A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion
Abstract As a basic task in energy consumption monitoring system, load forecasting has great effects on system operation safety, generation costs and economic benefits. In this paper, a long-term load forecasting algorithm using data dimension expansion and deep feature extraction is proposed. First...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2023-10-01
|
| Series: | EURASIP Journal on Advances in Signal Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13634-023-01068-1 |
| _version_ | 1797451069137092608 |
|---|---|
| author | Wei-guo Zhang Qing Zhu Lin-Lin Gu Hui-Jie Lin |
| author_facet | Wei-guo Zhang Qing Zhu Lin-Lin Gu Hui-Jie Lin |
| author_sort | Wei-guo Zhang |
| collection | DOAJ |
| description | Abstract As a basic task in energy consumption monitoring system, load forecasting has great effects on system operation safety, generation costs and economic benefits. In this paper, a long-term load forecasting algorithm using data dimension expansion and deep feature extraction is proposed. First, the outliers of the meteorological measurements are removed by median filter method, and then the time information is encoded to form the fingerprint of the training data. Next, the full connected network (FCN) is used to expand the dimensions of the fingerprint, and the convolutional neural network (CNN) is used to extract the deep features which can obtain better feature representation. Finally, the FCN, the CNN and regression learning model are combined for jointly offline training. The optimal parameters of these network can be obtained under global solution. Experimental results show that the proposed algorithm has better load forecasting performance than existing methods. |
| first_indexed | 2024-03-09T14:49:37Z |
| format | Article |
| id | doaj.art-b4b62cfd3ee14557acdca0c47fee2364 |
| institution | Directory Open Access Journal |
| issn | 1687-6180 |
| language | English |
| last_indexed | 2024-03-09T14:49:37Z |
| publishDate | 2023-10-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | EURASIP Journal on Advances in Signal Processing |
| spelling | doaj.art-b4b62cfd3ee14557acdca0c47fee23642023-11-26T14:33:12ZengSpringerOpenEURASIP Journal on Advances in Signal Processing1687-61802023-10-012023111310.1186/s13634-023-01068-1A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansionWei-guo Zhang0Qing Zhu1Lin-Lin Gu2Hui-Jie Lin3Southeast UniversityNari Technology Co., Ltd.Nari Technology Co., Ltd.Nari Technology Co., Ltd.Abstract As a basic task in energy consumption monitoring system, load forecasting has great effects on system operation safety, generation costs and economic benefits. In this paper, a long-term load forecasting algorithm using data dimension expansion and deep feature extraction is proposed. First, the outliers of the meteorological measurements are removed by median filter method, and then the time information is encoded to form the fingerprint of the training data. Next, the full connected network (FCN) is used to expand the dimensions of the fingerprint, and the convolutional neural network (CNN) is used to extract the deep features which can obtain better feature representation. Finally, the FCN, the CNN and regression learning model are combined for jointly offline training. The optimal parameters of these network can be obtained under global solution. Experimental results show that the proposed algorithm has better load forecasting performance than existing methods.https://doi.org/10.1186/s13634-023-01068-1Data dimension expansionDeep learningFeature extractionLoad forecasting |
| spellingShingle | Wei-guo Zhang Qing Zhu Lin-Lin Gu Hui-Jie Lin A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion EURASIP Journal on Advances in Signal Processing Data dimension expansion Deep learning Feature extraction Load forecasting |
| title | A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| title_full | A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| title_fullStr | A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| title_full_unstemmed | A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| title_short | A deep learning-based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| title_sort | deep learning based load forecasting algorithm for energy consumption monitoring system using dimension expansion |
| topic | Data dimension expansion Deep learning Feature extraction Load forecasting |
| url | https://doi.org/10.1186/s13634-023-01068-1 |
| work_keys_str_mv | AT weiguozhang adeeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT qingzhu adeeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT linlingu adeeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT huijielin adeeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT weiguozhang deeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT qingzhu deeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT linlingu deeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion AT huijielin deeplearningbasedloadforecastingalgorithmforenergyconsumptionmonitoringsystemusingdimensionexpansion |