Using a Statistical Method for the Determination of Spatial Systematic Reasons in Urban Accidents

Among the various reasons of accidents, some of them are under the influence of the outside environment of vehicle. The accident site is an important parameter in determining and resolving these factors. The same effect of environmental factors on different vehicles, causing spatial systematic behavior in accidents. Spatial point of view, the systematic behavior of point events such as accidents tend to cluster formation is shown. The use of cluster detection methods and spatial statistics such as K-function will be provided a useful tool for identifying accident-prone areas with a peripheral origin. Since the road accidents occur on the network, instead of using the K-function statistics in Euclidean space, of statistics based network K-function is used. Based on Monte-Carlo simulation, distribution of accidents on the three traffic areas are obtained and compared with the actual sample. The results show a tendency to cluster and systematic behavior in the areas of environmental factors such as the geometric imperfection of road.