A GIS SYSTEM FOR MANAGING AND ANALYZING URBAN ROAD TRAFFIC ACCIDENT

Master Student
留學生畢業(yè)dissertationDept. of Civil Engineering
Dalian University of Technology,
Dept. of Architecture

Abstract: We develop a GIS database to store urban traffic accident data in digital style, andanalyze these data in terms of their attribute and spatial characteristics to exploit the reasons of
accidents and rank urban area with dangerous degrees. We establish a road network in MapInfofor a sample city. Second we collect three years’ death traffic accidents data of the city and linkthese data with road network database with OLE connection between MapInfo and Visual Basic.We analyze reasons of traffic accidents to summarize major factors related to the accidents.Finally we create a hazard map to show the accident dangerous degree in the city.

Keywords: Traffic Accident, Hazard Map, MapInfo, Road Network and OLE

1. INTRODUCTION
As the soaring increment of car ownership in China’s cities and the poor traffic control system,road traffic accidents have been increasing rapidly and induced lots of economic lose. For
example, in year 2000 traffic accident per 10 thousand motor vehicles of Dalian city is 147.0,while 21.7% of them are death accidents. In order to handle traffic accidents, lots of informationhas been collected and large amount of documents are stored in police offices. However, worksof systematic analyses of the data are hardly done. Almost all studies on traffic accident in Chinaconcentrated on theoretical and conceptual analyses. For example, Mao (Mao et al, 2002) gavesome concept and structure to explain the causation of the accident, but no data or actual caseswere used in the analysis. Based on system theory Zheng (Zheng, 2002) used macro data,namely statistic traffic accident data on motorway, to analyze reasons of the accident in terms ofdriver, pedestrian, vehicle road condition and traffic management. Since actual urban traffic datahave not been analyzed based on city configuration, traffic and road condition and driver orpedestrian. Concrete reasons of the accident are not understood and effective counter-measurefor decreasing traffic accidents have not been found. One of the main reasons is that almost allconcerned documents are compiled and stored in analogue style. It is very difficult to search andProceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003analyze accident information systematically. Some studies in Japan (SHIRAISHI et al, 2000,SUGIURA et al, 2000) took city as study area and use GIS to analyze the urban traffic accidentsand Danger Perception, while they just used GIS as a platform for storing data and somespatialcalculation.
Here we develops a GIS system to store urban traffic accident data in digital style, and analyzethese data in terms of their spatial and attribute characteristics to exploit the reasons of accidents#p#分頁標題#e#
and rank urban area with dangerous degrees. Taking a China’s middle size city (Panjin City) asthe case study area, the system uses MapInfo to store urban road network, which contains names,
widths, design speeds and levels of service of road and names of intersections. Traffic accidentsinformation in the past 2.5 years is created as texture and photographic electric files. And link
between road network database and traffic accident file is established with OLE connectionbetween MapInfo and Visual Basic. Several command buttons and interfaces, which are used forsearching and displaying accident information, are created with Visual Basic and are installedinto MapInfo system. With them, users can easily find accident information in GIS windows
through some query method. We further analyze the reasons of death traffic accidents in detailsto summarize the major reasons related to them. With above results we rank major urban area
with accident dangerous degree. The rank results are created as thematic map in GIS, and thenaccident hazard map is displayed in GIS windows.
2. TRAFFIC ACCIDENT DATABASE
Traffic accidents data such as statement and pictures are often stored in straight sheet style. It isdifficult for user to carry out management, maintenance and query. Therefore, management,
query and analysis of traffic accidents data will be more convenience and efficiency if wecombine them with road traffic database in GIS. Fig. 1 illustrates the structure of our road
network database. In the system road and traffic information are store in MapInfo style, whileMapBasic is used to handle attribute and graphic data in MapInfo and MS-VB is used to createthe interfaces.
Even though MapBasic is powerful on operating objects in MapInfo, it cannot deal with externadatabase, while MS-VB can easily operate database through standard database channel (Yang et
al, 2002). Therefore, in traffic accident database we link MapInfo and MS-VB through OLEmethod. As Fi.2 shows road network database and Access tables of traffic accident are the main
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003foundation of traffic accident data management system. Based on the network and accident
MS-Access tables, MapBasic manage graphic objects and tables in MapInfo in the backstage totransfer the data to MS-VB through global variables, while MS-VB responds communicate withexternal database to display and manage the external data, more over MS-VB is also used toshow pictures. On the other hand, MS-VB can also transfer data through global variables from
external data to MapInfo. Here traffic accidents data are stored in two MS-Access tables, whichstore traffic accident describing text and traffic accident pictures respectively. With unique IDcode the two tables can be linked.
Fig. 1 Interface of Road Crossing and Section Information Database
Fig. 2 Interface of Traffic Accident Database
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003#p#分頁標題#e#
3. ANALYZING THE REASONS OF THE DEATH TRAFFIC ACCIDENTSFrom traffic accident database we can pick up the death cases and read their situation and look inthe pictures. There are totally 81 death accidents in the database, while 60% happened onroadway sections and 40% happened within intersections. This is some different from generalrealization of that intersection is the most potential site for traffic accident. Fig. 3 shown that thedeath traffic accidents happened mostly between motor vehicle and bike rider and pedestrian,this kinds ofaccidents account for 35% and 24% respectively. We sum up the main reasons ofthese accidents to get Table 1.
Table 1 Main Reason for the Accidents
Party Concerned Road Intersection
Motor-No Motor No motor vehicles occupied motor lane Left-turn of no motor vehicle
Motor-Motor Overtaking Left-turn of motor vehicle
Motor-Pedestrian Pedestrian cross road on no zebra stripes
Pedestrian cross road on no
zebra stripes
All Accidents
Drivers and pedestrian without safety coconscious, neglect traffic
regulation, bad traffic management, no license drivers
We can see that the main reason of the accidents are drivers or pedestrian neglect the safetyregulation and violate traffic regulation. Till now motorization in China’s kept at a low level,persons there did not know the importance of traffic regulation and danger of the motor vehicle.Therefore, pedestrians often cross roads wantonly or during red single phase, and bicycles (nomotor vehicles) often occupy motor vehicle lanes and neglect single. There are also some
Motor-Motor
Motor-NoMotor
Motor-Object
Motor -Pedestrian 35%
18%
24%
8%
Fig. 3 Proportion of Bodies concerned in Death Traffic Accidents
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003problems in traffic control system, especially for the accident in the case of left-turn situation.Accident between motor vehicles happened in left-turn are mainly due to the shortage of left-turnlane or left-turn phase. Accidents between motor vehicle and bicycle happened in left-turnsituation aremainly due to theshortage of mark for left turn vehicle to waiting in the intersectionor regulation for bicycle to left-turn in two phases.
4. ANALYZING THE SPATIAL DISTRIBUTION OF THE DEATH TRAFFIC
ACCIDENTS
Since we establish a database to manage the accident information, it is very easy to analyze the
spatial distribution of the accidents. We also pick up death traffic accidents from the database
and plot them in the road network layer. With traffic zone, road network, population and plotted
accidents in GIS we made some spatial calculations. Fig. 4 and Fig. 5 illustrate road and
intersection density based on traffic analysis zones respectively, while Fig. 6 and Fig. 7 shows
the population density and death traffic accidents quantity with the same spatial units
respectively. We can understand that spatial distribution of death accident does not coincide with#p#分頁標題#e#
road density, intersection density and population density except in CBD area.
Fig.4 Road Density (Road Area/Zone Area) Fig.5 Intersection Density (Node Number/Zone Area)
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003
In order to explain the characteristics of the spatial distribution of the death traffic accidents, we
first describe the spatial configuration of the city. Panjin city is about 14 km long in south-north
and 9 km wide in east-west. A river flows through the city in west-east direction and divides it
into two districts. There are three artery roads connecting the two districts and traffic between
districts is heavy. Moreover, main entrances of the city are in north and south tips. As the results,
through traffic also goes in south-north direction. From Fig. 8 we can see that death traffic
accidents first concentrated along the three north-south arteries, especially the south and north
urban entrance areas. Another accident area is in the urban center, since there are dens road
network and large traffic and pedestrian flows, and roads intersect each other, the central zone
became a main traffic accident area. Interesting thing is that area around the arteries and a little
far from the central zone is also a dangerous area. This is because that the middle area between
south and north districts has a little residential and commercial infrastructures, then traffic and
pedestrian there get less. Most traffic here are shuttled vehicle between the districts, especially
traffic from south to north will accelerate there since they just escaped from a heavier traffic
zone and face a good road and traffic conditions. However, signal lamps and zebra stripes in this
area are less than in central area, therefore, death traffic accidents often happened in this area.
Fig. 6 Population Density (Population/Zone Area) Fig. 7 Spatial Distribution of Death Accident
5. CONCLUSIONS
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003
Taking a China’s city as study object we developed a GIS system to store and manage its traffic
accident data. In the system we can query, analyze and look in the accident information with GUI
method. We further analyze and sum up the reasons and situations of the death traffic accidents
to find that neglecting of traffic regulation, short of safety consciousness and bad traffic control
method are the main reasons of the accident, while bike rider and pedestrians are most involved
in the accidents. We finally analyze the spatial distribution of the death traffic accidents and find
black point of the accident and make a traffic accident hazard map.
Fig. 8 Spatial Distribution of Death Traffic Accident in the City
REFERENCES
MAO Min & YU Xiang (2002), Analysis of Traffic Accident Causation, Journal of highway#p#分頁標題#e#
and transportation research and development, Vol.19, No.5, 125-127.
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003
SHIRAISHI H. et al (2000). Study On Relationship Between Traffic Accident Location And
Perception Of Hazardous Location By Considering The Type Of Road, Proceeding Of
Infrastructure Planning Of JSCE, No.23(1), 727-730
SUGIURA S. & IZUHARA K. (2000). Analysis Of Traffic Accidents and Danger Perception – A
Case Study Elementary School District Of Toyota City, Proceeding Of Infrastructure Planning
Of JSCE, No.23(1), 731-734.
Yang Zhongzhen & Sun Zhuo (2002). Urban Road Traffic management System based on GIS,
留學生dissertationComputer and Communications, Vol.20, No.5, 18-21
ZHENG An-wen (2002), Basic Characteristic And Prevention Measures Of Motorway Traffic
Accident In China, Journal of highway and transportation research and development, Vol.19,
No.4, 110-112.
Proceedings of the Eastern Asia Society for Transportation Studies, Vol.4, October, 2003