Significant wave height prediction model based on LSTM cell spatiotemporal network

DOI

10.2991/978-94-6463-398-6_64How to use a DOI?

Keywords

significant wave height prediction; deep learning; spatiotemporal network; PredRNN

Abstract

Marine operations, engineering activities, and transportation are highly influenced by sea waves, and accurately predicting the 2D wave field is crucial for ensuring safe and efficient project execution and avoiding the sea wave disaster. While existing deep learning applications for wave height prediction primarily focus on single-point forecasts, it's important to note that single-point data may not capture the overall regional trends effectively. Therefore, this paper establishes two spatiotemporal network models based on LSTM cells for Lianyungang Port and compares their performance. The results demonstrate that PredRNN, with its newly designed spatiotemporal memory cell which is able to deliver memory states through layers and laysers on current node and can learn the short-term non-linear variation, outperforms ConvLSTM, especially when dealing with unbalanced input samples. With the forecasting leading time increasing from 6h to 12h the correlation coefficient of PreRNN is over 0.9, ConvLSTM decreases to 0.849. Under the lower accuracy of input samples condition, PredRNN also performs better. In summary, PredRNN is less affected by the quality of input samples, which has engineering value in significant wave forecasting for marine operations.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Qinze Chen
AU  - Hanghang Lyu
AU  - Jiaming Qu
AU  - Yuchi Hao
PY  - 2024
DA  - 2024/04/24
TI  - Significant wave height prediction model based on LSTM cell spatiotemporal network
BT  - Proceedings of the 2023 5th International Conference on Hydraulic, Civil and Construction Engineering (HCCE 2023)
PB  - Atlantis Press
SP  - 653
EP  - 665
SN  - 2589-4943
UR  - https://doi.org/10.2991/978-94-6463-398-6_64
DO  - 10.2991/978-94-6463-398-6_64
ID  - Chen2024
ER  -