基于模型预测控制的智能车横纵向控制器设计

针对智能车横纵向控制中路径跟踪精度、行驶稳定性以及乘坐舒适性等问题,提出了基于模型预测控制(MPC)的横纵向综合控制方法。速度规则系统根据参考路径曲率与车辆跟踪位移误差计算出期望速度曲线,速度跟踪控制采用分层式控制器,上层控制器利用MPC算法计算期望加速度,下层控制器利用车辆逆纵向动力学模型对车辆的驱动和制动进行协调控制。横向控制器根据参考路径、车辆反馈状态以及纵向上层控制器的期望速度计算车辆前轮转角。最后通过实验对比本算法与恒速MPC横向控制算法的轨迹误差,结果表明:本算法控制的车辆横向位移均方根误差减小了0.051 m,有效提高了车辆轨迹跟踪的控制精度。

Design of Lateral and Longitudinal Controller of Intelligent Vehicle Based on Model Predictive Control

Aiming at the tracking accuracy,vehicle stability and riding comfort in the lateral and longitudinal control of intelligent vehicles,a method of lateral and longitudinal tracking control based on model predictive control(MPC)was proposed according to the vehicle dynamics model.The desired speed curve was calculated by the speed planning system based on the reference path curvature and vehicle lateral displacement error feedback.The expected acceleration was calculated by the upper controller based on MPC algorithm.The driving and braking of the vehicle were controlled by the lowerlevel controller based on the vehicle inverse longitudinal dynamics model.The front wheel angle of the vehicle was calculated by the lateral controller based on the reference path information,the vehicle feedback status information and the desired speed information of the longitudinal upper controller.Finally,through experiments,the errors of this algorithm and the constant speed MPC lateral control algorithm and the reference trajectory were compared.The experimental results show that the root mean square error of the vehicle lateral displacement controlled by this algorithm is reduced by 0.051 m,which improves the control accuracy of vehicle trajectory tracking effectively.