Finite-time adaptive optimal control of uncertain strict-feedback nonlinear systems based on fuzzy observer and reinforcement learning (Q6577108)

In this paper, a finite-time adaptive optimal control strategy is applied for high-order uncertain strict feedback nonlinear systems using reinforcement learning based on an optimal control scheme which enables to design an optimal controller and to achieve global optimization. Furthermore, taking into consideration the unmeasurable states, a fuzzy observer is constructed and some fuzzy logic systems to approximate the unknown functions are used. The controller design is simplified by the inclusion of command filtering and time-based control which improves system response speed. Improved speed response, steady-state performance, improved system resilience against uncertainties and disturbances, improved stability, control optimization efficiency, and reinforcement learning-based optimal control improving control strategies through interactions with the environment are the main results of this paper. Finally, some simulations illustrate these results.