Hydrodynamic performance analysis and resistance prediction of submersible surface ship in cross wave based on CNN-Bi-LSTM neural network

To study the resistance and motion response of the submersible surface ship (SSS) in cross-waves and build a neural network forecast model to provide a decision-making basis for the design and navigation of the SSS, this paper is based on the computational fluid dynamics (CFD) method. It uses the momentum source method for wave generation and the force method and applies the overlapping grid technology and the Dynamic Fluid Body Interaction (DFBI) module technology to build a numerical wave pool. Then, it adopts a cross-wave simulation approach to study the motion response of the SSS in cross-waves with different speeds and wave steepnesses. Based on the obtained data, the CNN-Bi-LSTM neural network is used to predict the resistance and motion response of the SSS. The model has good prediction accuracy and can provide a decision-making basis for the future design of the SSS and its sailing in cross-waves.