摘要
针对现有汽车自适应巡航控制系统(adaptive cruise control,ACC)在弯道上经常出现的追踪目标丢失问题,利用微波雷达、三轴陀螺仪、车道线识别系统等传感器构建了自然驾驶行为试验车。在高速公路、国道等道路下进行了多位驾驶人的实际道路自然驾驶试验,提取了自然跟车过程中的自车运动状态数据以及道路交通环境数据。采用车速与车身横摆角速度,基于非线性三自由度车辆动力学模型建立了横摆角速度的卡尔曼滤波器,实现了对道路曲率的在线实时估算。在此基础上,以前方目标车辆的后侧中心是否处于本车车道为判断依据,建立了ACC系统有效目标辨识模型。研究结果表明:建立的模型能够快速识别ACC系统的有效目标,对多个目标的区分能力较强;尽管该道路曲率估算误差较大,但模型依然能够准确辨识得到ACC系统的有效目标。
In order to establish valid target distinguish method of adaptive cruise control（ACC）,microwave radar,3-axis gyroscope,lane mark detect system and other sensors were used to set up natural driving behavior test vehicle.Real road natural driving test was carried out on expressway and national road,and vehicle running state data and road traffic condition data during following process were extracted.Based on nonlinear three degrees of freedom vehicle model,Kalman filter was established for filtering yaw rate.Vehicle speed and yaw rate after filtering were used to carry out road curvature online real-time estimation.On this basis,ACC valid target distinguish model was set up on the foundation whether the center of target vehicle was in the own vehicle＇s lane.The results show that this model can recognize the valid target of ACC quickly,and it can distinguish many objects quite well.Although the curvature estimation errors arelarge,this model can still recognize the ACC system＇s valid target accurately.
出处
《长安大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2014年第3期137-144,共8页
Journal of Chang’an University(Natural Science Edition)
基金
国家自然科学基金项目(51178053
61374196)
宁波市自然科学基金项目(2012A610153)
中央高校基本科研业务费专项资金项目(2013G1221024
2013G1221025)
关键词
汽车工程
自适应巡航
目标辨识
曲率估算
卡尔曼滤波
automotive engineering
adaptive cruise control
target distinguish
curvature estimation
Kalman filter
