基于自适应采样周期和预测时域MPC的车辆路径跟踪控制

Vehicle Path Tracking Control Based on Adaptive Sampling Period and Predictive Time Domain MPC

为解决自动驾驶车辆在低附着路面路径跟踪控制精度较低的问题,设计了一种自适应采样周期和预测时域MPC控制器。首先,结合车辆动力学模型和MPC算法设计了MPC控制器,并加入轮胎侧偏角约束;然后,分析控制器的采样周期和预测时域对控制效果的影响,提出一种综合考虑采样周期和预测时域的自适应控制策略,通过车辆前轮转向角更新采样周期,通过车速更新预测时域;最后,使用Carsim和Matlab/Simulink联合仿真平台,在低附着路面的不同车速条件下进行仿真实验。结果表明,当车速为25 km/h和45 km/h时,相较于固定控制参数MPC控制器,自适应采样周期和预测时域MPC控制器的最大横向误差分别减小140.2mm和40.8mm,其在不同车速下的路径跟踪控制精度均更高,横摆角速度和质心侧偏角均在合理范围内,车辆稳定性较好,证明所提路径跟踪控制器在低附着路面具有较高的控制精度和可行性。

In order to solve the problem of low accuracy of path tracking control of autonomous vehicle on low adhesion road surfaces,an adaptive sampling period and prediction time domain MPC controller was designed.Firstly,the MPC controller was designed by combining the vehicle dynamics model and the MPC algorithm with the tire sideslip angle constraints.Then,the influence of sampling period and predictive time domain of controllers on the control effect was analyzed.A adaptive control strategy that comprehensively considered the sampling period and predictive time domain was proposed,in which the sampling period was updated by the front wheel steering angle,and the predictive time domain was updated by vehicle speed.Finally,using Carsim and Matlab/Simulink co-simulation platform,simulation experiments were carried out under different vehicle speeds on low adhesion road surface.The results showed that when the vehicle speed was 25 km/h and 45 km/h,compared with the fixed control parameter MPC controller,the maximal lateral errors of the adaptive sampling period and predictive time domain MPC controller were reduced by 140.2 mm and 40.8 mm respectively;its path tracking control accuracy was higher at different vehicle speeds;the yaw rate and sideslip angle were all within reasonable limits,and the vehicle stability was good.It means the proposed path tracking controller has high control accuracy and feasibility on low adhesion road surfaces.