Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation
The development of electric power systems in the future will be more complex. Because of that, for the electric power system's operation to remain reliable, monitoring technology or algorithms is needed to support more advanced information delivery. One of the technologies used is the phasor me...
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Published: |
2021
|
| Subjects: |
| _version_ | 1797037449996664832 |
|---|---|
| author | Azhar, I.F. Putranto, L.M. Irnawan, R. |
| author_facet | Azhar, I.F. Putranto, L.M. Irnawan, R. |
| author_sort | Azhar, I.F. |
| collection | UGM |
| description | The development of electric power systems in the future will be more complex. Because of that, for the electric power system's operation to remain reliable, monitoring technology or algorithms is needed to support more advanced information delivery. One of the technologies used is the phasor measurement unit (PMU). The more PMUs used in the electric power system network, the more data will be generated from the PMU because the PMU has a high data sample resolution and is able to observe transient conditions. This paper discussed the transient stability prediction using CNN-LSTM for time step prediction using PMU data. The proposed method is used for predicting stable and unstable cases in time series data. The research focuses on stability conditions due to network changes, such as line detachment and out-of-step protection on generators when there is a loss of synchronism after the occurrence of three-phase fault. The proposed method is simulated using IEEE 39 bus test system in DIgSILENT PowerFactory. The resulting model can reach an accuracy of 99.62, with an average time of simulation per epoch is 247 s. The proposed method has a higher accuracy than the CNN and convLSTM methods and can overcome the weakness of the CNN method which consumes a lot of time during the training process. © 2021 IEEE. |
| first_indexed | 2024-03-14T00:03:30Z |
| format | Conference or Workshop Item |
| id | oai:generic.eprints.org:280232 |
| institution | Universiti Gadjah Mada |
| last_indexed | 2024-03-14T00:03:30Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | oai:generic.eprints.org:2802322023-11-07T01:08:41Z https://repository.ugm.ac.id/280232/ Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation Azhar, I.F. Putranto, L.M. Irnawan, R. Power and Energy Systems Engineering (excl. Renewable Power) Electrical and Electronic Engineering The development of electric power systems in the future will be more complex. Because of that, for the electric power system's operation to remain reliable, monitoring technology or algorithms is needed to support more advanced information delivery. One of the technologies used is the phasor measurement unit (PMU). The more PMUs used in the electric power system network, the more data will be generated from the PMU because the PMU has a high data sample resolution and is able to observe transient conditions. This paper discussed the transient stability prediction using CNN-LSTM for time step prediction using PMU data. The proposed method is used for predicting stable and unstable cases in time series data. The research focuses on stability conditions due to network changes, such as line detachment and out-of-step protection on generators when there is a loss of synchronism after the occurrence of three-phase fault. The proposed method is simulated using IEEE 39 bus test system in DIgSILENT PowerFactory. The resulting model can reach an accuracy of 99.62, with an average time of simulation per epoch is 247 s. The proposed method has a higher accuracy than the CNN and convLSTM methods and can overcome the weakness of the CNN method which consumes a lot of time during the training process. © 2021 IEEE. 2021 Conference or Workshop Item PeerReviewed Azhar, I.F. and Putranto, L.M. and Irnawan, R. (2021) Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation. In: 2021 International Conference on Technology and Policy in Energy and Electric Power (ICT-PEP). https://www.scopus.com/inward/record.uri?eid=2-s2.0-85123479458&doi=10.1109%2fICT-PEP53949.2021.9601021&partnerID=40&md5=8f349521c67817680eec91c9b4be73ed |
| spellingShingle | Power and Energy Systems Engineering (excl. Renewable Power) Electrical and Electronic Engineering Azhar, I.F. Putranto, L.M. Irnawan, R. Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title | Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title_full | Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title_fullStr | Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title_full_unstemmed | Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title_short | Transient Stability Detection Using CNN-LSTM Considering Time Frame of Observation |
| title_sort | transient stability detection using cnn lstm considering time frame of observation |
| topic | Power and Energy Systems Engineering (excl. Renewable Power) Electrical and Electronic Engineering |
| work_keys_str_mv | AT azharif transientstabilitydetectionusingcnnlstmconsideringtimeframeofobservation AT putrantolm transientstabilitydetectionusingcnnlstmconsideringtimeframeofobservation AT irnawanr transientstabilitydetectionusingcnnlstmconsideringtimeframeofobservation |