Actor-Critic-Based Predefined-Time Fuzzy Adaptive Optimal Control for Uncertain Nonlinear Systems with Input Saturation

The widely studied finite/fixed-time control guarantees fast convergence of the controlled systems. Yet, the adjustment of settling time remains complex, and the optimality of control signal is not considered. In this article, a predefined-time optimal tracking control scheme is proposed for uncertain nonlinear systems with input saturation. With the aid of fuzzy approximation, the reinforcement learning actorcritic structure is established, in which the actor and critic network are used to implement control actions and evaluate execution costs, respectively. Then, by introducing the actorcritic structure into the command filtered backstepping design framework, the approximated optimal control signals containing the predefined-time parameter are derived, and an easily tunable upper bound on the settling time with respect to the predefinedtime parameter is obtained. With the approximation of saturated nonlinearity using tanh function, the input saturation constraint is satisfied. Stability analysis proves that all signals in the closedloop system can converge to a small neighborhood near the origin in a predefined time. Eventually, comparative simulations on quadrotor attitude system are carried out to assess the validity of the developed control strategy.