Abstract: With the improvement of the living standards of Chinese people, the popularity of motor vehicles continues to increase, and with the increase in the number of vehicles, traffic congestion, frequent accidents and other problems are becoming more and more serious, for passenger car drivers with immature driving ability, it is more likely to cause traffic accidents due to improper parking, so automatic parking has become an urgent need to develop technology. The main research content of this paper is the control algorithm of automatic parking system trajectory tracking, which adopts the vertical parking environment, uses the model predictive control (MPC) algorithm for path tracking according to the feasible path that has been planned, and increases the control amount increment in the cost function to achieve the accuracy and stability of the trajectory tracking process. In order to verify the control effect, the Carsim-Simulink joint simulation platform is built, and the results show that the MPC control algorithm can achieve good trajectory tracking effect when parking.

Key words: Automatic parking; Trajectory tracking; MPC algorithm

摘要： 随着中国人民的生活水平日益改善，机动车的普及程度不断提升，而随着车辆数量的 增加，交通拥堵、事故频发等问题日益严重，对于驾驶能力不成熟的乘用车驾驶员而言，更 易因泊车不当而造成交通事故，因此，自动泊车应运而生，成为了亟需开发的技术。文章主 要研究自动泊车系统轨迹跟踪的控制算法，采用垂直泊车环境，根据已规划出的可行路径， 利用模型预测控制（MPC）算法进行路径跟踪，通过在代价函数中增加控制量增量，以求实 现轨迹跟踪过程的精确与平稳。为验证控制效果，搭建了 Carsim-Simulink 联合仿真平台，结 果表明 MPC 控制算法可以实现泊车时良好的轨迹跟踪效果。

关键词: 自动泊车；轨迹跟踪；MPC 算法