高速公路隧道交通事故严重程度的影响因素分析

根据京珠高速公路韶关段4个隧道的交通事故资料，从时间因素、隧道环境因素和交通动态因素3个方面选取9个输入变量，以交通事故严重程度为输出变量，建立高速公路隧道交通事故严重程度预测模型；然后，通过灵敏度分析方法，研究各个输入变量对输出变量的影响程度，并对各个输入变量的灵敏度分析结果进行比较分析．研究结果表明，日交通量与年平均日交通量之比和大型车混入率对交通事故严重程度的影响最大，天气、线形、坡度和事故发生地点在隧道中的位置对交通事故严重程度的影响基本相等，事故发生时段对交通事故严重程度的影响可以忽略不计．

Analysis of Influence Factors on Severity for Traffic Accidents of Expressway Tunnel

According to the accident information of Shaoguan tunnels for Beijing-Zhuhai expressway in China, a neural network model was constructed for the purpose of predicting the severity of accidents. In this model, 9 input variables were selected from three aspects, which are the time of traffic accident, tunnel environment and traffic dynamic factors, and the output variable was the severity of accident. Then the sensitivity analysis method was selected to study the effects of input variables on output variable. Three conclusions can be obtained. Firstly, the most contribution to the severity of accidents are the ratio of daily traffic volume and the AADT, and the proportion of large vehicles. Secondly, four input variables, which are weather, alignment, grade and accident location, have equal contribution to the severity of accidents. Thirdly, it is negligible that the time of accidents happened contributes to the severity of accidents.