Modeling of Coupling Effects in Neural Networks for Ship Motion Prediction
preprintposted on 15.07.2021, 04:20 by Chong Zhang, Peida Hu, Rong Zhang, Qiuping Wu, Anlan Yang
This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.
In this article, a novel NN structure that combines factorization machines, wavelet transform, and the RNN structure is proposed to improve the accuracy of ship motion prediction. In comparison with traditional NN methods, the proposed structure explicitly models the coupling effects in ship motions, thereby significantly improving the performance of RNN-based models. The proposed structure also provides interpretability of the coupling effects, which helps build white-box NN models for ship motion prediction. Moreover, the proposed structure is lightweight enough for real-time prediction while remaining frequency-aware and sequence-aware.