Path tracking control of autonomous vehicle based on MPC

This article investigates in detail the application of path tracking control in autonomous driving technology based on model predictive control (MPC), including discussions on vehicle kinematic and discrete models, providing a comprehensive overview of non-linear vehicle models. In addition, it delves into the design of the MPC controller, including prediction models, problem descriptions, solver designs, control law designs, constraint conditions, and improved objective functions, and outlines the experimental design, including working conditions such as straight and circular tracks, and analyzes the experimental results. The optimized MPC algorithm was evaluated through simulation tests to assess its tracking ability under different trajectories, particularly targeting low-speed conditions such as parking, and showed significant improvements in control performance, including speed control performance, to support further development of autonomous driving technology.