基于贝叶斯网的道路事故机理分析

Traffic Accident Mechanism Analysis Based on Bayesian Network

在累计频率法原理的基础上,运用excel和数学软件MATLAB分析交通事故数据,进行道路交通事故黑点鉴别,并绘制出事故频率曲线。确定事故黑点位置后,应用贝叶斯网理论全面整理、分析事故数据,并结合贝叶斯网推理进行事故机理分析,利用贝叶斯网定性和定量分析相结合的特点,为改善黑点交通安全提供参考和指导。

This paper firstly use Matlab to identify the black-spot on Jiaxing segment of Shanghai-Hangzhou Freeway, based on the principles of cumulative frequency method. Afterwards adaptability of using Bayesian network in accident analysis is argued and fundamental knowledge of it is introduced. Finally, software Netica is employed to establish the topological structure of Bayesian network and sensitivity analysis and accident diagnosis are done with the help of the software as well as the network model, thus the mechanism of accidents is analyzed.