Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios
Fast reactive power optimization policy-making for various operating scenarios is an important part of power system dispatch. Existing reinforcement learning algorithms alleviate the computational complexity in optimization but suffer from the inefficiency of model retraining for different operating...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | International Journal of Electrical Power & Energy Systems |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0142061524005994 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832595397303861248 |
|---|---|
| author | Congbo Bi Di Liu Lipeng Zhu Chao Lu Shiyang Li Yingqi Tang |
| author_facet | Congbo Bi Di Liu Lipeng Zhu Chao Lu Shiyang Li Yingqi Tang |
| author_sort | Congbo Bi |
| collection | DOAJ |
| description | Fast reactive power optimization policy-making for various operating scenarios is an important part of power system dispatch. Existing reinforcement learning algorithms alleviate the computational complexity in optimization but suffer from the inefficiency of model retraining for different operating scenarios. To solve the above problems, this paper raises a data-efficient transfer reinforcement learning-based reactive power optimization framework. The proposed framework transfers knowledge through two phases: generic state representation in the original scenario and specific dynamic learning in multiple target scenarios. A Q-network structure that separately extracts state and action dynamics is designed to learn generalizable state representations and enable generic knowledge transfer. Supervised learning is applied in specific dynamic learning for extracting unique dynamics from offline data, which improves data efficiency and speeds up knowledge transfer. Finally, the proposed framework is tested on the IEEE 39-bus system and the realistic Guangdong provincial power grid, demonstrating its effectiveness and reliability. |
| format | Article |
| id | doaj-art-5830f1afbde14059980bb7cf88aaa04d |
| institution | Kabale University |
| issn | 0142-0615 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | International Journal of Electrical Power & Energy Systems |
| spelling | doaj-art-5830f1afbde14059980bb7cf88aaa04d2025-01-19T06:23:52ZengElsevierInternational Journal of Electrical Power & Energy Systems0142-06152025-03-01164110376Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenariosCongbo Bi0Di Liu1Lipeng Zhu2Chao Lu3Shiyang Li4Yingqi Tang5Department of Electrical Engineering, Tsinghua University, Haidian District, 100084, Beijing, ChinaDepartment of Electrical Engineering, Tsinghua University, Haidian District, 100084, Beijing, ChinaCollege of Electrical and Information Engineering, Hunan University, Changsha, 410082, Hunan, ChinaDepartment of Electrical Engineering, Tsinghua University, Haidian District, 100084, Beijing, China; Corresponding author.China Southern Grid Electric Power Research Institute, China Southern Grid, Guangzhou, 510663, Guangdong, ChinaChina Southern Grid Electric Power Research Institute, China Southern Grid, Guangzhou, 510663, Guangdong, ChinaFast reactive power optimization policy-making for various operating scenarios is an important part of power system dispatch. Existing reinforcement learning algorithms alleviate the computational complexity in optimization but suffer from the inefficiency of model retraining for different operating scenarios. To solve the above problems, this paper raises a data-efficient transfer reinforcement learning-based reactive power optimization framework. The proposed framework transfers knowledge through two phases: generic state representation in the original scenario and specific dynamic learning in multiple target scenarios. A Q-network structure that separately extracts state and action dynamics is designed to learn generalizable state representations and enable generic knowledge transfer. Supervised learning is applied in specific dynamic learning for extracting unique dynamics from offline data, which improves data efficiency and speeds up knowledge transfer. Finally, the proposed framework is tested on the IEEE 39-bus system and the realistic Guangdong provincial power grid, demonstrating its effectiveness and reliability.http://www.sciencedirect.com/science/article/pii/S0142061524005994Reactive power optimizationTransfer reinforcement learningData efficiencyPower system |
| spellingShingle | Congbo Bi Di Liu Lipeng Zhu Chao Lu Shiyang Li Yingqi Tang Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios International Journal of Electrical Power & Energy Systems Reactive power optimization Transfer reinforcement learning Data efficiency Power system |
| title | Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| title_full | Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| title_fullStr | Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| title_full_unstemmed | Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| title_short | Reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| title_sort | reactive power optimization via deep transfer reinforcement learning for efficient adaptation to multiple scenarios |
| topic | Reactive power optimization Transfer reinforcement learning Data efficiency Power system |
| url | http://www.sciencedirect.com/science/article/pii/S0142061524005994 |
| work_keys_str_mv | AT congbobi reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios AT diliu reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios AT lipengzhu reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios AT chaolu reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios AT shiyangli reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios AT yingqitang reactivepoweroptimizationviadeeptransferreinforcementlearningforefficientadaptationtomultiplescenarios |