相邻车道车辆并道预测与跟车控制改进的研究  

A Study on the Merging Prediction of the Vehicle on Adjacent Lane and the Improvement of Vehicle-following Control

作　　者：付锐[1,2] 袁小慧 郭应时[1] 马勇[1] 

机构地区：[1]长安大学汽车学院,西安710064 [2]长安大学,汽车运输安全保障技术交通行业重点实验室,西安710064

出　　处：《汽车工程》2017年第10期1106-1113,共8页Automotive Engineering

基　　金：教育部创新团队发展计划项目(IRT_17R95);国家自然科学基金(61473046,61374196);中央高校基本科研业务费专项资金项目(310822153101,310822161006)资助

摘　　要：为在自适应跟车控制中,预测相邻车道车辆的并道行为,提前调整自车速度控制策略,以提高自车的乘坐舒适性和安全性,通过采集和分析实际道路上自车与周围车辆的运动状态信息,建立相邻车道车辆并道行为预测模型,并在此基础上对现有期望间距跟车模型进行改进。仿真结果表明:自车车速vM、自车与前车的车头时距t1、自车与前车的相对速度vML和自车与相邻车道车辆的相对速度vMF可作为预测并道行为的表征参数;所建立的预测模型可提前2s预测并道行为的类型,对两种并道行为预测的准确率分别达到88%和90%;改进的期望间距模型使跟车过程更加平稳,可提高车辆的舒适性与安全性。For predicting the merging manoeuvre of the vehicle on adjacent lane,and adjusting the control strategy of ego-vehicle speed in advance to enhance the ride comfort and safety of ego-vehicle in adaptive car- follow-ing control,a predictive model for the merging manoeuvre of the vehicle on adjacent lane is established collecting and analyzing the real world motion state information of ego-vehicle and surrounding vehicles,and based on which the existing desired vehicle spacing model for vehicle following is improved. The results of simulation show that four variables can be regarded as characteristics parameters to predict merging behavior. They are the speed of ego- vehi- cle,the time headway and relative speed between ego-vehicle and preceding vehicle and the relative speed between ego-vehicle and the vehicle on adjacent lane. The predictive model can predict the type of merging behavior 2 sec-onds in advance,and the prediction accuracy rates for two merging types reach 88% and 90 % respectively. The im-proved desired vehicle spacing model makes vehicle following process more stable,with the ride comfort and safety of vehicle enhanced.

关 键 词：跟车 并道预测 期望间距 Fisher判别法 

分 类 号：U463.6[机械工程—车辆工程]