한국 해역 관측 데이터를 이용한 앙상블 학습 기반 파고 예측
Wave Height Prediction based on Ensemble Learning Using Observed Data from South Korea
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
파고 예측은 해양 분야에서 매우 중요하며, 자율운항 선박의 최적 경로 설정, 파력 발전기 설치 및 운영, 해양 플랫폼 설치 등 여러 분야에 폭넓게 활용되고 있다. 과거에는 물리적 원리에 기초한 수치 모델을 기반으로 파고를 예측했으나, 이 방법은 많은 계산 자원이 필요하다는 단점이 있다. 이러한 단점을 극복하기 위해 최근에는 기계학습 방법을 적용하여 파고를 예측하는 연구가 이루어지고 있다. 본 연구에서는 기계학습 방법 중 앙상블 학습을 이용하여 해양 관측소에서 수집된 데이터를 기반으로 파고를 예측하였다. 상관관계 분석을 통해 모델의 입력 변수를 결정한 후 앙상블 기법 중 배깅 기법과 부스팅 기법을 적용한 모델들을 적용하여 성능을 비교하였다. 그 결과, 앙상블 학습 모델 중 배깅 모델보다 부스팅 모델의 성능이 더 좋은 것을 확인하였다. 본 연구는 단일 관측소의 데이터를 기반으로 한 실험이라는 한계가 있다. 따라서, 향후에는 여러 해역의 관측소 데이터를 활용하여 제안한 모델의 보편성을 검증하는 연구를 수행할 계획이다.
Abstract
Wave height prediction is crucial in the marine sector and has wide-ranging applications, including determining the optimal route for autonomous ships, installing and operating wave power generators, and deploying offshore platforms. In the past, wave prediction was primarily based on numerical models grounded in physical principles; however, these methods require extensive computational resources. Recently, many researchers have applied machine learning methods to overcome these limitations. In this study, we utilized ensemble learning, a type of machine learning approach, to predict wave heights based on data collected from ocean observation stations in Korea. After determining the input variables of the model through correlation analysis, we applied and compared models using both bagging and boosting ensemble methods. The results indicated that, among the ensemble learning models, the boosting models outperformed the bagging models. This study is limited by its reliance on data from a single observation station. Therefore, future work will involve conducting research using data from multiple observation stations across various sea areas to validate the generalizability of the proposed model.
참고문헌 (27)
- Akiba, T., "Optuna: A Next-generation Hyperparameter Optimization Framework", Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.2623~2631 (2019)
- Berbić, J., "Application of neural networks and support vector machine for significant wave height prediction", Oceanologia, 59(3), pp.331~349 (2017)
- Booij, N. R. R. C., "A third-generation wave model for coastal regions 1. Model description and validation", Journal of Geophysical Research : Oceans, 104(4), pp.7649~7666 (1999)
- Breiman, L., "Random forests", Machine Learning, 45(10), pp.5~32 (2001)
- Chen, T., "XGBoost : A scalable tree boosting system", Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp.785~794 (2016)
- Dietterich, T. G., "Ensemble methods in machine learning", International workshop on multiple classifier systems, 1857, pp.1~15 (2000)
- Feng, X., "A multi-layer perceptron approach for accelerated wave forecasting in Lake Michigan", Ocean Engineering, 211(9), pp.1~11 (2020)
- Friedman, J. H., "Greedy function approximation : a gradient boosting machine", The Annals of Statistics, 29(5), pp.1189~1232 (2001)
- Fushiki, T., "Estimation of prediction error by using K-fold cross-validation", Statistics and Computing, pp.137~146 (2011)
- Gang, Y. S., "Design of Ship-type Floating LiDAR Buoy System for Wind Resource Measurement in the Korean West Sea and Numerical Analysis of Stability Assessment of Mooring System", Journal of Korean Navigation and Port Research, 46(6), pp.483~490 (2022)
- Gao, R., "Significant wave height forecasting using hybrid ensemble deep randomized networks with neurons pruning", Engineering Applications of Artificial Intelligence, 117, pp.105535~ (2023)
- Gracia, S., "Improving accuracy on wave height estimation through machine learning techniques", Ocean Engineering, 236, pp.805~817 (2021)
- Hasan, A., "Generalized machine learning models to predict significant wave height utilizing wind and atmospheric parameters", Energy Conversion and Management, 23, pp.100623~ (2024)
- Huang, W., "A training strategy to improve the generalization capability of deep learning-based significant wave height prediction models in offshore China", Ocean Engineering, 283(9), pp.1~21 (2023)
- Janssen, P. A., "Progress in ocean wave forecasting", Journal of Computational Physics, 227(7), pp.3572~3594 (2008)
- Jeong, S., "The Study on the Effects of Breakwater Energy Conversion System by Horizontal Plate Installation", Journal of Korean Navigation and Port Research, 38(1), pp.39~44 (2014)
- Jung, S. J., "Experimental and Numerical Study of Berthing and Unberthing of LNG-Bunkering Vessel", Journal of Korean Navigation and Port Research, 44(6), pp.439~446 (2020)
- Jung, S. J., "Numerical Assessment of LNGC Berthing Operation to FLBT", Journal of Korean Navigation and Port Research, 45(3), pp.87~94 (2021)
- Karabulut, N., "Wave height prediction with single input parameter by using regression methods", Energy Sources, Part A : Recovery, Utilization, and Environmental Effects, 42(24), pp.2972~2989 (2020)
- Ke, G., "LightGBM : A Highly Efficient Gradient Boosting Decision Tree", Advances in Neural Information Processing Systems, 30, pp.1~9 (2017)
- Kim, D., "Optimal Route Planning for Maritime Autonomous Surface Ships Using a Nonlinear Model Predictive Control", Journal of Korean Navigation and Port Research, 47(2), pp.66~74 (2023)
- Kim, J., "STG-OceanWaveNet : Spatio-temporal geographic information guided ocean wave prediction network", Ocean Engineering, 257(8), pp.1~13 (2022)
- "Ocean Data in Grid Framework" (2024)[URL]
- Lou, R., "Prediction of Ocean Wave Height Suitable for Ship Autopilot", IEEE Transactions on Intelligent Transportation Systems, 23(12), pp.25557~25566 (2022)
- Shamshirband, S., "Prediction of significant wave height; comparison between nested grid numerical model, and machine learning models of artificial neural networks, extreme learning and support vector machines", Engineering Applications of Computational Fluid Mechanics, 14(1), pp.805~817 (2020)
- Tolman, H., "User manual and system documentation of WAVEWATCH III TM version 3. 14", MMAB Contribution, 1~220 (2009)
- WAMDI, "The WAM Model—A Third Generation Ocean Wave Prediction Model", Journal of Computational Physics, pp.1775~1810 (1988)