Research on Three- Degree-of- Freedom Wave Compensation Gangway Compensation Algorithm and Compensation Amount Prediction

In the randomly changing marine environment, using traditional prediction models to forecast gangway compensation in advance sometimes leads to significant random fluctuations in the predicted values, increasing the risk factor for offshore platform personnel landing. This paper proposes a PSO-improved LSTM prediction model, which accurately predicts gangway joint compensation to facilitate better the smooth docking between the gangway and the offshore platform. The PSO-improved LSTM model is based on the improvement of the traditional LSTM prediction model, utilizing the PSO algorithm to optimize the topology, number of iterations, and learning rate of the improved LSTM model, resulting in the PSO-improved LSTM prediction model. By utilizing the traditional RNN model, the LSTM model, the improved LSTM model, and the PSO-improved LSTM model, compensation amounts for the rotation, pitch, and extension of a three-degree-of-freedom gangway are forecasted and compared with actual values. The results indicate that the PSO-improved LSTM model provides more accurate predictions for the compensation amounts of the three joints of the three-degree-of-freedom gangway.