Active heave compensation for remotely operated vehicle recovery operations under random wave disturbances

Active heave compensation for remotely operated vehicle (ROV) recovery operations presents significant challenges due to the system's nonlinear dynamics and the influence of random and uncertain external wave disturbances. This study proposes a control strategy that integrates feedforward model predictive control (MPC) with model reference adaptive control (MRAC) to enhance the robustness and performance of heave compensation systems. The MPC utilizes position feedback from the ROV and system modeling to predict future outputs, enabling proactive control adjustments. A hybrid model, combining high-order dynamic mode decomposition with long short-term memory neural networks, is developed to forecast vessel motion. The MPC mitigates the effects of uncertain wave conditions and system delays by incorporating the prediction. In addition, the MRAC dynamically adjusts the proportional–integral–derivative controller parameters in real time, compensating for variations in rotational inertia and nonlinear characteristics caused by winch cable retractions. A comprehensive system dynamics model is presented, which includes barge motion under random wave excitation, the electric-driven winch model, and umbilical cable dynamics. Numerical simulations demonstrate that the feedforward MPC-MRAC controller outperforms conventional control strategies of heave compensation performance, significantly enhancing the control system's disturbance rejection capabilities and operational stability. The proposed control strategy exhibits stable and superior heave compensation performance across sea conditions. These findings provide valuable insights for improving the safety and efficiency of offshore ROV recovery operations in dynamic marine environments.