基于模型预测控制的车辆轨迹跟踪稳定性研究  

A Study of the Vehicle Track Tracking Stability Based on the Model Predictive Control

作　　者：宋晓华[1,2] 邵毅明 屈治华[2] 束海波 SONG Xiaohua;SHAO Yiming;QU Zhihua;SHUHaibo(School of Mechanotronics and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China;College of Traffic and Transportation,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学机电与车辆工程学院,重庆400074 [2]重庆交通大学交通运输学院,重庆400074

出　　处：《重庆理工大学学报（自然科学）》2020年第8期1-8,共8页Journal of Chongqing University of Technology：Natural Science

基　　金：国家重点研发计划子课题(2016YFB0100905)。

摘　　要：因车辆的实际行驶环境非常复杂,导致车辆轨迹跟踪算法的稳定性问题成为智能驾驶技术发展的重点和难点之一。然而,目前基于模型预测控制的车辆轨迹跟踪算法在车辆轨迹预测求解过程中,二次规划方程变量上下限设置不当可能会产生预测轨迹偏离期望轨迹的问题。针对这一问题,基于车辆的2自由度动力学模型,建立了运动学的空间状态方程,提出了基于车辆2自由度动力学模型的车辆轨迹预测改进算法,采用MPC控制算法,建立以控制增量和状态增量为变量的优化控制目标,并根据文献中的车辆机械特性极限和车辆纵横向综合能力实验数据对模型添加适当的控制增量约束和状态增量约束,通过滚动优化和反馈校正实现轨迹跟踪,解决了车辆轨迹预测求解过程中二次规划方程变量上下限设置不当导致的车辆预测轨迹偏离期望轨迹的问题。研究结果表明:基于车辆横向动力学模型的车辆轨迹预测改进方法可以更加准确地预测车辆轨迹,并能更加稳定地跟踪期望轨迹。Because the actual driving environment of vehicles is very complex,the stability of vehicle trajectory tracking algorithm has become one of the key and difficult problems of the intelligent driving technology.However,during the solving process of vehicle trajectory prediction based on the model predictive control,the improper upper and lower bounds of the quadratic programming equation variables may cause the difference between the predicted trajectory and desired trajectory.In order to solve this problem,the space state equation of kinematics based on the lateral dynamic model of vehicle is established and an improved vehicle trajectory prediction algorithm based on the two-degreeof-freedom dynamics model of vehicle is proposed.The MPC control algorithm is adopted to control the optimal control objectives with controling the increment and state increment as variables.According to the experimental data of vehicle mechanical characteristics and vehicle longitudinal and lateral comprehensive capability in the listed reference,appropriate control increment constraints and state increment constraints are used to achieve the trajectory tracking by rolling optimization and feedback correction,which can overcome the difference between the predicted trajectory and the expected trajectory caused by the improper upper and lower bounds of the quadratic programming equation variables during the process of the vehicle trajectory prediction.The results show that the improved vehicle trajectory prediction method based on the vehicle lateral dynamics model can predict the vehicle trajectory more accurately and track the desired trajectory more stably.

关 键 词：模型预测控制 车辆轨迹 轨迹跟踪算法 稳定性 

分 类 号：U461[机械工程—车辆工程]