Prediction of ship future motions at different predictable time scales for shipboard helicopter landing in regular and irregular waves

Differing from the traditional helicopter landing on land, due to ship motions induced by actual waves, the landing safety of shipboard helicopter on vessels in waves deserves to be studied. To provide a judgment basis for the timing of shipboard helicopter landing, a hydrodynamic analysis method based on the Long-Short-Term Memory (LSTM) neural network is proposed to achieve the real-time prediction of ship motions in this study. Firstly, the hydrodynamic formulations based on potential flow theory for nonlinear ship motions in waves are presented. The implementation details of the LSTM neural networks are described. Secondly, the numerical verification study is carried out in regular and irregular waves, respectively. Furthermore, a destroyer model is employed to illustrate the proposed framework, and the results corresponding to the different time scales in waves are presented. Finally, the work of this paper is summarized and the certain conclusions are drawn.