DEVELOPING ACCIDENT PREDICTION MODEL OF TWO-LANE HIGHWAY ON THE PLATEAU
Based on the analysis of the traffic flow,in this research,roadside environment and traffic safety data of Qinghai-Tibet highway,the characteristics of road traffic and the influence of altitude,oxygen content,traffic flow and environment on the traffic accidents were studied.A non-rural road accident prediction model based on zero-inflation Poisson and an accident prediction model based on negative binomial distribution were established.In the paper,the effect of oxygen content,the proportion of large vehicles and the direction distribution coefficient of the traffic flow on the number of accidents in the rural roads were studied.The results indicate that there are significant differences between the characteristics of rural road accidents and non-rural road accidents.Roadside hazard level and the access density respectively have the most significant influence on the number of non-rural road and rural road accidents.
traffic engineering Qinghai-Tibet Plateau two-lane highway zero-inflation Poisson distribution negative binomial distribution accident prediction model
Bingbing Feng Yulong He Guixian Qu Mingyang Jia
Beijing University of Technology,Beijing Key Laboratory of traffic engineering,Beijing 100124
国际会议
2016 International Symposium on Civil and Environmental Engineering(2016年土木与环境工程国际研讨会)
武汉
英文
1-6
2016-12-20(万方平台首次上网日期,不代表论文的发表时间)