大型营运客车在高速公路的换道安全性辨识

为了给大型营运客车换道预警系统设计提供参考，采用毫米波雷达、激光雷达、车道线识别传感器、GPS、视频监控系统以及控制器局域网（CAN）总线数据采集仪等设备，基于小型乘用车搭建浮动车采集平台。通过在试验线路上进行1.5×10⁴ km的驾驶试验，获取1 200余次营运客车的真实换道数据。以Jula提出的换道安全性模型为基础，结合营运客车的换道行为特征，通过分析换道进程结束后客车需要与周围车辆保持的安全距离，建立适合于营运客车的3类换道安全性识别模型（客车与自车道前方车辆、目标车道前方车辆、目标车道后方车辆），并利用真实数据对3类模型进行验证。研究结果表明：客车换道持续时间均值为10.4 s，换道起始时刻与目标车道后方车辆的距离为10.0~40.0 m；所有换道样本中，73.3%的换道过程中客车速度要高于目标车道后方车辆，且超过90%的换道过程是由前方慢车引起；不同的速度区间下，车速和航向角联合变化情况下，驾驶人控制营运客车的横向偏移速度保持稳定，可认为客车驾驶人的心理预期换道进程存在固定经验模式，这与小型车换道的研究结论存在较大差异，传统的TTC预警算法识别率较低，在不同速度区间情况下，所提出的模型对客车与自车道前方车辆、目标车道前方车辆、目标车道后方车辆的换道安全识别评价准确率均超过了90%。

Identification of Lane Change Safety Characteristic of Large Commercial Bus on Expressway

In order to provide a reference for designing lane change warning systems of commercial buses, a floating car system with millimeter wave radar, laser radar, a lane mark detection sensor, GPS, a video monitoring system, and a CAN-bus data capture device was developed based on a small passenger vehicle. More than 1 200 lane changes were acquired from a 15 000 km real road driving test. Analysis results of the lane changes show that the average lane change duration time was 10.4 s. At the lane change start time, the distance distribution between the commercial bus and the rear vehicle in the target lane was 10.0-40.0 m. In more than 73.3% of the lane change samples, the velocity of the commercial bus was higher than the rear vehicle in the target lane, and about 90% of lane changes were caused by slow vehicles in front of the bus. In different velocity situations, while the velocity and heading angle changed separately, the lateral deviation velocity of the commercial bus remained stable. It can be considered that the psychological expectations of the lane change process for the commercial bus driver existed in a fixed empirical mode, and this result was different with that of small vehicles. Based on the lane change safety model presented by Jula, considering the lane change characteristics of the commercial bus, lane change safety models of a commercial bus between different surrounding vehicles were established separately. The surrounding vehicles include a vehicle ahead in the same lane, a vehicle ahead in the target lane, and a rear vehicle in the target lane. Real data was used to identify the three model types, and analysis shows that while the identification results of the traditional TTC warning algorithm were poor, the recognition rates of the models for a vehicle in the same lane, a vehicle ahead in the target lane, and a rear vehicle in the target lane all exceeded 90%.