Adaptive Critic Learning-based Prescribed Performance Optimal Control for Nonlinear System with Input Saturation

Prescribed performance control has emerged as an effective approach to address the user-specified requirements of steady-state accuracy and settling time in control systems. However, in practical scenarios, a new challenge arises when considering input saturation limitations. Existing prescribed performance control methods often come at the cost of large control inputs leading to stability issues when meets input saturation. A novel approach integrating input saturation and prescribed performance optimal tracking control is proposed. First, A reference signal modification mechanism is introduced to dynamically adapt the reference signal based on the level of input saturation. Additionally, a performance scaling functions and an error transformation functions are employed to capture userspecified accuracy and settling time requirements, transforming the prescribed performance control problem into an optimal control problem. To achieve optimal control, the utilization of integral reinforcement learning and adaptive critic learning is explored, enabling adaptive optimal tracking control in the presence of uncertain system dynamics. This approach offers a simpler structure and lower computational complexity compared to previous methods. All signals in the closed-loop system are bounded. The simulation results verify the proposed method.