Advancing unmanned surface vehicle heading control: real-sea validation of a practical model predictive controller for a medium-scale hybrid-power trimaran

Abstract The development of medium- to large-scale, high-performance, and environmentally-friendly unmanned surface vehicles (USVs) is a burgeoning trend in intelligent marine systems. The heading controller is crucial for USVs to execute various missions, especially given their inherent underactuation characteristics. Despite extensive exploration of numerous control strategies, the majority of studies have been confined to smaller, conventional USVs with an emphasis on simulation validation. This limitation stems from the significant gap between theory and practical engineering. We here introduce a model predictive controller for the heading control of a medium-scale hybrid-power unmanned trimaran and validate its performance in real sea. This controller directly accounts for the amplitude and increment constraints of the water-jet thrust angle, and its asymptotic stability is established with a terminal cost. Experimental results demonstrate that the controller’s ability to effectively track and maintain constant desired heading angles without any disturbance compensation, surpassing the traditional proportional-integral-derivative controller in terms of response speed, overshoot, steady-state error, and robustness. We further showcase the successful application of this controller in path-following missions, evidencing its precise adherence to desired paths. A critical aspect of our research is monitoring the controller’s execution time in all experiments, affirming its real-time response capabilities given the onboard hardware’s limited computational power. This research not only contributes a significant advancement in the field of USV control but also provide a new solution for the integration of theoretical control strategies into practical marine engineering.