Event-triggered Discrete-time Output Feedback Adaptive Neural Control of Robot Manipulators

Abstract

In this paper, an event-triggered control (ETC) scheme is developed for a class of networked n-link rigid robotic manipulators with unknown dynamics and unmeasured joint velocity. By employing neural networks (NNs) to approximate the unknown dynamic of the manipulators, a neural state observer is first designed to estimate the immeasurable joint velocity of the robotic manipulators. A novel NN weight updating law is designed such that the estimation errors are ultimately uniformly bounded. Based on the designed state observer, an output feedback ETC scheme is developed for rigid robotic manipulators with n-degree of freedom. Simulation study demonstrates that the proposed ETC scheme not only guarantees the ultimately uniformly boundedness of the robotic manipulator control system, but also alleviates the transmission burden of the controller-to-actuator communication network.