Session: 08-06-01 non-presentations

Paper Number: 135752

135752 - Prediction of ship rolling motion in waves based on deep operator network 

The roll motion of the ship in the wave is one of the important factors affecting the safety of the ships. Due to the nonlinearity of the roll motion, it is difficult to predict it accurately. With the development of intelligent ships, real-time prediction of roll motion in wave environment is one of the current problems. DeepOnet is a functional prediction deep learning model, which uses wave height data as the input of the branch network and ship roll motion data as the input of the trunk network to couple the correlation between wave and ship roll motion. The model structure has a dual network structure, and at the same time, the wave height and hull motion are used as inputs, and the subsequent hull motion is output. In this paper, based on the experimental data of ship model 5512, the roll prediction under regular wave and irregular wave conditions are carried out using the model DeepOnet, and compared with the classical time series prediction model LSTM. The forecast model is divided into single-condition model and multi-condition model, single-step prediction and multi-step prediction. The results show that the prediction accuracy of the DeepOnet model has obvious advantages. DeepOnet model may have an optimistic application prospect for real-time and accurate roll motion  prediction in the wave environment for ships.

Presenting Author: Yong Zhao Dalian Maritime University

Presenting Author Biography: Dr. Yong Zhao is an associate professor at the School of Ship and Ocean Engineering at Dalian Maritime University. His research interests include ship hydrodynamic analysis, artificial intelligence methods for ship and ocean engineering, wavelet analysis, and more. Dr. Yong Zhao has completed two National Natural Science Foundation projects and published over 30 papers in academic journals. He visited the Department of Mathematics at Brown University from August 2019 to March 2023 to conduct research on machine learning methods and applications.

Authors:

Yong Zhao Dalian Maritime University
Jinxiu Zhao Dalian Maritime University