摘要
Partial discharge (PD) diagnosis is a momentous measure to extend the service life of gas insulated switchgear (GIS). To fully mine the temporal feature and global spatial topological feature contained in PD and improve the performance of GIS PD diagnosis, a novel method based on spatiotemporal information fusion is proposed. Firstly, the deep features of the original PD signal are mined by convolutional neural network (CNN). On the one hand, it is input into the bidirectional attention long short-term memory network to further mine the temporal features. On the other hand, it is used as a graph node, and the graph data is obtained after the adjacency matrix is formed by K nearest neighbor algorithm matching, and then input into the graph CNN to mine the spatial topological information. Then, the improved D-S evidence theory is employed to fuse the spatiotemporal feature evidence, so as to fully mine the richer implicit information in the PD spectrogram. Results display that our method has obvious merits over the traditional single feature extraction method, and fully utilize the luxuriant discriminable feature in original PD signal.
| 源语言 | 英语 |
|---|---|
| 主期刊名 | 2025 IEEE 8th International Electrical and Energy Conference, CIEEC 2025 |
| 出版商 | Institute of Electrical and Electronics Engineers Inc. |
| 页 | 858-863 |
| 页数 | 6 |
| ISBN(电子版) | 9798331542979 |
| DOI | |
| 出版状态 | 已出版 - 2025 |
| 活动 | 8th IEEE International Electrical and Energy Conference, CIEEC 2025 - Changsha, 中国 期限: 16 5月 2025 → 18 5月 2025 |
丛书
| 姓名 | 2025 IEEE 8th International Electrical and Energy Conference, CIEEC 2025 |
|---|
会议
| 会议 | 8th IEEE International Electrical and Energy Conference, CIEEC 2025 |
|---|---|
| 国家/地区 | 中国 |
| 市 | Changsha |
| 时期 | 16/05/25 → 18/05/25 |
联合国可持续发展目标
此成果有助于实现下列可持续发展目标:
-
可持续发展目标 7 经济适用的清洁能源
学术指纹
探究 'A novel GIS Partial Discharge Diagnosis Method Based on Spatiotemporal Information Fusion and Improved D-S Evidence Theory' 的科研主题。它们共同构成独一无二的学术指纹。引用此
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver