Parametric dynamic modeling-based robust nonlinear model predictive control design for tracking unmanned surface vehicle trajectory

Abstract

This paper proposes a robust nonlinear model predictive control scheme based on parameter identification (EF-NMPC) for trajectory tracking control of unmanned surface vehicle (USV) under uncertain time-varying disturbances and actuator saturation. An extended Kalman filter (EKF)-based parameter identification method is employed to construct the maneuvering motion model of the USV, enhancing the accuracy of internal dynamic estimation. A novel nonlinear disturbance observer (FDO) is designed to estimate unknown disturbances and uncertainties affecting the system, and this observer is integrated into the predictive model to mitigate the impact of model inaccuracies on control performance. The Cybership II model ship is used as the simulation experiment object, with various operating conditions and trajectory tracking scenarios designed to test the tracking performance of the robust nonlinear model predictive control (NMPC) scheme based on parameter identification. Comparative simulations with EF-LMPC and other control schemes demonstrate that the proposed control scheme exhibits strong robustness, with lower average tracking error and faster error convergence speed compared to current mainstream control schemes, resulting in better tracking performance.