Adaptive Critic Learning for Optimized Tracking Control of an Unmanned Surface Vehicle with Guaranteed Performance
Lin Chen, Shi‐Lu Dai · 2024
This paper develops a critic-based optimal control strategy for an unmanned surface vehicle (USV). To address the constraint requirement on tracking errors, universal barrier function is incorporated with the optimal control scheme, such that the tracking errors satisfy the transient and steady-state performance. Then, to obtain the optimal control law, a single-critic architecture is constructed to solve the Hamilton-Jacobi-Bellman (HJB) equation. Meanwhile, the basis function in the critic network is selected as Gaussian function. With the improved critic weight updating law, the proposed optimal control algorithm not only relaxes the persistence of excitation (PE) condition but also guarantees the uniformly ultimately boundedness (UUB) of critic network weight. Finally, simulation results validate the effectiveness the presented optimized control scheme.