Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
2018/11/21
1
• Chinese geodesist and educator;
• President of Duke Kunshan University from 2012.
• Academician of Chinese Academy of Engineering since 1999;
• Former president of Wuhan University from 2003 to 2008;
• Have developed the first GPS satellite positioning data processing system in China;
• Participating the design of National High Precision GPS Network;
Introduction to Pro. Liu Jiangnan
Intelligent Transportation and Autonomous Vehicles
Organizer:
2018/11/21
2
Historic evolution of using maps in the car
����� ���
����� � ���������� � ��� ���
���������� ��������� �!" #$$
� ��� %�&' �(')*
+� %�&' � ��� ,-
Foreword
Contents
1. Background and demands
2. Compared with traditional map
3. Main elements and layers
4. Outlook of technologies and standards
5. Summary and thoughts
2018/11/21
3
� ./0123 4560758029 :58/;;08/ :<=8/>?.1:@
• AB095 0= C30;D097 .1:EFG 59D H0;; IJ260D/ =/J60K/=
09K;3D097 :LM?:/5JKB 59D M/=K3/@ 59D :N:?:B2J8
>/==57/ =/J60K/ @ H2J;DH0D/ 5J239D OPOP
• QB/ RS�GRT� ./0123EF =58/;;08/ H5= ;539KB/D 29 426URT
• .1: 0= 5C238 82 =/J6/ ./;8 59D M25D 8B0= </5J
1.1 Advancement of GNSS technology
� ./0123 VJ239DEC5=/D L37>/9858029 :<=8/>?.1V.L:@
• WXYZX[\ ]Y\ ^\_`abcdeff ghg]\i jh jkaZl[Zg]_Xm gZ]\nn_]\
g_mXZn \kkak [akk\[]_aXg
• ^`fop q_nn j\ Zjn\ ]a rkas_l\ k\Zno]_i\ rk\[_g\ rag_]_aX ZXl
Zbmi\X] g\ks_[\ tak XZs_mZ]_aX q_]Y i\]\k ]a [\X]_i\]\k
n\s\n Z[[bkZ[h ]a [Y_XZ
� LKK3J58/ D<95>0K I2=08029097 u v07B IJ/K0=029 >5I=
9560758029 9//D= w2J L38292>23= 1J06097
• [aXX\[] rk\[_g\ deff ]\[YXanamh q_]Y Y_mY rk\[_g_aX iZrg
]a i\\] ]Y\ l_s\kg\ X\\lg _X ]Y\ _X]\nn_m\X] \kZ
x'�yz{ �)z{|y x*'y ,{�}'|�z| �(*'}
1.1 Advancement of GNSS technology
2018/11/21
4
� VJ2H097 9//D= 2w QJ5ww0K N5957/>/98 59D
QJ59=I2J858029 ~;599097 J/�30J/ IJ/K0=029 928 29;<
J25D ;/6/;�C38 5;=2 ;59/ ;/6/; H08B >/8/J ;/6/;
• 4560758029 w2J J/6/J=0C;/ ;59/= 59D v�� �59/=
• �59/ =3I/J60=029 09 8J5ww0K 602;58029
• �9;09/ D/8/J>0958029 2w J/=I29=0C0;08< 59D ;2==
w2J 6/B0K;/ 5KK0D/98=
• �=57/ .5=/D �9=3J59K/?�.�@
1.2 Accurate and intelligent trends of traffic management
1.2 Accurate and intelligent trends of traffic management
� ��������� � ����������� ��������������� ����
• �biZXo[\X]\k\l
• �airk\Y\Xg_s\nh rkaia]\ aXn_X\ _X]\mkZ]_aX ZXl gYZk_Xm at
]kZXgrak]Z]_aX k\gabk[\g� gb[Y Zg ]kZXgrak]Z]_aX _XtkZg]kb[]bk\�
]kZXgrak]Z]_aX ]aang� ]kZXgrak]Z]_aX ghg]\i� \][�
• gZt\ ZXl [aXs\X_\X] iaj_n_]h� mk\\X� _X]\nn_m\X]� bj_�b_]abg g\ks_[\
• rkas_l\ na[Z]_aX g\ks_[\ q_]Y rk\[_g\ ]_i\ ZXl rag_]_aX _X q_l\ Zk\Z
2018/11/21
5
� ������� �������� ��������
Intelligent High-Precision Map(IHPM) is the carrier of high-
precision environment perception and the foundation of real-time
road control for autonomous/unmanned driving
��' |�'y �'* ���� y'&'�z �|'����'| -|')|' z¡ ¢'��£�'* * �' ¤'(�z)¤ £z|'| ¡z) �' y'&'�z }'| z¡ �|'����'| )|* z)�z| *(*'}*��' ¡'y')� *'�¡¥y)�&�|� &'��£�'* z��£( �* ¦''| {¦��*�'y �| §¨©ª
��' �z&')|}'| £�&'�( )z}z'* £)z**¥*'£z)� *(|')�( |y )z}z'*�' �} �'}'|�z| z¡ �|'����'| |'�z)¤'y {z}z�&' )z«'£*¬- �* �||'y z ��z� {z}'y &'��£�'* �| )'*)�£'y )'* �| §¨§¨ |y¡z)} ¡{��( *'�¡¥y)�&�|� £) })¤' )�' �| | ¦( §¨§®
¯-||z&�&' y'&'�z }'| *)'�( ¡z) -|'����'| &'��£�'*° {¡z)�)y �' �y' z ¦{��y |�z|� ¦*�£ } *(*'} ¡z) &'��£�' ���{|�¡�'y *|y)y*± |y z '*¦��*� |y �} )z&' ²� �'z¥�|¡z)}�z|*(*'}± ¡{)�') }z)' z )z&�y' )'�¥�}' y(|}�£ } y *')&�£'
1.3 Intelligent and Connected Vehicles and Autonomous Driving
� fZt\]h k\�b_k\i\X]g at ³b]aXaiabg `k_s_Xm
• ´µ¶ ·¸¹º¶» !¼ ½��¶ ¶"��¾!"¿¶" �½ $¶¾»¶À �!" »µ�"�¶Á ¼¾!¿
µ¸¿�" ! ¿�»µ�"¶
• ¾���"� ·Ã· ¶¿ À¸ · µ��µ Á¶¿�"Á· !" ·�¼¶ à �"Á · �¹�½� à �"Á
¾¶Ä¸�¾¶· � µ��µ ¿� »µ ¹¶ Ŷ¶" �»»¸¾� ¶ ¾¶�½Æ �¿¶ À!·� �!"�"�
�"Á µ��µÆÀ¾¶»�·�!" ¾!�Á ¿�À
Contents
1. Background and demands
2. Compared with traditional map
3. Main elements and layers
4. Outlook of technologies and standards
5. Summary and thoughts
2018/11/21
6
2.1 Definition and classification of Intelligent High-Precision Map(IPHM)
� IPHM is an intelligent map that meet the following constraints:
• B5= 5C=2;38/ 5KK3J5K< 2w 8B/ K22JD0958/= C/88/J 8B59 PUR>
• K298509= 8H2 Ç09D= 2w 09w2J>58029G 29/ 0= =8580K 09w2J>58029 =3KB 5= =B5I/ 2w J25D= 59D ;59/=G
8J5ww0K K29=8J5098= 59D =3JJ239D097 8J5ww0K /960J29>/98G 5928B/J 0= =/>0ED<95>0K 2J D<95>0K
09w2J>58029 =3KB 5= J/5;E80>/ 8J5ww0K 59D 2C=85K;/=
• 5C;/ 82 IJ2K/== 09w2J>58029 H08B K2;;5C2J58029 2w K;23D K2>I38097 59D 8B/ �2Q
• 5C;/ 82 =/J6/ >3;80I;/ 5J/5= =3KB 5= D/;0K5K< >5957/>/98 2w 098/;;07/98 8J59=I2J858029 G
L38292>23= 1J06097 59D M2C28 9560758029 /8KU
2.1 Definition and classification of Intelligent High-precision Maps(IPHM)
,{z|z}z{* �)�&�|� �)¡¡�£ £z|)z� |y�|*{)|£' ��£�z| Èz¦z* É�z{y
�According to Application Scenarios,it can be divided into four
categories:high precision maps for vehicle, for control center, for
robots, and for cloud
2018/11/21
7
ÊkZl_]_aXZn ËZrg
• Ì�Í��Î����� �Ï Ð���Ð�Ð ���� Ð���Ñ� ��Í��Î����� ��Ð ��Ð���Ñ� ��Í��Î�����
� Ò���Ñ� ��Í��Î����� �Ï Ï�Î�ÓÔ ��ÍÓ�Ñ��Ð ÕÔ ����Ö�ÑÏ ��Ð ÏÔÎÕ�ÓÏ
� Ì�Ð���Ñ� ��Í��Î����� Ð����ÐÏ �� ×Ï��ØÏ �Ù� ×�Ð��Ï���Ð��� ��Ð Ï�����Ó Ð��� Î�����
�� ��Ï�Ú���Ñ�ÏÏ���
• ÛÖ� Î�� ��Í��Î����� �Ï ×�Ð���Ð Î��×�ÓÓÔ
� Information load and expression(What is the information?)
2.2 Theory development of high-precision map for vehicles
ÜX]\nn_m\X] �_mY Ýk\[_g_aX ËZrg
• Þ���� Î��� ��Í���Ðß ÐÔ��Î�Ñ ��Ð ���ÓÚ��Î�ß �Ö�Ô �×� Î��� �Î�Ö�Ï�Ï �� Ð��� Î����� ��Ð
�×��Î���Ñ �Ñà×�Ï����� �Í ��Ð���Ñ� ��Í��Î�����
• áÏ��ØÏ ×�Ð��Ï���Ð��� �Í �Ö� �Õâ�Ñ���� Ù��ÓÐ �Ï ��Ö��Ñ�Ð Í��Î Î�� Ï�����Ó ���Ñ������ ��
ÐÔ��Î�Ñ Ñ��������
• ÛÖ� Î�� ��Í��Î����� �Ï ×�Ð���Ð ÏÔ�ÑÖ����×ÏÓÔ ÙÖ�Ó� Î�� �Ï Õ���� ×Ï�Ð
� Information transmission (Where is the information from?)
ÊkZl_]_aXZn ËZrg
• �Ï �Ö� Ï�����Ó Î�Ð�ÓÏ �Í �Ö� �Õâ�Ñ���� Ù��ÓÐß �Ö�Ô ��� �Ö� Ñ��������Ö��ØÏ
×�Ð��Ï���Ð��� �Í �Ö� �Õâ�Ñ���� Ù��ÓÐ Ù��Ö�� Ñ������ ���ÎÏ
• Ì�Í��Î����� ����ÏÎ�ÏÏ��� �Ï � ���ÚÙ�Ô ���Ñ�ÏÏ Í��Î Ñ��������Ö�� �� ×Ï��
ÜX]\nn_m\X] �_mY Ýk\[_g_aX ËZrg
• �� �ã���Ï��� �Í ���Ð������Ó Î��Ï ÙÖ�ÑÖ Ñ�� Õ� ×�Ð��Ï���Ð ÕÔ Î�ÑÖ���
• ä�ÓÓ�Õ�������Ï Õ��Ù��� ���Í�ÏÏ����Ó Ñ��������Ö��Ï ��Ð Ñ��ÙÐÏ�×�Ñ��� Ð���
• áÏ��Ï �� Ó����� â×Ï� ��Ñ���� Ð���ß Õ×� �ÓÏ� �����Ñ����� �� Î�� ���Ð×Ñ����
• áÏ��ÏØ Ñ�������� ��Ð ���Ï���Ó�å�Ð ���ÐÏ Ù�ÓÓ �ÍÍ�Ñ� Í���Ó ���Ï�������� �Í �Ö�
Î�� Î�Ð�Óß �� ���Ó�å� �Ö� Ï�ÓÍÚ�Ð�������� Õ��Ù��� �Ö� Î��Ï ��Ð ×Ï��ÏØ
��à×���Î���Ï
2.2 Theory development of high-precision map for vehicles
2018/11/21
8
2.2 Theory development of high-precision map for vehicles
Traditional Maps
• QB/ 5==0=8/D D/K0=029E>5Ç097 5C0;08< ?;0Ç/ I;599097@ 2w
>5I= 0= C5=/D 29 3=/J=æ 39D/J=859D097 2w 8B/ /960J29>/98
• v3>59 0= 8B/ =3Cç/K8 09 8B/ IJ2K/== 2w >5I 3=097U .5=/D
29 8B/0J 2H9 60=35; I/JK/I8029 59D ;270K5; 8B09Ç097 5C0;08<G
3=/J= J/;< 29 7/2E09w2J>58029 K5JJ0/D C< 7J5IB0K5;;<
/èIJ/==/D >5I= 82 K2>I;/8/ =I/K0w0K 85=Ç
� The use of information (How to use the information?)
2.2 Theory development of high-precision map for vehicles
� The use of information (How to use the information?)
����������� é��ê ��������� ë���
• ÊY\ iZ[Y_X\ j\[ai\g ZXa]Y\k gbjì\[] _X ]Y\ rka[\gg at iZr bg_Xm�
ÊY\ bgZm\ _g íYbiZXoiZ[Y_X\oiZrî _X]\kZ[]_Xm q_]Y \Z[Y a]Y\k
• ïbZX]_]Z]_s\ ZXl l_m_]Zn Y_mYork\[_g_aX iZrg rkas_l\ Y_mYnh
l\]Z_n\l ZXl lhXZi_[ \Xs_kaXi\X]Zn _XtakiZ]_aX
• ÊY\ n_s\ iZr q_]Y k\Zno]_i\ r\k[\r]_aX ibg] rZk]_[_rZ]\ _X ]Y\
l\[_g_aXoiZð_Xm ZXl k\Zno]_i\ [aX]kan at lk_s_Xm� ZXl j\ Zjn\ ]a
g\nton\ZkX_Xm� g\ntoZlZr]Z]_aX ZXl g\nto\sZnbZ]_aX
2018/11/21
9
Contents
1. Background and demands
2. Compared with traditional map
3. Main elements and layers
4. Outlook of technologies and standards
5. Summary and thoughts
ñ\t\k\X[\ nZh\kg at �_mYork\[_g_aX kaZl XZs_mZ]_aX iZr
� � ��
• ³[[bkZ]\nh k\tn\[] kaZl \Xs_kaXi\X]
• ³[Y_\s\ ibn]_og[Zn\ [Zn_jkZ]_aX ZXl Y_mY
gr\\l Z[[\gg
• Ë\\] grZ]_Zn _Xl\ô_Xm k\�b_k\i\X]g tak
Ýag_]_aX_Xm� rZ]Y rnZXX_Xm ZXl XZs_mZ]_aX
� õö÷øùú
• û����Ñ Ó�Ô��ü ���Ðý �ÖÔÏ�Ñ�Ó ��Ð ���Î����Ñ
ÑÖ���Ñ����Ï��ÑÏ �Í Ó���Ïý ��Í��Ï��×Ñ�×��
• þ��ÓÚ��Î� Ó�Ô��ü ���ÓÚ��Î� ���ÍÍ�Ñ ��Ð �ÕÏ��ÑÓ�Ï
• ÒÔ��Î�Ñ Ó�Ô��ü �×����Î�×Ï Ï��Ï��� Ð��� ýÿ��
���
• áÏ�� Ó�Ô��ü Ò������ ��Ï�ý Ò������ Õ�Ö�����ý
��Ö�ÑÓ� Ñ��Í��×������
1. Static layer
• Road model
• Traffic infrastructure
• Position information
3.1 Layers for Intelligent High-precision Maps
3. Dynamic layer
• Vehicles
• Pedestrian
4. User layer
• Vehicle configuration
• Scene information
• Behavior monitoring
• Cognitive characteristics
2. Real-time layer
• Road Construction
• Traffic jam
• Traffic management
• Weather prediction
2018/11/21
10
3.2 Examples of representative elements
Light pole
Lane
Lane line
Curbs
Road boundar
y line
Lane connection
Traffic jam
Scene
occupancy grid
Vehicle perception
Trajectory record
þ��Ð ä��Ï��×Ñ���� Reversible lanes
Road sign
Traffic management
PedestrianV2X
Traffic accident
Traffic sign
Contents
1. Background and demands
2. Compared with traditional map
3. Main elements and layers
4. Outlook of technologies and standards
5. Summary and thoughts
2018/11/21
11
Data collection and updating
• Centralized professional collection
• Crowdsourcing
• Integration of professional collection
and crowdsourcing
Dynamic data interaction
• Machine vision and target detection
• Multi-source sensor collaboration
• V2X
Computing Pattern
• Cloud computing
• Edge calculation (cloud and terminal
collaboration )
Application and usage
• Lane level map matching
• Map-assisted perception and position
• planning and obstacle avoidance
Autonomous intelligent control
• Accurate match between traffic signs and
dynamic position of vehicles
• Real-time driving control through
information in Internet of Vehicle
• Environmental and behavioral monitoring
• ������� ������ �� ������ �������
Map production and data processing
• Multi-semantic segmentation
• Computer vision
• Geometric rendering and topology
4.1 Several technical problems that need to be solved
4.2 Standards development
� ISO TC204/WG3—Database construction and data exchange standards for intelligent traffic maps
� OpenDRIVE (standardize the logical road description to facilitate the data exchange between
different driving simulators)——OpenDRIVE® V 1.4 Format Specification, Revision H
� ADASIS(Define a interface specifications)——ADASIS v1�ADASIS v2�ADASIS v3
� ETSI——ETSI TR 102 863 V1.1.1 (2011-06)
� NDS Steering Committee put HD map on standardization agenda in 2013 by installing dedicated
Working Group 3,then in 2017 NDS standard is ready for 2020 autonomous cars
� Open AutoDrive Forum is the cross-domain platform driving standardizations in the area of
autonomous driving
� DMP (Dynamic Map Planning Co., Ltd. )——Study the methodologies of developing and
maintaining high-precision 3D map data for Automated Driving
� Lead by Wuhan University—— “Data Specifications of Road High-precision Electronic Navigation Map”
� Lead by Beijing University of Civil Engineering and Architecture—— “Production Technology Specifications of
Road High-precision Electronic Navigation Map”
� China ITS Industry Alliance(C-ITS) ——”Digital Map for Intelligent Vehicle Data Model & Exchange Format
Specification”
� CAICV HD MAP WG—— 3 subjects: ”Research on Automated Driving Map Standard System”(Lead by
NavInfo)�” Research on Map deflection and Communication between mapping management departments”(Lead by
Amap)�”Research on Autopilot Map Technology and Application Roadmap” (Lead by Wuhan KOTEI Informatics
Co., Ltd.�� Baidu—— Apollo OpenDrive specifcations
International standards Standards in China
� 2018-20190 Ministry of Natural Resources 1National Administration of Surveying, Mapping and Geoinformation of China2entrusts Wuhan
University, Tongji University, Beijing University of Civil Engineering and Architecture, etc. to set standard related to Surveying, Mapping and
Geoinformation Industry, and release the draft of “Data Specifications of Road High-precision Electronic Navigation Map "
2018/11/21
12
4.3 Roadmap
:���������Roadmap of Intelligent high-precision maps � “The intelligent vehicle innovation and
development strategy” released by the national
development and reform commission of China
specifies the national intelligent vehicle
innovation and development strategy: the
framework of road network facilities,
regulations and standards for standard
intelligent vehicles will be basically formed by
2020, and be fully formed by 2025, by that time ,
"human-vehicle-road-cloud" could be highly
coordinated.
� The European Union plans to develop
intelligent transport from the two aspects of
Internet road environmental resource
aggregation and intelligent network vehicles,
and aims to achieve that goal by 2050.
� Japan conducted a static data survey of autono
mous driving in 2015; Conduct dynamic data re
search in 2016, and establish dynamic map pla-
tform to discuss dynamic map data use cases an
d element contents. Japan will achieve full cove
rage of high-precision maps by 2020
�� !Types �� " ���� ���� ���# ���$ ���! ��%� ��&&
Intelligent
high-
precision
maps
Dynamic
high-
precision
maps
Static
high-
precisio
n maps
� Establish basic map system for vehicles
covering whole country, complete the
standard and uniform basic maps for
vehicle.
� High precision maps realize autonomous
intelligent control, with self-adaptive, self-
evaluation and self-learning
� Optimize global control of the entire
network, provide comfortable, convenient, ,
green, civilized trip and dynamic, real-time,
intelligent, ubiquitous service
ÝYZg\]Yk\\'
ÝYZg\aX\'
ÝYZg\]qa '
� Complete the static data survey and
formulation of related standard
� Establish data acquisition system with high
precision and update ecological chain
� Carry out dynamic data survey and determine
types, contents, and definitions of dynamic map
data
()*(+
(,*(-
� High precision map related standard syst-
em will be formed in an all-round way
� ”human-vehicle-road-cloud
” will achieve highly collaborative
� Wireless communication network for
vehicles will basically meet the demand
(.
Contents
1. Background and demands
2. Compared with traditional map
3. Main elements and layers
4. Outlook of technologies and standards
5. Summary and thoughts
2018/11/21
13
� �v~N 5J/ 09D0=I/9=5C;/ 09wJ5=8J3K83J/= w2J 098/;;07/98 K299/K8/D 6/B0K;/=U QB/0J D585C5=/ K59
C/ /=85C;0=B/D 59D 3ID58/D D<95>0K5;;<G 5KK2JD097 82 D0ww/J/98 D/>59D= 59D J3;/=
� �9 �v~NG ;59/ ;09/=G F1 K22JD0958/= 2w 8J5ww0K =079= 59D J/;58/D I5J5>/8/J= ?=3KB 5=G 83J9097
J5D03= G 7J5D0/98@G B56/ DJ06097 K298J2; K5I5C0;08<U QB/< 5J/ 8B/ 3;80>58/ K298J2; C5=0= HB0;/
8B/ /960J29>/985; I/JK/I8029 =<=8/> =3KB 5= 60=029 2J J5D5J w50;=
2018/11/21
14
� :859D5JD= w2J �v~N 5J/ DJ06/9 C< 8/KB92;270/= 59D D/>59D=U :09K/ =5w/8< J/�30J/>/98=
5J/ /è8J/>/;< D/>59D097G >59< =859D5JD= 9//D 82 C/ =/8 =0>3;859/23=;< H08B ;5H=
� AJ2HD=23JK097 H0;; C/ 5 9/K/==5J< D585 =23JK/ 82 /9=3J/ J/5;E
80>/ 09 �v~NU D9=3J097 D585 5650;5C0;08<G J/;05C0;08<G J/5D5C0;08<G D/8/J>09097 D585 65;0D08<
59D w0;8/J097G 59D 8B/ D585 098/J5K8029 IJ02J08< 5= H/;; 5= D585 D/;5< J3;/= 5J/ Ç/< I2098= 59D
>5ç2J D0ww0K3;80/=
2018/11/21
15
� .5=/D 29 /960J29>/985; I/JK/I8029 59D >5I >58KB097 8/KB92;270/=G �v~N B56/ 8B/
5C0;08< 82 K298J2; L38292>23= 1J06097 H08B =I/K0w0K58029= =3KB 5= 8J5ww0K J3;/= 59D >5I
D585 K29=8J5098=G HB0KB 8J5D080295; >5I= K59 928 D2
� �v~N 5J/ B/;Iw3; 82 5KB0/6/ /ww0K0/98 >2C0;08< H08B B07B �35;08<G 8B3= 8B/ 60=029 2w
EN2C0;08<E5=E5E=/J60K/FK59 C/ J/5;0//D
2018/11/21
16
4.1 Several technical problems that need to be solved
1585
K2;;/K8029
3�Ù �� ��Ð×Ñ� Ö��Ö Ñ�Ï�Ï �Í ���Í�ÏÏ����ÓÑ�ÓÓ�Ñ���� �ÑÑ��Ð��� �� Ö��Ö Ð��� ×�Ð�������Ï4
N5I
IJ2D3K8029
3�Ù �� ��Ð×Ñ� Ñ�Ï�Ï �Í Ñ�Î�×������Ï�×�Ñ� ×�Ð�� �Ö� Ñ��Ð����� �Í Õ��Ð���5����� ÑÓ�×Ðß �Î���ß Ó�Ñ�����ß ��Ñ67 4
1585
098/J5K8029
3�Ù �� Õ×�ÓÐ ���ÓÚÙ��ÓÐ �������Î���ÏÙ��Ö ÐÔ��Î�Ñ ���ÓÚ��Î� Ð��� 4
A2>I38097
I588/J9
3�Ù �� ����Ð Ð��� Õ�Ñ�Ó�� Ð×��� Ó�Î���Ð Ñ�Î�×���� ��Ù�� �Í �Ö� ÑÓ�×Ð�Ó��Í��Î4
LII;0K58029=
3�Ù �� ����Ð ÐÔ��Î�Ñ �ÕÏ��ÑÓ�Ï à×�Ñ�ÓÔ4��Ð Ö�Ù �� ���Î��� �Ñ���� Ï�Ñ×���Ô��Í��Î����� Í��Î ÑÓ�×Ð �Ó��Í��Î ���Ð���Ñ� ��Ð Ö�Ù Ñ�� ×Ï��Ï �Õ���� ��4
�98/;;07/98
K298J2;
3�Ù �� �Î�Ó�Î��� ���ÓÚ��Î� ��Ð Ö��Ö���Ñ�Ï��� ���Ñ������ Í�� ���ÍÍ�Ñ Ï���Ï �Ö����� ×Ï�Ð �� ��Ð �×����Î�×Ï ����ÓÓ�����Ð������4
ÜX]\mkZ]_aX rkat\gg_aXZn [ann\[]_aX q_]Y[kaqlgabk[_Xm
³k]_t_[_Zn _X]\nn_m\X[\ _g Z X\q ZrrkaZ[Y
[annZjakZ]_s\ ZqZk\X\gg at ËZ[Y_X\
s_g_aX ZXl 89;
[aorka[\gg_Xm j\]q\\X �nabl ZXl
]\ki_XZn \Xlg
Zgg_g]_s\ \Xs_kaXi\X]Zn r\k[\r]_aXcrag_]_aX_XmcrnZXX_Xmc[aX]kan
³b]aXaiabg _X]\nn_m\X] [aX]kan q_]Yg\nton\ZkX_Xm� g\ntoZlZr]Z]_aX ZXl g\nto\sZnbZ]_aX