基于ACC系统的目标车辆换道与出入弯道状态辨识算法

An Algorithm for State Identification of Lane Change and Curve Entry/Exit of Target Vehicle Based on ACC System

自适应巡航控制(ACC)系统利用雷达对前方目标进行追踪,当前车进行换道或进出弯道时,ACC系统无法区分这两种状态,容易引发交通冲突。针对此问题,本文中通过实际驾驶试验,获取了前车不同运动状态的数据,采用道路曲率估算值、前车行驶轨迹的斜率及其变化率和前车与自车之间横向距离作为表征参数,结合车-路协同运动特征,建立了前车换道与进出弯道的识别模型,并利用实测数据对模型的有效性进行了验证。结果表明:当自车处于直道时,对前车换道和进入弯道的识别率分别达到91.46%和89.81%;当自车处于弯道时,对前车换道和驶出弯道的识别率分别达到87.06%和90.42%。

Adaptive cruise control (ACC) system uses radar to track preceding target,but it is unable to distinguish whether the preceding vehicle is doing lane change or entering/exiting curved road,easily leading to traffic conflict.Aiming at this problem,the data of preceding vehicle in different running states are collected by driving test.The estimated values of road curvature,the slope and its changing rate of preceding vehicle trajectory,and the lateral distance between host vehicle and preceding vehicle are taken as characteristic parameters.With consideration of the cooperative characteristics between vehicle and road,an identification model for lane change and curve entry/exit of preceding vehicle is established with its effectiveness verified by road test data.The results show that when the host vehicle is on a straight road,the recognition rate of lane change and curve entry of preceding vehicle reaches 91.46% and 89.81% respectively;and when the host vehicle is on a curved road,the recognition rate of lane change and curve exit of preceding vehicle reaches 87.06% and 90.42% respectively.