SACI framework-based fixed-time learning control for nonlinear systems with asymmetric constraints

This paper investigates a fixed-time learning control scheme for constrained nonlinear systems using a Sub-Actor-Critic-Identifier framework, considering asymmetric states and input constraints. Specifically, a universal barrier function and a prescribed performance function are introduced to address dynamic asymmetric state constraints. Fuzzy logic systems are utilized to compensate for asymmetric input constraints and to approximate unknown nonlinear dynamics. Subsequently, by combining fixed-time control with reinforcement learning, an Sub-Actor-Critic-Identifier framework is constructed via optimal backstepping techniques, thereby enabling the approximate solution of Hamilton-Jacobi-Bellman equation and achieving rapid system convergence. Additionally, a dynamic event-triggered mechanism is developed to reduce the controller's sampling frequency, thereby alleviating computational and communication burden. Finally, Lyapunov stability analysis demonstrates that all signals remain bounded and converge within a fixed time. Simulation results verify the effectiveness of the proposed control strategy.