Adaptive Neural Network Output-Feedback Control for Uncertain Nonlinear Systems via Event-Triggered Output

This article systematically studies the issue of adaptive neural network (NN) output-feedback control for uncertain nonlinear systems using event-triggered output. First, to tackle the problem of unmeasurable states, a compact state observer using event-triggered output is constructed. Then, since the event-triggered output signals are discontinuous, the virtual control laws in backstepping design are no longer differentiable. Hence, the dynamic surface control scheme is introduced to resolve this problem. Unlike existing work requiring system functions to satisfy Lipschitz continuity condition, adaptive NN control is incorporated into the designed algorithm to relax the above constraint. What is more, the event-triggered mechanism is also used for parameter estimation to avoid waste of computing and communication resources. Finally, the results of comparative simulations and the DC brush motor experiment are depicted to demonstrate the practicality and effectiveness of the proposed method.