Robust NMPC with Disturbance Compensation for Unmanned Surface Vehicle

Abstract

Autonomous navigation requires an unmanned surface vehicle (USV) to maintain accurate heading and trajectory tracking despite the presence of environmental disturbances. In practice, disturbances such as waves and wind can cause the vessel to deviate from its desired course, reducing navigation accuracy and overall control performance. This paper proposes a disturbance compensating nonlinear model predictive control (DC-NMPC) strategy to improve the heading control of a USV operating under such conditions. The proposed controller extends the conventional nonlinear model predictive control (NMPC) framework by incorporating a disturbance compensation mechanism that estimates external disturbances and generates an additional control action to mitigate their effects. The compensated control input is then optimized to determine the rudder angle while satisfying the system constraints. The controller is developed using a two-degree-of-freedom USV model that captures the sway and yaw dynamics, with deterministic disturbances represented by constant and sinusoidal disturbance models. Simulation results show that the proposed DC-NMPC consistently achieves more accurate trajectory tracking than the conventional NMPC, as reflected by lower root mean square error (RMSE) values under both disturbance scenarios. These results demonstrate that incorporating disturbance compensation into the NMPC framework enhances the tracking accuracy and robustness of USV heading control under environmental disturbances.