Adaptive self-triggered control for a nonlinear uncertain system based on neural observer

The issue of adaptive tracking control for nonlinear uncertain system with self-triggered input and immeasurable states is investigated. The Radial Basis Function of Neural Networks (RBFNNs) is introduced to approximate unknown nonlinear functions. Based on this, an observer is constructed to estimate the immeasurable states. Considering that the self-triggered input may cause poor tracking performance, the prescribed performance bound method is combined into co-designing. Furthermore, a self-triggered adaptive control scheme is presented to guarantee close-loop system signals are bounded. Compared with the event-triggered scheme, the presented self-triggered scheme can compute the next trigger point through the current one. It is not required to continuously monitor the measurement error to determine whether the triggering condition is reached. By the given control scheme, the tracking error can be bounded in prescribing performance bounds, and reaches the balance of network resource saving and tracking performance. These are verified by the given simulation example.