Abstract: This paper focuses on the path tracking of the unmanned vehicle, if the vehicle encounters an obstacle, a feasible path needs to be re-planned locally. Firstly, this paper use the vehicle point mass model and the MPC local path planning algorithm to obtain the vehicle dynamics constraints and Realize the local path of the obstacle avoidance function, and then track the path based on the two-degree-of-freedom vehicle dynamics model and MPC algorithm, and finally the algorithm is verified by Carsim& Simulink platform. The results show that the local path planning and path tracking algorithm can be reliable to plan a local path to avoid obstacles, and realize path tracking while at high speed.

Key words: Unmanned driving vehicle; Model predictive control; Local path planning; Path tracking

摘要： 文章主要针对无人驾驶车辆在进行路径跟踪遇到障碍物时，需要局部重新规划出一条可行 路径的问题，首先基于车辆点质量模型的 MPC 局部路径规划算法，得到满足车辆动力学约束并实 现避障功能的局部路径，然后在二自由度车辆动力学模型的基础上基于 MPC 进行路径的跟踪，最 后使用 Simulink/Carsim 进行联合仿真验证，结果表明基于该局部路径规划与路径跟踪算法能够可 靠地规划出避开障碍物的局部路径，实现高速下的路径跟踪。

关键词: 无人驾驶；MPC；局部路径规划；路径跟踪