Actor-critic-based adaptive predefined-time optimal tracking control for a QUAV

Abstract

This paper proposes a predefined-time adaptive optimal control strategy to address the trajectory tracking problem for a quadrotor unmanned aerial vehicle (QUAV). By employing the command filter and nonsmooth error compensation mechanism, the issue of “explosion of complexity” (EOC) and the effect of filtered errors are successfully addressed, respectively. By integrating the actor-critic structure from reinforcement learning into command filtered backstepping design method, both feedback and feedforward signals are optimized to minimize the performance index. It is proved that all signals of the closed-loop system are predefined-time bounded, and the tracking errors converge to a neighborhood of the origin within a predefined time. Finally, simulation examples are conducted to verify the effectiveness and superiority of the developed predefined-time adaptive optimal control strategy.