An adaptive NMPC for ROVs trajectory tracking with environmental disturbances and model uncertainties

A novel hybrid adaptive control method is presented for trajectory tracking of remotely operated underwater vehicles (ROVs) that addresses unknown disturbances and model uncertainties in this paper. Traditional nonlinear control methods struggle to handle external disturbances and uncertainty in system model. To address the trajectory tracking control needs of ROVs in complex underwater environments, a kinematic and dynamic model is first developed for a fully actuated ROV with six degrees of freedom (6-DOF). The trajectory tracking problem is formulated as an online, nonlinear receding horizon optimization process. Control increments are computed as inputs to this nonlinear optimization problem. An L1 adaptive control method (L1AC) is then developed, incorporating a state observer, adaptive control law, and time filter. The framework retains the rolling optimization process of nonlinear model predictive control (NMPC) while integrating the L1 adaptive component for instant compensation of unknown disturbances and model parameter mismatches. Numerical simulations were conducted to compare the trajectory tracking performance of the proposed hybrid adaptive method with the NMPC method under various disturbances, including ocean currents, waves, random forces, and model uncertainties. The results confirm that the proposed hybrid adaptive control scheme is more effective and robust than the standalone NMPC approach across various scenarios.