Adaptive model-based dynamic event-triggered output feedback control of a robotic manipulator with disturbance
Jie Gao,Erlong Kang,Wei He,Hong Qiao
DOI: https://doi.org/10.1016/j.isatra.2021.04.023
IF: 7.3
2022-03-01
ISA Transactions
Abstract:This paper focuses on the stable tracking control of the manipulator with constrained communication, unmeasurable velocity, and nonlinear uncertainties. An NN observer-depended output feedback scheme in the discrete-time domain is developed by virtue of the model-based dynamic event-triggered backstepping technique in the channel of sensor to controller. For generalizing the zero-order-holder (ZOH) implementation, a plant model is built to approximate the triggered states in the time flow, and according to which, the control law is fabricated. Based on model-based error events, we construct a dead-zone triggered condition with a dynamically adjustable threshold, making the threshold evolve with the system performance, to achieve flexible communication scheduling and avoid the accumulation of triggers in small tracking errors. The internal and external nonlinear uncertainties are online compensated by the neural network, and the aperiodic adaptive law is derived in the sense of control stability to save the computation. Finally, the conditions for semi-global ultimate uniform bounded (SGUUB) of all variables are given via impulse Lyapunov analysis, and a positive lower bound in the time interval between consecutive executions to guarantee the Zeno free behavior is obtained. Simulations are conducted on a three-link manipulator to illustrate the effectiveness of our method.
automation & control systems,instruments & instrumentation,engineering, multidisciplinary