Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information
Predicting wind speed over the ocean is difficult due to the unequal distribution of buoy stations and the occasional fluctuations in the wind field. This study proposes a dynamic graph embedding-based graph neural network—long short-term memory joint framework (DGE-GAT-LSTM) to estimate wind speed...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Journal of Marine Science and Engineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-1312/12/3/502 |
| _version_ | 1797240387489759232 |
|---|---|
| author | Dibo Dong Shangwei Wang Qiaoying Guo Yiting Ding Xing Li Zicheng You |
| author_facet | Dibo Dong Shangwei Wang Qiaoying Guo Yiting Ding Xing Li Zicheng You |
| author_sort | Dibo Dong |
| collection | DOAJ |
| description | Predicting wind speed over the ocean is difficult due to the unequal distribution of buoy stations and the occasional fluctuations in the wind field. This study proposes a dynamic graph embedding-based graph neural network—long short-term memory joint framework (DGE-GAT-LSTM) to estimate wind speed at numerous stations by considering their spatio-temporal information properties. To begin, the buoys that are pertinent to the target station are chosen based on their geographic position. Then, the local graph structures connecting the stations are represented using cosine similarity at each time interval. Subsequently, the graph neural network captures intricate spatial characteristics, while the LSTM module acquires knowledge of temporal interdependence. The graph neural network and LSTM module are sequentially interconnected to collectively capture spatio-temporal correlations. Ultimately, the multi-step prediction outcomes are produced in a sequential way, where each step relies on the previous predictions. The empirical data are derived from direct measurements made by NDBC buoys. The results indicate that the suggested method achieves a mean absolute error reduction ranging from 1% to 36% when compared to other benchmark methods. This improvement in accuracy is statistically significant. This approach effectively addresses the challenges of inadequate information integration and the complexity of modeling temporal correlations in the forecast of ocean wind speed. It offers valuable insights for optimizing the selection of offshore wind farm locations and enhancing operational and management capabilities. |
| first_indexed | 2024-04-24T18:06:37Z |
| format | Article |
| id | doaj.art-121e18ab4d544c2aab9f9d3f923af0f8 |
| institution | Directory Open Access Journal |
| issn | 2077-1312 |
| language | English |
| last_indexed | 2024-04-24T18:06:37Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Marine Science and Engineering |
| spelling | doaj.art-121e18ab4d544c2aab9f9d3f923af0f82024-03-27T13:49:27ZengMDPI AGJournal of Marine Science and Engineering2077-13122024-03-0112350210.3390/jmse12030502Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal InformationDibo Dong0Shangwei Wang1Qiaoying Guo2Yiting Ding3Xing Li4Zicheng You5Institute of Smart Marine and Engineering, Fujian University of Technology, Fuzhou 350118, ChinaInstitute of Smart Marine and Engineering, Fujian University of Technology, Fuzhou 350118, ChinaInstitute of Smart Marine and Engineering, Fujian University of Technology, Fuzhou 350118, ChinaFinance and Economics College, Jimei University, Xiamen 361021, ChinaMarine Forecasting Center of Fujian Province, Fuzhou 350003, ChinaInstitute of Smart Marine and Engineering, Fujian University of Technology, Fuzhou 350118, ChinaPredicting wind speed over the ocean is difficult due to the unequal distribution of buoy stations and the occasional fluctuations in the wind field. This study proposes a dynamic graph embedding-based graph neural network—long short-term memory joint framework (DGE-GAT-LSTM) to estimate wind speed at numerous stations by considering their spatio-temporal information properties. To begin, the buoys that are pertinent to the target station are chosen based on their geographic position. Then, the local graph structures connecting the stations are represented using cosine similarity at each time interval. Subsequently, the graph neural network captures intricate spatial characteristics, while the LSTM module acquires knowledge of temporal interdependence. The graph neural network and LSTM module are sequentially interconnected to collectively capture spatio-temporal correlations. Ultimately, the multi-step prediction outcomes are produced in a sequential way, where each step relies on the previous predictions. The empirical data are derived from direct measurements made by NDBC buoys. The results indicate that the suggested method achieves a mean absolute error reduction ranging from 1% to 36% when compared to other benchmark methods. This improvement in accuracy is statistically significant. This approach effectively addresses the challenges of inadequate information integration and the complexity of modeling temporal correlations in the forecast of ocean wind speed. It offers valuable insights for optimizing the selection of offshore wind farm locations and enhancing operational and management capabilities.https://www.mdpi.com/2077-1312/12/3/502graph embeddinggraph neural networkspatio-temporal informationwind data |
| spellingShingle | Dibo Dong Shangwei Wang Qiaoying Guo Yiting Ding Xing Li Zicheng You Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information Journal of Marine Science and Engineering graph embedding graph neural network spatio-temporal information wind data |
| title | Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information |
| title_full | Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information |
| title_fullStr | Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information |
| title_full_unstemmed | Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information |
| title_short | Short-Term Marine Wind Speed Forecasting Based on Dynamic Graph Embedding and Spatiotemporal Information |
| title_sort | short term marine wind speed forecasting based on dynamic graph embedding and spatiotemporal information |
| topic | graph embedding graph neural network spatio-temporal information wind data |
| url | https://www.mdpi.com/2077-1312/12/3/502 |
| work_keys_str_mv | AT dibodong shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation AT shangweiwang shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation AT qiaoyingguo shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation AT yitingding shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation AT xingli shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation AT zichengyou shorttermmarinewindspeedforecastingbasedondynamicgraphembeddingandspatiotemporalinformation |