Journal of Navigation and Port Research (한국항해항만학회지)

Korean Institute of Navigation and Port Research (한국항해항만학회)
권호 표지
DOI QR코드

DOI QR Code

Fishing Boat Rolling Movement of Time Series Prediction based on Deep Network Model

심층 네트워크 모델에 기반한 어선 횡동요 시계열 예측

  • Donggyun Kim (Division of Navigation Science, Mokpo National Maritime University) ;
  • Nam-Kyun Im (Division of Navigation Science, Mokpo National Maritime University)
  • 김동균 (목포해양대학교 항해학부) ;
  • 임남균 (목포해양대학교 항해학부)
  • Received : 2023.10.04
  • Accepted : 2023.11.08
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Fishing boat capsizing accidents account for more than half of all capsize accidents. These can occur for a variety of reasons, including inexperienced operation, bad weather, and poor maintenance. Due to the size and influence of the industry, technological complexity, and regional diversity, fishing ships are relatively under-researched compared to commercial ships. This study aimed to predict the rolling motion time series of fishing boats using an image-based deep learning model. Image-based deep learning can achieve high performance by learning various patterns in a time series. Three image-based deep learning models were used for this purpose: Xception, ResNet50, and CRNN. Xception and ResNet50 are composed of 177 and 184 layers, respectively, while CRNN is composed of 22 relatively thin layers. The experimental results showed that the Xception deep learning model recorded the lowest Symmetric mean absolute percentage error(sMAPE) of 0.04291 and Root Mean Squared Error(RMSE) of 0.0198. ResNet50 and CRNN recorded an RMSE of 0.0217 and 0.022, respectively. This confirms that the models with relatively deeper layers had higher accuracy.

통계에 따르면 어선의 전복 사고는 전체 전복 사고의 절반 이상을 차지한다. 이는 미숙한 조업, 기상 악화, 정비 미흡 등 다양한 원인으로 발생할 수 있다. 업계 규모와 영향도, 기술 복잡성, 지역적 다양성 등으로 인해 어선은 상선에 비해 상대적으로 연구가 부족한 실정이다. 본 연구에서는 이미지 기반 딥러닝 모델을 활용하여 어선의 횡동요 시계열을 예측하고자 한다. 이미지 기반 딥러닝은 시계열의 다양한 패턴을 학습하여 높은 성능을 낼 수 있다. 이를 위해 Xception, ResNet50, CRNN의 3가지의 이미지 기반 딥러닝 모델을 활용하였다. Xception과 ResNet50은 각각 177, 184개의 층으로 구성되어 있으며 이에 반해 CRNN은 22개의 비교적 얇은 층으로 구성되어 있다. 실험 결과 Xception 딥러닝 모델이 가장 낮은 0.04291의 sMAPE와 0.0198의 RMSE를 기록하였다. ResNet50과 CRNN은 각각 0.0217, 0.022의 RMSE를 기록하였다. 이를 통해 상대적으로 층이 더 깊은 모델의 정확도가 높음을 확인할 수 있다.

Keywords

Acknowledgement

This research was supported by Korea Institute of Marine Science & Technology Promotion (KIMST) funded by the Ministry of Oceans and Fisheries (RS-2023-00238653).

References

  1. Alvarellos, A., Figuero, A., Carro, H., Costas, R., Sande, J., Guerra, A., Pena, E. and Rabunal J.(2021)"Machine Learning Based Moored Ship Movement Prediction. Journal of Marine Science and Engineering", Vol. 9, No. 8.
  2. Ashish, V., Noam, S., Niki, P., Jakob, U.,Llion, J., Aidan, G., Kaiser, L. and Illia, P.(2017), "Attention is All you Need", Advances in Neural Information Processing Systems, Vol. 30.
  3. BBC News Korea(2023), "Fishing boat capsize accident", https://www.bbc.com/korean/news-64527530
  4. Cho, K., Bart, V. M., Caglar, G., Dzmitry, B., Fethi, B., Holger, S. and Yoshua, B.(2014), "Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation", Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1724-1734.
  5. Chollet, F.(2017), "Xception: Deep Learning With Depthwise Separable Convolutions", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1251-1258.
  6. El M. S., Benabbou L., Caron, S. and Berrado A.(2023), "Deep Learning-Based Ship Speed Prediction for Intelligent Maritime Traffic Management", Journal of Marine Science and Engineering, Vol. 11, No. 1.
  7. He, K., Zhang, X., Ren, S. and Sun, J.(2016), "Deep Residual Learning for Image Recognition", 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, pp. 770-778.
  8. Hochreiter, S. and Schmidhuber J.(1997), "Long Short-term Memory", Neural Computation, Vol. 9, No. 8, pp. 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  9. Hoseinzade, E. and Haratizadeh, S.(2019), "CNNpred: CNN-based stock market prediction using a diverse set of variables", Expert Systems with Applications, Vol. 129, No. 1, pp. 273-285. https://doi.org/10.1016/j.eswa.2019.03.029
  10. Kim, Y., Park, J. and Moon, S.(2018), "Prediction of Ship Roll Motion using Machine Learning-based Surrogate Model", Journal of Navigation and Port Research, Vol. 42, No. 6, pp. 395, 405.
  11. Kim, H., Kim, J. and Lee, Y.(2020), "A Study on Securing a Stable GM for Each Ship Type Considering the Ship's Operating Status", Vol. 44, No. 4, pp. 275-282. https://doi.org/10.5394/KINPR.2020.44.4.275
  12. Kim, M. and Lee, H.(2023), "A Study on Ship Route Generation with Deep Q Network and Route Following Control", Journal of Navigation and Port Research, Vol. 47, No. 2, pp. 75-84. https://doi.org/10.5394/KINPR.2023.47.2.75
  13. Kim, H., Kim, K., Hwang S. and Lee, J. H.(2022), "The Fault Diagnosis Model of Ship Fuel System Equipment Reflecting Time Dependency in Conv1D Algorithm Based on the Convolution Network", Journal of Navigation and Port Research, Vol. 46, No. 4, pp. 367-374. https://doi.org/10.5394/KINPR.2022.46.4.367
  14. Kim, Y., Shin, J. Y. and Park, H. J.(2022), "A Study on the Prediction of Gate In-Out Truck Waiting Time in the Container Terminal", Journal of Navigation and Port Research, Vol. 46, No. 4, pp. 344-350. https://doi.org/10.5394/KINPR.2022.46.4.344
  15. Korean Maritime Safety Tribunal(2023), https://wwwkmst.go.kr/web/stcAnnualReport.do?menuIdx=126.
  16. Lin, J., Han, Y., Guo, C., Su, Y. and Zhong, R. (2022), "Intelligent ship anti-rolling control system based on a deep deterministic policy gradient algorithm and the Magnus effect", Physics of Fluids, Vol. 34, No. 5.
  17. Suhermi, N., Suhartono, Prastyo, D. D. and Ali, B.(2018), "Roll motion prediction using a hybrid deep learning and ARIMA model", Procedia Computer Science, Vol. 144, pp. 251-258. https://doi.org/10.1016/j.procs.2018.10.526
  18. Yao, G., Lei, T. and Zhong, J(2019). "A review of convolutional-neural-network-based action recognition", Pattern Recogn Lett, Vol. 118, pp. 14-22. https://doi.org/10.1016/j.patrec.2018.05.018
  19. Zhou, X., Li, H., Huang, Y. and Liu, Y.(2023), "Deep learning machine based ship parametric rolling simulation and recognition algorithms", Ocean Engineering, Vol. 276.
  20. Wibawa, A. P., Utama, A. B., Elmunsyah, H., Pujianto, U., Dwiyanto, F. and Hernandez, L.(2022), "Time-series analysis with smoothed Convolutional Neural Network", Journal of Big Data, Vol. 9, No. 44.