DE-optimized LQR for trajectory tracking of omnidirectional mobile robots

Abstract

Omnidirectional mobile robots (OMRs) have become increasingly important in modern automation due to their ability to navigate freely in any planar direction, making them highly suitable for industrial logistics, service robotics, and collaborative systems. However, achieving a balance between tracking accuracy and actuator limitations under physical constraints remains challenging. The Linear Quadratic Regulator (LQR) is an effective optimal controller, but its performance depends heavily on manually tuned weighting matrices. To address this issue, this study integrates the Differential Evolution (DE) algorithm with the LQR controller to automatically optimize these parameters. The proposed controller significantly improves tracking accuracy and eliminates the need for manual tuning making it highly suitable for advanced control of OMRs. Simulation results compared with a conventional Proportional Integral (PI) controller confirm that the DE-LQR strategy provides superior convergence behavior, lower tracking error, and enhanced stability.