基于增量MPC和转向角优化的泊车轨迹跟踪研究
Research on parking trajectory tracking based on incremental MPC and steering angle optimization
徐佳宝 1张国良 1汪坤1
作者信息
- 1. 四川轻化工大学自动化与信息工程学院,四川宜宾 644000;人工智能四川省重点实验室,四川宜宾 644000
- 折叠
摘要
针对传统模型预测控制(model predictive control,MPC)下泊车过程中轨迹跟踪精度较低且实时性较差的问题,提出一种改进的MPC控制方法.首先,基于车辆运动学模型,加入控制量和控制增量约束,设计增量MPC控制器,提升轨迹跟踪精度;其次,根据车辆当前状态,利用卡尔曼滤波算法优化转向角,防止转向角出现突变的情况;最后,通过搭建Matlab/Simulink与Carsim联合仿真环境,将增量MPC控制器与MPC控制器进行对比分析.实验结果表明:在平行泊车工况下,增量MPC比MPC的最大横向误差小 6.3 cm,最大航向误差小 1.23°.在垂直泊车工况下,增量MPC最大横向误差比MPC小 7 cm,最大航向误差小 0.46°.仿真结果验证该轨迹跟踪控制器具有较好的控制性能.
Abstract
Aiming at the problem of low tracking accuracy and poor real-time performance in parking process under traditional model predictive control(MPC),an improved MPC control method is proposed.Firstly,based on the vehicle kinematics model,the incremental MPC controller is designed by adding control quantity and control increment constraints to improve the tracking accuracy;Secondly,according to the current state of the vehicle,the Kalman filter algorithm is used to optimize the steering angle to prevent sudden changes in the steering angle;Finally,by building a joint simulation environment of Matlab/Simulink and Carsim,the incremental MPC controller is compared with the MPC controller.The experimental results show that under parallel parking conditions,the maximum lateral error of incremental MPC is 6.3 cm less than that of MPC,and the maximum heading error is 1.23 ° less.Under the condition of vertical parking,the maximum lateral error of incremental MPC is 7cm smaller than MPC,and the maximum heading error is 0.46 ° smaller.The simulation results show that the trajectory tracking controller has good control performance.
关键词
轨迹跟踪/模型预测控制/转向角/卡尔曼滤波Key words
trajectory tracking/model predictive control/steering angle/Kalman filtering引用本文复制引用
基金项目
四川省应用基础研究项目(2019YJ4013)
四川轻化工大学研究生创新基金(Y2022142)
出版年
2024
NETL
NSTL
万方数据