / C / O S / C / O

S S / C / O

S / C / O / PS

S / PS S / C / O / PS

S / C / O S / C / O

7

8

Adaptive Traffic Signal Control Processes

Offset Optimizer Split Optimizer Cycle Optimizer

Offset weights

Splitweights

Cycleweights

Intersection control

Platoon modeling

VolumeOccupancy

Traffic Engineer

B

A

A + BA or B

Expert Rules

Control Center

Traffic Flow / Queue Model

€

€Street Observer for Calibration

Operator / Traffic Engineer

Traffic data at detection ponts

Intersection structure

Mean travel timeQueue lengthsExit flowsQueue capacity

Operations of Adaptive Signal Control

Real Time DataPro Active Timing Changes

9

Weighting Factors Apply Expert Rules to Adaptive Control Strategies

•ITACA Expert Rules System •Usable range is -10 to +10• Positive weights apply importance to affected links• Negative weights reduce importance of associated links• 2 weighting parameters for each link (split and offset)• 1 weighting parameter for each intersection (cycle)• Can be adjusted in real time manually or by expert rules• Each final weight may reflect adjustments by different sources

If Then UntilRoute A is highly congested

Reduce weight for Transversal Route B

Intersection B20has maximum queue

10

Cabinet

OPAC Hardware ConfigurationAdvanced Traffic Controllers OPAC Single Board Computer (local)Upstream Advance Detectors

Sto

p B

ar

Phase Detector

Phase Detector

Phase Detector

Adaptive Detector

Adaptive Detector

Adaptive Detector

~15 secs travel time at prevailing speeds Single Board Computer (Linux OS) running OPAC

software

Traffic Signal Controller:

Type 2070 or NEMA (TS1 / TS2)

Force offs & Holds

Detector Data

11

ACS-Lite Architecture

12

ACS-Lite Computer

Communications Interface and/orNTCIP Translators

NTCIP controllers orProprietary protocol controllers

Minimum 9600 baudCommunications(serial or IP) NTCIP

On-street Master(optional)

Advantages of Predictive Modeling

Continuous adjustment to changing demand

Pro-active timing changes

Optimal use of available capacity

Adjusts to incident conditions

The primary differentiator for Traffic Adaptive Control is the use of a traffic flow model to predict demand and adapt timings to meet expected traffic flows

13

Cycle Length Adaptability

14

Incident

Occurs

References:Stevanovic, Aleksandar, “Adaptive Traffic Control Systems: Domestic and Foreign State of Practice” NCHRP Synthesis 403, Transportation Research Board, Washington, DC, 2010

Gordon, Robert L, and Tighe, Warren “Traffic Control Systems Handbook.” Federal Highway Administration Report FHWA-HOP-06-006, Washington, DC, 2005.

Gartner, N.H., F.J. Poorhan, and C.M. Andrews. “Implementations and Field Testing of the OPAC Adaptive Control Strategy in RT-TRACS.” TRB Paper No. 02-3667 - Presented at the 81st Annual Meeting of the Transportation Research Board, Washington, DC, 2002.

Riveira Martínez, José Carlos. “The Road to ITACA”, internal white paper, Telvent, 2000.

MacGowan, J., and I. J. Fullerton. 1979. Development and testing of advanced control strategies in the urban traffic control system. Public Roads 43, no. 2, 3, and 4.

15

Smart Information for a Sustainable World

Thank You

Smart Information for a Sustainable World

19 Oct. 2011