高速公路施工区车辆强制换道耗时生存模型

摘要

To explore the mandatory lane-changing ( MLC ) behavior of vehicles in work zones on freeways and factors which influence vehicles' MLC behavior, semi-parametric method is used to establish the multiplicative hazard model of mandatory lane-changing duration. The traffic data on the duration of the lane-changing and factors of the vehicles in freeway maintenance construction area were collected by an Unmanned Aerial Vehicle, and lane-changing data has been analyzed by Cox regression analysis. The results show that nearly 77% of the lane-changing duration is in 10 s. The effect of different vehicle types on the duration of mandatory lane-changing is not significant. The cumulative survival rate of off-peak period was significantly lower than that of peak period and transition period, and the cumulative survival rate of the peak period is the highest. The influence of lane changing vehicle types and traffic time periods on MLC for work zones on freeways can be conducted effectively using the survival model proposed. The model proposed and results from analysis can provide a theoretical basis for traffic management and control of work zones, and the modeling and simulation of freeway lane-changing behavior in high-speed and dynamic traffic environment.%为研究高速公路施工区车辆强制换道行为及其影响因素,采用生存分析中的半参数分析方法建立强制换道耗时的乘法风险率模型.通过无人机拍摄采集高速公路施工区的车辆换道耗时及其影响因素数据,最终确立换道耗时Cox比例风险模型,对换道耗时数据进行Cox回归建模分析.结果表明:近77%的换道车辆在10 s内完成换道;小型车和中型车经养护施工区的换道耗时未发现显著性差别;对于相同的换道耗时,平峰期的累积生存率明显低于高峰期和过渡时期,而高峰期的累积生存率最高.建立的强制换道耗时生存模型可有效的定量分析车型和交通时段对高速公路施工区车辆换道行为的影响,可为高速公路施工区交通管理控制及车辆换道行为建模及仿真奠定一定的理论基础.