Research And Development Of Highway Dangerous Section Identification And Analysis System Based On ArcGIS

With the gradual growth of vehicles and drivers, traffic volume assumes the geometric series growth in recent years. And a worsening road safety situation follows gradually. Because of the high speed, large flow and relatively high level of service, highway traffic safety problems are more serious. The research and development of identification and analysis system for highway dangerous sections are the key link of the study on road safety. In this context, the key problem to prevent accidents is to excavate the dangerous sections effectively and find the potential causes of traffic accidents using advanced information technology.Geographical information system is the main technique to deal with issues about spatial location. Due to the obvious spatial location features of traffic accidents, we utilize the best GIS software—ArcGIS to develop the identification and analysis system for highway dangerous sections. According to the safe grade ratings criteria in the notice of a comprehen-sive investigation for highway dangerous sections issued by the national conference about the work of road traffic safety, this dissertation attempts to establish an identification and analysis system having the spatial visualization interface and excellent spatial analysis capabilities.First of all, this article introduces the background of accidents safety management and the domestic and foreign research status about analysis methods and systems of traffic accidents. Then it elaborates and analyzes the conceptions, functions and characteristics of ArcGIS software architecture. On this basis, it analyzes the two key issues including finding the accident-prone points and identifying the cause of accidents. In the light of the core issues, this dissertation designs two functional modules, identification module of dangerous sections and analysis module of association rules. To realize the safe grade rating function under different standard of dividing sections, it applies the classical spatial clustering algorithm—DBSCAN and designs an identification algorithm for highway dangerous sections. Then it introduces association rule and spatial data mining and study the process and improvement of classical Apriori algorithm. Therefore, to achieve accident association analysis function, there is a multidimensional association algorithm directed against the traits of traffic accident attribute proposed. Finally, combining with ArcGIS Engine component package, this dissertation establishes an integral identification and analysis system using the above two algorithms. Through importing example data, it demonstrates the effectiveness of these design methods mentioned.