To enhance intelligent vehicle path tracking accuracy and adaptability across different road conditions and speeds, this paper introduces a parameter-adaptive MPC method combined with a PSO-BP neural network. The MPC algorithm was formulated with path tracking accuracy and control increment as the key components of its cost function. The PSO-BP neural network was employed to dynamically adjust the weights of the MPC cost function in real time. The controller was implemented using a co-simulation framework built with CarSim and MATLAB/Simulink. Simulation results under different road adhesion levels and vehicle speeds demonstrated the proposed algorithm's effectiveness.
Intelligent Vehicle Path Tracking Control based on Adaptive Model Predictive Control
2024-10-25
1040490 byte
Conference paper
Electronic Resource
English
Adaptive Nonlinear Model Predictive Path Tracking Control for a Fixed-Wing Unmanned Aerial Vehicle
British Library Conference Proceedings | 2009
|