基于CA模型的交通事故叠加风险量化分析方法

A Quantitative Analysis Method of Traffic Crash Risk Based on Cellular Automata Model

考虑多种风险因素的共同作用,探讨了多风险因素叠加作用下的交通事故风险量化方法。在交通事故风险因素分类和基于事故统计数据的交通事故风险量化分析的基础上,将分心行为(手机通话)、违法行车(超载)、不利环境(雾霾天气)3种风险因素加入到基于NS模型规则的元胞自动机(CA)模型中,通过设定危险情境,量化了各风险因素组合下的交通事故风险,然后利用多因素方差分析方法计算了多风险因素下各风险因素组合的叠加作用所产生的交通事故风险。仿真结果表明,驾驶员手机通话时的事故风险是没有手机通话时的14倍,同时,当雾霾天气因素与手机通话因素同时作用时,出现了显著的叠加作用,产生的叠加风险约为雾霾天气单独作用所产生风险的2倍,表明了多风险因素下交通事故风险具有非线性叠加性。

A quantitative method of traffic crash risk under superposition effects of multiple risk factors is studied.On the basis of classification of traffic accident risk factors and quantitative analysis of traffic accident risk based on accident statistics,three risk factors,including distraction(use mobile phone),illegal driving(vehicle overload),and adverse weather(fog and haze weather),are added to a cellular automata traffic flow model based on the Nagel-Schreckenberg model.Then,by adjusting factors of dangerous situations,the risk of traffic crash under various combinations of risk factors is quantified,and a multi-factor analysis of variance is used to calculate superposition effects.The simulation result shows that when drivers dealing with phone calls,the risk level is fourteen times of the normal level.When drivers are using mobile phone under fog and haze weather,there is a significant superposition effect,which is about twice the risk of fog and haze weather alone.It indicates that traffic accident risk has non-linear superposition under multiple risk factors.