Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm
The classification of voltage sag sources is essential for the establishment of controlling scheme and reasonable division of responsibilities in voltage sag-associated accidents. Existing methods for classifying voltage sag sources usually ignore the interpretability of the classification model, an...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9749094/ |
| _version_ | 1828378703478915072 |
|---|---|
| author | Zhang Yikun He Yingjie Zhang Haixiao Li Jiahao Li Yijin Liu Jinjun |
| author_facet | Zhang Yikun He Yingjie Zhang Haixiao Li Jiahao Li Yijin Liu Jinjun |
| author_sort | Zhang Yikun |
| collection | DOAJ |
| description | The classification of voltage sag sources is essential for the establishment of controlling scheme and reasonable division of responsibilities in voltage sag-associated accidents. Existing methods for classifying voltage sag sources usually ignore the interpretability of the classification model, and are only dedicated to improving the accuracy of voltage classification, which cannot provide a reliable classification basis for users and power enterprises. Therefore, this article proposes an effective and interpretable voltage sag sources classification method based on sequential trajectory feature learning and Random Forest algorithm. Firstly, to fully consider the interpretable sequential trajectory features of voltage sag signals so as to improve the quality and interpretability of calculation process, the fused lasso generalized eigenvector (FLAG) algorithm is adopted to quickly search for interpretable shapelets sub-sequences from the labeled voltage sag data. After that, the labeled data and samples to-be-classified are subjected to shapelet transformation through the shapelet sub-sequences to obtain the sequential trajectory features. Finally, the random forest is trained on the labeled sequential trajectory feature data to achieve supervised sample classification, which inherits the interpretability of shapelet. To test the feasibility and validity of the proposed voltage sag sources classification method, the simulation cases based on the simulated voltage sag signals were studied. The simulation results show that the proposed method has significant advantages in terms of accuracy and interpretability of voltage sag sources classification. |
| first_indexed | 2024-04-14T08:31:58Z |
| format | Article |
| id | doaj.art-1add1013923141fe96fa1a9d8ab87638 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-14T08:31:58Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-1add1013923141fe96fa1a9d8ab876382022-12-22T02:03:54ZengIEEEIEEE Access2169-35362022-01-0110385023851010.1109/ACCESS.2022.31646759749094Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning AlgorithmZhang Yikun0https://orcid.org/0000-0001-8790-8254He Yingjie1https://orcid.org/0000-0002-4869-2507Zhang Haixiao2Li Jiahao3Li Yijin4Liu Jinjun5https://orcid.org/0000-0003-0050-2548School of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaSchool of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaSchool of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaSchool of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaSchool of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaSchool of Electrical Engineering, Xi’an Jiaotong University, Xi’an, ChinaThe classification of voltage sag sources is essential for the establishment of controlling scheme and reasonable division of responsibilities in voltage sag-associated accidents. Existing methods for classifying voltage sag sources usually ignore the interpretability of the classification model, and are only dedicated to improving the accuracy of voltage classification, which cannot provide a reliable classification basis for users and power enterprises. Therefore, this article proposes an effective and interpretable voltage sag sources classification method based on sequential trajectory feature learning and Random Forest algorithm. Firstly, to fully consider the interpretable sequential trajectory features of voltage sag signals so as to improve the quality and interpretability of calculation process, the fused lasso generalized eigenvector (FLAG) algorithm is adopted to quickly search for interpretable shapelets sub-sequences from the labeled voltage sag data. After that, the labeled data and samples to-be-classified are subjected to shapelet transformation through the shapelet sub-sequences to obtain the sequential trajectory features. Finally, the random forest is trained on the labeled sequential trajectory feature data to achieve supervised sample classification, which inherits the interpretability of shapelet. To test the feasibility and validity of the proposed voltage sag sources classification method, the simulation cases based on the simulated voltage sag signals were studied. The simulation results show that the proposed method has significant advantages in terms of accuracy and interpretability of voltage sag sources classification.https://ieeexplore.ieee.org/document/9749094/Random forestsequential trajectory featuresshapeletvoltage sag sources classification |
| spellingShingle | Zhang Yikun He Yingjie Zhang Haixiao Li Jiahao Li Yijin Liu Jinjun Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm IEEE Access Random forest sequential trajectory features shapelet voltage sag sources classification |
| title | Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm |
| title_full | Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm |
| title_fullStr | Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm |
| title_full_unstemmed | Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm |
| title_short | Classification Method of Voltage Sag Sources Based on Sequential Trajectory Feature Learning Algorithm |
| title_sort | classification method of voltage sag sources based on sequential trajectory feature learning algorithm |
| topic | Random forest sequential trajectory features shapelet voltage sag sources classification |
| url | https://ieeexplore.ieee.org/document/9749094/ |
| work_keys_str_mv | AT zhangyikun classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm AT heyingjie classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm AT zhanghaixiao classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm AT lijiahao classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm AT liyijin classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm AT liujinjun classificationmethodofvoltagesagsourcesbasedonsequentialtrajectoryfeaturelearningalgorithm |