Predefined-time trajectory tracking control for USVs via dual-observer architecture and particle swarm optimization

This study proposes a predefined-time trajectory tracking framework for unmanned surface vehicles (USVs) under ocean-current disturbances and external time-varying perturbations. A layered compensation architecture combines an ocean-current disturbance observer for kinematic disturbances with an extended state observer for dynamic perturbations. Based on their estimates, a predefined-time sliding mode controller is developed to drive velocity errors into a bounded neighborhood of the origin within a user-specified time, irrespective of initial conditions. Particle swarm optimization is used to tune the controller gains and improve overall performance. Comparisons with a PSO-free predefined-time controller, a switching nonsingular terminal sliding mode controller, a finite-time adaptive integral sliding mode controller, and an observer-free baseline demonstrate that the proposed method provides higher tracking accuracy, faster convergence, and stronger robustness under compound disturbances.