大跨桥梁全样本数据交通安全评价

Safety Evaluation of Long Scan Bridge in the Estuary Region with Full Sample Data

随着恶劣多变的天气和日益增加的交通量,大跨桥梁往往成为交通事故的重灾区。为了提高大跨桥梁的整体服务水平,更加科学的评价大桥交通安全情况,基于全样本数据分析的方法对桥梁数据进行分析,对面板数据时间自相关效应进行检验,同时使用普通最大似然法和惩罚最大似然法(firth方法)对模型进行参数估计,并将计算结果进行对比分析。研究结果表明,湿度、平均风速、小时流量、季节、时段是影响大桥交通事故概率的显著因素。

With the bad and changeable weather and the increasing traffic volume, the large span bridge often becomes the disaster area of traffic accident. In order to improve the overall service level of large span bridge, the traffic safety of Bridge is evaluated more scientifically. Based on the method of full sample data analysis, the bridge data is analyzed, the time autocorrelation effect of panel data is tested, and the parameter estimation of the model is carried out by using the common maximum likelihood method and the penalty maximum likelihood method (Firth method), and the calculation results are compared and analyzed. The results show that humidity, average wind speed, hourly flow rate, season and time period are the significant factors affecting the probability of traffic accidents on the bridge.