Low Frequency Oscillation Mode Estimation Using Synchrophasor Data
Deep learning techniques have been widely used for power system operations as a very hot topic in recent years. This paper proposed a spatiotemporal deep learning-based low frequency oscillation mode estimation method using synchrophasor data. The proposed deep learning method consists of the graph...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9046012/ |
| _version_ | 1818665118013063168 |
|---|---|
| author | Jinhuan Zhang Haixia An Na Wu |
| author_facet | Jinhuan Zhang Haixia An Na Wu |
| author_sort | Jinhuan Zhang |
| collection | DOAJ |
| description | Deep learning techniques have been widely used for power system operations as a very hot topic in recent years. This paper proposed a spatiotemporal deep learning-based low frequency oscillation mode estimation method using synchrophasor data. The proposed deep learning method consists of the graph convolutional network (GCN) and the long short-term memory (LSTM). A graph network is used to model the power network in which the edges represent the connection relationships between system buses. Specifically, the GCN method is used to capture the topological structure of the power networks to obtain the spatial dependence. The LSTM model is used to capture the dynamic change of electrical variables that can be monitored by PMU devices through synchrophasor data to obtain the temporal dependence. Eventually, the proposed GCN-LSTM model is used to capture the spatiotemporal patterns and features of the system-wide synchrophasor data. To validate the effectiveness of the proposed method, we evaluate it on two classic IEEE simulation platform, i.e., IEEE 39-bus system and IEEE 118-bus system, by comparing with the Nonlinear Autoregressive Neural Network with External Input (NARX), the Gated Recurrent Unit (GRU) network, and single LSTM methods. It is demonstrated that the proposed GCN-LSTM method can achieve the best estimation results for different simulation platforms. |
| first_indexed | 2024-12-17T05:43:33Z |
| format | Article |
| id | doaj.art-2b9f6322f3f249dc91156f463edf4769 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-17T05:43:33Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-2b9f6322f3f249dc91156f463edf47692022-12-21T22:01:23ZengIEEEIEEE Access2169-35362020-01-018594445945510.1109/ACCESS.2020.29829799046012Low Frequency Oscillation Mode Estimation Using Synchrophasor DataJinhuan Zhang0https://orcid.org/0000-0001-9637-8558Haixia An1Na Wu2Department of Electric-Mechanical Engineering & Automation, Tianjin Vocational Institute, Tianjin, ChinaDepartment of Electric-Mechanical Engineering & Automation, Tianjin Vocational Institute, Tianjin, ChinaDepartment of Electric-Mechanical Engineering & Automation, Tianjin Vocational Institute, Tianjin, ChinaDeep learning techniques have been widely used for power system operations as a very hot topic in recent years. This paper proposed a spatiotemporal deep learning-based low frequency oscillation mode estimation method using synchrophasor data. The proposed deep learning method consists of the graph convolutional network (GCN) and the long short-term memory (LSTM). A graph network is used to model the power network in which the edges represent the connection relationships between system buses. Specifically, the GCN method is used to capture the topological structure of the power networks to obtain the spatial dependence. The LSTM model is used to capture the dynamic change of electrical variables that can be monitored by PMU devices through synchrophasor data to obtain the temporal dependence. Eventually, the proposed GCN-LSTM model is used to capture the spatiotemporal patterns and features of the system-wide synchrophasor data. To validate the effectiveness of the proposed method, we evaluate it on two classic IEEE simulation platform, i.e., IEEE 39-bus system and IEEE 118-bus system, by comparing with the Nonlinear Autoregressive Neural Network with External Input (NARX), the Gated Recurrent Unit (GRU) network, and single LSTM methods. It is demonstrated that the proposed GCN-LSTM method can achieve the best estimation results for different simulation platforms.https://ieeexplore.ieee.org/document/9046012/Mode estimationsimulation platformsynchrophasor datalow frequency |
| spellingShingle | Jinhuan Zhang Haixia An Na Wu Low Frequency Oscillation Mode Estimation Using Synchrophasor Data IEEE Access Mode estimation simulation platform synchrophasor data low frequency |
| title | Low Frequency Oscillation Mode Estimation Using Synchrophasor Data |
| title_full | Low Frequency Oscillation Mode Estimation Using Synchrophasor Data |
| title_fullStr | Low Frequency Oscillation Mode Estimation Using Synchrophasor Data |
| title_full_unstemmed | Low Frequency Oscillation Mode Estimation Using Synchrophasor Data |
| title_short | Low Frequency Oscillation Mode Estimation Using Synchrophasor Data |
| title_sort | low frequency oscillation mode estimation using synchrophasor data |
| topic | Mode estimation simulation platform synchrophasor data low frequency |
| url | https://ieeexplore.ieee.org/document/9046012/ |
| work_keys_str_mv | AT jinhuanzhang lowfrequencyoscillationmodeestimationusingsynchrophasordata AT haixiaan lowfrequencyoscillationmodeestimationusingsynchrophasordata AT nawu lowfrequencyoscillationmodeestimationusingsynchrophasordata |