Event-Triggered Control for Nonlinear Uncertain Strict-Feedback Systems: An Adaptive Filtering Approach

Abstract

This note proposes a novel output-feedback event-triggered control method for nonlinear uncertain strict-feedback systems. It incorporates dual asynchronous triggering mechanisms for both the system's output and control input, utilizing a specifically designed adaptive filtering method. The first mechanism aims to reduce the burden on sensor to controller communication, while the second determines when the controller needs to be updated. Particularly, an adaptive neural state observer, reliant on the filtered version of sampled output, is designed to estimate the system's states. Then, differentiable virtual controls are formulated using the estimated states within the framework of the dynamic surface control. Hence, the proposed approach reduces the number of triggering mechanisms and required communication channels compared to existing results. By using the online approximation technique with adaptation schemes, the unknown nonlinearities are approximated without the need for global Lipschitz and linear growth conditions, as well as without encountering overparameterization issue. Finally, the closed-loop stability is analyzed, and proofs for the avoidance of Zeno behavior are provided.