Mecanum Wheel AGV Trajectory Tracking Control Based on Efficient MPC Algorithm

Abstract

In response to the challenge of insufficient trajectory tracking accuracy and low solution efficiency of Mecanum wheel AGV (Automated Guided Vehicle) under complex and constrained working conditions, this paper proposes an efficient Model Predictive Control (MPC) method to achieve superior tracking performance and robustness. Initially, a linear error model of the mobile platform is established based on pose error, serving as the predictive model for the MPC controller. A target function is designed to transform the trajectory tracking control problem into an optimal control problem. To handle inequality constraints, penalty terms are introduced into the objective function, and the resulting constrained problem is subsequently solved to approximate the optimal solution for the original inequalities. To alleviate the computational burden associated with real-time optimization problem-solving, an efficient MPC algorithm. has been developed. To ensure closed-loop stability under the MPC control method, stability constraints are imposed on the new optimization problem. Simulation results demonstrate that, in comparison to traditional Model Predictive Control methods, the proposed approach reduces the average solution calculation time by 5.1% and the maximum single calculation time by 13.7%, all while maintaining trajectory tracking accuracy. These results validate the algorithm's feasibility, effectively addressing the challenges associated with solving Model Predictive Control problems.