Research On The Identification Methods Of Urban Road Traffic Accident Black-spots

There are many urban road traffic accidents every year. The number of accidents accounted for 42.39% of the total number, and the number of injuries accounted for 40.97% of the total only in 2013. In recent years, the research of urban traffic accidents is mostly based on statistical analysis with little considering some factors such as roads and environment affecting accidents. Therefore, the key point of studying urban traffic accidents is to find the root causation and accident-prone locations based on the combination of specific road conditions and accident cases, which is very important to reduce traffic accidents and the severity of accidents.Firstly, on the basis of analysis and comparison of existing accident black spot identification methods, the accident-prone points were studied from macro and micro aspects. For finding higher risk roads, all the traffic accidents from urban road network were analyzed with a minimum sum of squares clustering method. For finding the accident-prone points in a specific road, DBSCAN clustering algorithm was used to analyze the accidents in this road combined with cumulative frequency curve method.Secondly, the current traffic situation in Xi’an was analyzed from four aspects such as human geography, road network transportation and traffic accidents. Furthermore, the influencing factors of urban traffic accidents in Xi’an were analyzed by using FAHP from some aspects such as people, vechiles, road, environment and regulations. Then the hierarchy model of accident factors was established. In the end, we obtain the importance weights of each factor.Finally, a road traffic accident database was established on the basis of collecting accident cases in 2014 in Xi’an. All the traffic accidents from urban road network in Xi’an were analyzed with a minimum sum of squares clustering method, and we obtained the higher risk roads. For a specific road, Fuyu Road was taken for example. Cumulative frequency curve method combined with DBSCAN clustering algorithm was used to find accident-prone points in Fuyu Road...