Enhanced Lyapunov-Based Model Predictive Control for Wheeled Mobile Robot With Improved Tunnel Performance Constraints

This study addresses the maneuvering control problem of a wheeled mobile robot (WMR) operating in dynamic and uncertain environments, where smooth error convergence and robustness against external disturbances are critical challenges. An enhanced Lyapunov-based model predictive control (ELMPC) method is developed, in which an improved tunnel performance function is incorporated into an auxiliary controller to impose stricter contraction constraints on the WMR system, enabling smoother and faster error convergence. In addition, an improved tunnel performance constraint (ITPC) scheme is introduced to eliminate the initial-condition dependence of conventional prescribed performance control. By employing a performance transition function, the ITPC ensures that performance bounds are satisfied under arbitrary initial errors while avoiding excessive transient responses. To compensate for unknown and time-varying disturbances, a neural network–based disturbance predictor is integrated into the control architecture. The predictor estimates disturbances online and transforms tracking errors into prediction errors, thereby accelerating convergence and enhancing robustness. The proposed control strategy is implemented within a Lyapunov-based model predictive control framework with explicit consideration of actuator constraints, where contraction constraints guarantee closed-loop stability and recursive feasibility. Experimental results demonstrate improved tracking performance, robustness, and constraint satisfaction under complex operating conditions.