Fixed‐Time Self‐Triggered Adaptive Neural Practically Tracking Control for Nonlinear Systems With Predefined Output Constraints

ABSTRACT This paper investigates adaptive fixed‐time tracking control problems for strict‐feedback nonlinear systems under output constraints occurring in a limited time interval (OCOLT). Firstly, a shift function with the assistance of a barrier function is introduced so that systems under OCOLT are converted into unconstrained ones. By this approach, time‐varying/constant, symmetric/asymmetric, and delayed/instantaneous constraints can be treated in a unified manner without altering the system structure. Secondly, a theoretical framework is built through a generalized fixed‐time stability method, which offers a more accurate estimate of the settling time than usual fixed‐time theories. To handle unknown nonlinearities and avoid complexity explosion issues, an adaptive neural network and a generalized fixed‐time dynamic surface filter are designed, respectively. Furthermore, a self‐triggered control mechanism is developed to reduce communication burdens without additional hardware to monitor triggering errors continuously. Theoretical analysis proves that the closed‐loop system is practically fixed‐time stable, with the tracking error converging to a neighborhood of the origin within a fixed time interval, while ensuring all system signals remain bounded. Finally, the feasibility and effectiveness of the proposed control scheme are rigorously validated through a simulation.