Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder
The accurate and robust prediction of short-term solar power generation is significant for the management of modern smart grids, where solar power has become a major energy source due to its green and economical nature. However, the solar yield prediction can be difficult to conduct in the real worl...
| Autores principales: | , , , , |
|---|---|
| Otros Autores: | |
| Formato: | Conference Paper |
| Lenguaje: | English |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | https://hdl.handle.net/10356/152999 |
| _version_ | 1826122862527774720 |
|---|---|
| author | Shen, Meng Zhang, Huaizheng Cao, Yixin Yang, Fan Wen, Yonggang |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Shen, Meng Zhang, Huaizheng Cao, Yixin Yang, Fan Wen, Yonggang |
| author_sort | Shen, Meng |
| collection | NTU |
| description | The accurate and robust prediction of short-term solar power generation is significant for the management of modern smart grids, where solar power has become a major energy source due to its green and economical nature. However, the solar yield prediction can be difficult to conduct in the real world where hardware and network issues can make the sensors unreachable. Such data missing problem is so prevalent that it degrades the performance of deployed prediction models and even fails the model execution. In this paper, we propose a novel temporal multi-modal variational auto-encoder (TMMVAE) model, to enhance the robustness of short-term solar power yield prediction with missing data. It can impute the missing values in time-series sensor data, and reconstruct them by consolidating multi-modality data, which then facilitates more accurate solar power yield prediction. TMMVAE can be deployed efficiently with an end-to-end framework. The framework is verified at our real-world testbed on campus. The results of extensive experiments show that our proposed framework can significantly improve the imputation accuracy when the inference data is severely corrupted, and can hence dramatically improve the robustness of short-term solar energy yield forecasting. |
| first_indexed | 2024-10-01T05:55:11Z |
| format | Conference Paper |
| id | ntu-10356/152999 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T05:55:11Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | ntu-10356/1529992021-10-28T05:20:04Z Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder Shen, Meng Zhang, Huaizheng Cao, Yixin Yang, Fan Wen, Yonggang School of Computer Science and Engineering 29th ACM International Conference on Multimedia Engineering::Computer science and engineering::Computing methodologies Solar Forecasting Multimodal Learning The accurate and robust prediction of short-term solar power generation is significant for the management of modern smart grids, where solar power has become a major energy source due to its green and economical nature. However, the solar yield prediction can be difficult to conduct in the real world where hardware and network issues can make the sensors unreachable. Such data missing problem is so prevalent that it degrades the performance of deployed prediction models and even fails the model execution. In this paper, we propose a novel temporal multi-modal variational auto-encoder (TMMVAE) model, to enhance the robustness of short-term solar power yield prediction with missing data. It can impute the missing values in time-series sensor data, and reconstruct them by consolidating multi-modality data, which then facilitates more accurate solar power yield prediction. TMMVAE can be deployed efficiently with an end-to-end framework. The framework is verified at our real-world testbed on campus. The results of extensive experiments show that our proposed framework can significantly improve the imputation accuracy when the inference data is severely corrupted, and can hence dramatically improve the robustness of short-term solar energy yield forecasting. Energy Market Authority (EMA) Ministry of Education (MOE) National Research Foundation (NRF) This work is supported by the National Research Foundation, Singapore, the Energy Market Authority, under its Energy Programme (EP Award <NRF2017EWT-EP003-023>), and MOE under its grant call (RG96/20). 2021-10-28T05:20:03Z 2021-10-28T05:20:03Z 2021 Conference Paper Shen, M., Zhang, H., Cao, Y., Yang, F. & Wen, Y. (2021). Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder. 29th ACM International Conference on Multimedia, 2558-2566. https://dx.doi.org/10.1145/3474085.3475430 9781450386517 https://hdl.handle.net/10356/152999 10.1145/3474085.3475430 2558 2566 en NRF2017EWT-EP003-023 RG96/20 © 2021 Association for Computing Machinery. All rights reserved. |
| spellingShingle | Engineering::Computer science and engineering::Computing methodologies Solar Forecasting Multimodal Learning Shen, Meng Zhang, Huaizheng Cao, Yixin Yang, Fan Wen, Yonggang Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title | Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title_full | Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title_fullStr | Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title_full_unstemmed | Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title_short | Missing data imputation for solar yield prediction using temporal multi-modal variational auto-encoder |
| title_sort | missing data imputation for solar yield prediction using temporal multi modal variational auto encoder |
| topic | Engineering::Computer science and engineering::Computing methodologies Solar Forecasting Multimodal Learning |
| url | https://hdl.handle.net/10356/152999 |
| work_keys_str_mv | AT shenmeng missingdataimputationforsolaryieldpredictionusingtemporalmultimodalvariationalautoencoder AT zhanghuaizheng missingdataimputationforsolaryieldpredictionusingtemporalmultimodalvariationalautoencoder AT caoyixin missingdataimputationforsolaryieldpredictionusingtemporalmultimodalvariationalautoencoder AT yangfan missingdataimputationforsolaryieldpredictionusingtemporalmultimodalvariationalautoencoder AT wenyonggang missingdataimputationforsolaryieldpredictionusingtemporalmultimodalvariationalautoencoder |