Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM
Analyzing and predicting abnormal vibrations in optical fiber composite overhead ground wire (OPGW) transmission lines accurately is a challenging task. This paper proposes a distributed monitoring and forewarning method for OPGW abnormal vibrations using the long short-term memory (LSTM) algorithm,...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIP Publishing LLC
2025-02-01
|
| Series: | AIP Advances |
| Online Access: | http://dx.doi.org/10.1063/5.0249673 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850027071078662144 |
|---|---|
| author | Tianlong Bu Hanpeng Kou Dapei Zhang Zhenhua Feng Helen Law Bin Wang |
| author_facet | Tianlong Bu Hanpeng Kou Dapei Zhang Zhenhua Feng Helen Law Bin Wang |
| author_sort | Tianlong Bu |
| collection | DOAJ |
| description | Analyzing and predicting abnormal vibrations in optical fiber composite overhead ground wire (OPGW) transmission lines accurately is a challenging task. This paper proposes a distributed monitoring and forewarning method for OPGW abnormal vibrations using the long short-term memory (LSTM) algorithm, leveraging the regularity of abnormal vibrations related to climatic conditions. A distributed fiber Bragg grating array is employed to acquire monitoring signals, followed by the derivation of LSTM prediction steps. We effectively capture the long-term dependence of OPGW abnormal vibration signals by introducing cell state and gating mechanisms. In addition, the abnormal vibration forewarning method is analyzed by correlating predicted data with historical data. Experimental results in Hulunbuir demonstrate that the LSTM algorithm performs well in predictions over a 22-h period, evidenced by a root mean square error of 0.8729 and a determination coefficient (R2) of 0.9938 for the fitting curve with actual results. This performance surpasses that of the traditional GA-BP algorithm, facilitating effective abnormal vibration forewarning. This method holds significant potential for widespread application in the field of OPGW abnormal vibration engineering. |
| format | Article |
| id | doaj-art-ec8fc31424d843f98b4919fc208087f8 |
| institution | DOAJ |
| issn | 2158-3226 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | AIP Publishing LLC |
| record_format | Article |
| series | AIP Advances |
| spelling | doaj-art-ec8fc31424d843f98b4919fc208087f82025-08-20T03:00:21ZengAIP Publishing LLCAIP Advances2158-32262025-02-01152025034025034-710.1063/5.0249673Distributed OPGW abnormal vibration monitoring and forewarning based on LSTMTianlong Bu0Hanpeng Kou1Dapei Zhang2Zhenhua Feng3Helen Law4Bin Wang5State Grid Inner Mongolia Eastern Power Co., Ltd. Hulunbuir Power Supply Company, Hulunbuir 021000, Inner Mongolia, ChinaState Grid Inner Mongolia Eastern Power Co., Ltd. Hulunbuir Power Supply Company, Hulunbuir 021000, Inner Mongolia, ChinaState Grid Inner Mongolia Eastern Power Co., Ltd. Hulunbuir Power Supply Company, Hulunbuir 021000, Inner Mongolia, ChinaState Grid Inner Mongolia Eastern Power Co., Ltd. Hulunbuir Power Supply Company, Hulunbuir 021000, Inner Mongolia, ChinaWuhan Kangpu Evergreen Software Technology Co., Ltd., Wuhan 430000, Hubei, ChinaWuhan Kangpu Evergreen Software Technology Co., Ltd., Wuhan 430000, Hubei, ChinaAnalyzing and predicting abnormal vibrations in optical fiber composite overhead ground wire (OPGW) transmission lines accurately is a challenging task. This paper proposes a distributed monitoring and forewarning method for OPGW abnormal vibrations using the long short-term memory (LSTM) algorithm, leveraging the regularity of abnormal vibrations related to climatic conditions. A distributed fiber Bragg grating array is employed to acquire monitoring signals, followed by the derivation of LSTM prediction steps. We effectively capture the long-term dependence of OPGW abnormal vibration signals by introducing cell state and gating mechanisms. In addition, the abnormal vibration forewarning method is analyzed by correlating predicted data with historical data. Experimental results in Hulunbuir demonstrate that the LSTM algorithm performs well in predictions over a 22-h period, evidenced by a root mean square error of 0.8729 and a determination coefficient (R2) of 0.9938 for the fitting curve with actual results. This performance surpasses that of the traditional GA-BP algorithm, facilitating effective abnormal vibration forewarning. This method holds significant potential for widespread application in the field of OPGW abnormal vibration engineering.http://dx.doi.org/10.1063/5.0249673 |
| spellingShingle | Tianlong Bu Hanpeng Kou Dapei Zhang Zhenhua Feng Helen Law Bin Wang Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM AIP Advances |
| title | Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM |
| title_full | Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM |
| title_fullStr | Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM |
| title_full_unstemmed | Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM |
| title_short | Distributed OPGW abnormal vibration monitoring and forewarning based on LSTM |
| title_sort | distributed opgw abnormal vibration monitoring and forewarning based on lstm |
| url | http://dx.doi.org/10.1063/5.0249673 |
| work_keys_str_mv | AT tianlongbu distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm AT hanpengkou distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm AT dapeizhang distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm AT zhenhuafeng distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm AT helenlaw distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm AT binwang distributedopgwabnormalvibrationmonitoringandforewarningbasedonlstm |