Active Wave Compensation Control for Large Offshore Transfer Gangways Based on BP Neural Network Motion Prediction

With the rapid expansion of offshore wind farms into deeper waters, large-scale wave-compensated transfer gangways have become critical equipment for ensuring safe personnel transfer between maintenance vessels and wind turbines. However, the increased mass and inertia of large gangways introduce significant control delays in hydraulic actuation, data acquisition, and compensation calculation processes, which severely degrade compensation accuracy and threaten operational safety. This paper proposes a predictive active wave compensation control scheme based on Back Propagation (BP) neural network to address this challenge. First, a ship vertical-plane kinematic model was established to derive the mapping relationship between ship motions (roll, pitch, heave) and gangway end-point positions. The impact of 3–5 s control delays on gangway stability was systematically analyzed, revealing that uncompensated delays can cause end-point deviations exceeding 4 m. Then, an optimized BP neural network model was constructed for short-term ship six-degree-of-freedom (6-DOF) motion prediction, with key parameters (input duration, hidden layer nodes, random seed) tuned to achieve high prediction accuracy. Finally, the proposed predictive control scheme was validated through numerical simulations. The results demonstrate that the scheme can reduce the maximum end-point deviations by more than 80% for delays up to 5 s, effectively mitigating the adverse effects of control delays. This research provides a technical solution for improving the safety and reliability of large offshore transfer gangways.