Event-triggered optimal control for modular reconfigurable manipulators with input constraints based on model predictive control.

This paper proposes an event-triggered optimal control method for modular reconfigurable manipulators(MRMs) based on model predictive control(MPC). By using a decentralized optimization method based on MPC, the optimal control problem of MRMs is transformed into independent optimization tasks for each module, while a global MPC optimization framework is utilized to coordinate the modules, ultimately optimizing the overall performance of the entire system. In order to avoid the safety hazards caused by excessive torque, hyperbolic tangent function is added to constrain the input torque. Considering the inaccuracies in the modeling process, adaptive dynamic programming (ADP) is introduced into MPC to improve the robustness of the system. A critical neural network (NN) is designed to solve the Hamilton-Jacobi-Bellman (HJB) equation, yielding the system's optimal solution. Lyapunov theory is used to prove that the trajectory tracking error is uniformly ultimately bounded (UUB). Finally, the comparative experimental results demonstrate that the proposed method achieves significant improvements in reducing tracking error, minimizing resource consumption, and enhancing constrained torque capability.