Event-triggered decentralized optimal tracking control for modular manipulator with constrained input

In this paper, a decentralized optimal tracking control method considering constrained input is proposed for modular manipulator system. The dynamic model of the subsystem is established by using the joint torque feedback(JTF) technology. The model uncertainty of the system is observed by an adaptive observer, and the bounded function is used to deal with the constrained input. The controller is optimized by combining event-triggered mechanism and adaptive dynamic programming (ADP) control algorithm. Furthermore, in order to reduce the computation of solving Hamilton-Jacobi-Bellman(HJB) equation, critical neural network is used to solve it online. The weights of neural networks(NN) is adjusted by event-triggered conditions. By using Lyapunov theory, the trajectory tracking error of the modular manipulator is proved to be uniformly ultimately bounded(UUB) with constrained input. Finally, the effectiveness of the proposed scheme is verified by experiments.