Robust Event-Triggered MPC with Parallel Optimization and Disturbance Rejection for Mobile Robots

Abstract

This paper proposes an event-triggered parallel optimization model predictive control (MPC) strategy integrated with an extended state observer (ESO) for mobile robots operating under unknown disturbances. In the proposed framework, a conventional MPC formulation is employed to minimize the tracking error and control increments, thereby generating real-time control outputs. To improve control performance and computational efficiency, a parallel optimization strategy is introduced to optimize MPC parameters by jointly considering the accumulated tracking error, state convergence time, and computational cost. An improved Cuckoo Catfish Optimizer (ICCO) is adopted to perform the parallel parameter optimization in real time, enabling adaptive tuning of the MPC parameters. Furthermore, to address unknown external disturbances and model uncertainties, a nonlinear ESO is designed to estimate and compensate disturbances, thereby enhancing the robustness and tracking accuracy of the system. The effectiveness of the proposed control strategy is validated through experimental studies conducted on the MR500 mobile robot platform. The comparative experimental results demonstrate that the proposed method achieves an approximately 40% reduction in tracking error compared with the conventional MPC, while also exhibiting faster response speed, higher tracking accuracy, and stronger robustness against both internal and external disturbances.