Abstract:

In order to improve the path tracking accuracy and driving stability of autonomous vehicles under different operating conditions, a path tracking control algorithm based on vehicle dynamics model and model predictive control theory is proposed, which adaptively adjusts the weight coefficient of the cost function according to tracking deviation and road curvature. The weight coefficient matrix of the traditional MPC path tracking controller is dynamically optimized by fuzzy controller, so that when the deviation between the vehicle and the reference path is relatively large, the deviation can be quickly reduced to ensure the safety of the vehicle. When the path tracking deviation is relatively small and the reference path curvature is relatively small, the system puts more emphasis on the requirements of driving stability. In order to verify the performance of the designed path tracking controller, CarSim/Simulink co-simulation model is built. Compared with MPC path tracking controller with constant weight coefficient, the tracking accuracy and stability of MPC path tracking controller with adaptive weight coefficient are improved.

Key words: autonomous vehicles, path tracking, model predictive control, adaptive, weight coefficient, fuzzy controller

摘要：

为提高自动驾驶车辆在不同工况下的路径跟踪精度和行驶稳定性，基于车辆的单轨模型和模型预测控制（MPC）理论，提出一种依据跟踪偏差和道路曲率自适应调整成本函数权重系数的路径跟踪控制算法。该算法主要是通过模糊控制理论动态优化传统MPC路径跟踪控制器中权重系数矩阵，使得当车辆与参考路径偏差比较大时，能够快速减小跟踪偏差，保证车辆行驶的安全性；当路径跟踪偏差比较小，且参考路径曲率比较小时，使得系统更加侧重行驶稳定性的要求。为验证所设计的路径跟踪控制器的性能，搭建CarSim/Simulink联合仿真模型，在联合仿真过程中，基于权重系数自适应的MPC路径跟踪控制器与基于权重系数为常量的MPC路径跟踪控制器相比，路径跟踪精度和车辆的行驶稳定性均得到了提高。

关键词: 自动驾驶车辆, 路径跟踪, 模型预测控制, 自适应, 权重系数, 模糊控制