基于μ综合方法的智能车辆人机共驾的鲁棒横向控制  

Robust Lateral Control of Intelligent Vehicle in the Human-machine Sharing Based onμ-synthesis

作　　者：谢有浩 魏振亚 赵林峰 王家恩 陈无畏 XIE Youhao;WEI Zhenya;ZHAO Linfeng;WANG Jiaen;CHEN Wuwei(College of Mechanical and Electric Engineering,Chuzhou University,Chuzhou 239000;Anhui Leopaard Automobile Co.,Ltd.,Chuzhou 239064;School of Automotive and Transportation Engineering,Hefei University of Technology,Hefei 230009)

机构地区：[1]滁州学院机械与电气工程学院,滁州239000 [2]安徽猎豹汽车有限公司,滁州239064 [3]合肥工业大学汽车与交通工程学院,合肥230009

出　　处：《机械工程学报》2020年第4期104-114,共11页Journal of Mechanical Engineering

基　　金：安徽省科技重大专项(17030901060);国家自然科学基金(51675151,U1564201)资助项目。

摘　　要：在大多数智能车辆横向控制研究中,存在未考虑驾驶员误操作的影响这一不足。以人机共驾控制问题为研究对象,将驾驶员操纵转矩和车辆状态作为控制器输入。首先,建立转向系统和车辆二自由度模型,在车辆局部坐标系中,根据预瞄点曲率信息实现虚拟路径的规划,基于车辆状态和目标车道设计上层期望横摆角速度控制器。其次,将侧向风和驾驶员误操作作为干扰输入,以车辆状态中的横摆角速度、转向盘转角、转矩传感器测量值和期望横摆角速度作为控制器反馈变量,考虑车辆参数摄动及传感器测量噪声等影响,设计下层μ综合控制器,使车辆跟踪期望横摆角速度和期望的横向位移,确保车辆能稳定地跟踪目标路径。最后,进行自动换道和车道保持仿真,并基于Carsim/Labview的硬件在环试验台上进行硬件在环试验,仿真和试验结果均表明,提出的横向控制方法能辅助驾驶员更好的跟踪目标车道,且对侧向风和驾驶员误操作均有很好的干扰抑制性能。In most intelligent vehicle lateral control studies,there is a shortcoming that the influence of driver’s misoperation is not taken into account.The human-machine sharing control problem is taken as the research object.The driver steering torque and vehicle state are considered as inputs of the controller.Firstly,the steering system and 2-DOF vehicle model is established,and in the vehicle local coordinate system,a virtual path planning is realized according to the curvature information of the preview point.And an upper desired yaw rate controller is designed based on vehicle statue and the desired lane.Secondly,a lowerμ-synthesis robust controller considering vehicle parameter uncertainty and sensor noise is designed to make vehicle tracking the desired yaw rate and the desired lateral position,ensure vehicle tracking the target path stably,in which crosswind and driver’s misoperation are considered as disturbance input,and yaw rate,steering angle,steering torque and desired yaw rate are considered as feedback variables.Finally,lane changing and lane keeping simulation is carried out,and the HIL test is implemented on the HIL test bench based on CarSim/LabVIEW,the simulation and test results show that the proposed lateral control algorithm not only assist driver to control vehicle tracking desired lane more accurately but also can restrain the crosswind and prevent driver’s misoperation.

关 键 词：智能汽车 横向控制 人机共享控制 Μ综合控制 

分 类 号：TG156[金属学及工艺—热处理]