Study On Road Traffic Accident Prediction Of Medium-sized Cities

This article relies on Road Traffic Safety Planning Project of Medium-sized Cities, it analysis the current situation of Chinese road safety features and background profoundly. With the improvement of the socio-economic development and the rapid demand for people of travel and travel quality requirements, our vehicle’s quantity is growing rapidly. Rapid increasing of vehicles and drivers brings new challenges to the originally fragile traffic safety. The total number of road accidents has been always high, it brings serious harm to the lives and propert of people in our country. Therefore, it has become a top priority of road traffic management that how to prevent road traffic accidents and improve road traffic safety environment effectively.Therefore, this paper is based on the planning of the project, studying the urban road traffic accident prediction as an example of a medium-sized city, which purpose is to predict the future trend of road traffic accidents and to provide theoretical support for traffic management departments to develop prevention strategies and public awareness and education work. Predicting the recent year and future state of road accidents can provide a clear situation for the development of traffic safety administration and help the government develop appropriate social development goals, development strategy. Through analyzing the basic situation of the city social development, the basic conditions of road traffic development and road traffic accident statistics, it will get the factors related to the four traffic elements and summarizes the characteristics of the accident led to the accident to identify where the the key points is.Combined with qualitative analysis above, the paper choose three theoretical methods, which are a multiple linear regression theory, gray model theory and multi-hierarchical method, to establish the appropriate road traffic accident prediction model based on the medium-sized city road traffic accidents statistical data, then predict the road traffic accidents numbers of this city and make comprehensively comparative analysis of the degree of curve fitting, forecast error, advantages and disadvantages and scope of each model. Finally it will get a conclusion that multi-hierarchical method with standardized impact factors has high prediction accuracy in a long period future years forecasting of traffic accidents, and it applies to the traffic accident prediction work of the urban road.