| Presentation | 2022-01-21 Detection of fault locations in an unbranched power distribution line using deep learning algorithm Daiki Nagata, Tohlu Matsushima, Yuki Fukumoto, Hideaki Kawano, Shunya Fujioka, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In fault detection in overhead power distribution systems, it is desired to take immediate response with fewer people. Therefore, a novel method using TDR(Time Domain Reflecting) has been developed for fault detection in the overhead distribution system. However, it is known that in complex network with multiple branches and power distribution equipment such as transformers and switches, it is difficult to accurately identify the accident point due to waveform distortion and decrease in amplitude of TDR pulse. In this report, A method for detecting fault points from TDR waveforms using deep learning was proposed.The proposed method can be used to locate fault in complex power distribution networks where multiple reflected waves are observed. In this report the TDR waveform data was generated for a simple straight unbranched distribution line. A large amount of waveform data is required for deep learning. In this case, these data are obtained by circuit simulation. Therefore, for circuit simulation, the power distribution line is treated as a transmission line, and the primary constant of the line was obtained by calculating electromagnetic field of the cross-sectional structure. The transmission line was represented by cascading the fundamental matrix. The equivalent circuit model of the constructed power distribution network was calculated using MATLAB, and an environment was constructed to generate TDR waveform data. As a result, the TDR waveform data of a fault can locate the fault point, fault type and fault line. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | TDR method / distribution system / accident point probe / deep learning / fundamental matrix |
| Paper # | EMCJ2021-61 |
| Date of Issue | 2022-01-14 (EMCJ) |
| Conference Information | |
| Committee | EMCJ |
|---|---|
| Conference Date | 2022/1/21(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Atsuhiro Nishikata(Tokyo Inst. of Tech.) |
| Vice Chair | Kimihiro Tajima(NTT-AT) |
| Secretary | Kimihiro Tajima(NAIST) |
| Assistant | Kiyoto Matsushima(Hitachi) / Hiroyoshi Shida(EMC Tech.) / Tohlu Matsushima(Kyushu Inst. of Tech.) |
| Paper Information | |
| Registration To | Technical Committee on Electromagnetic Compatibility |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Detection of fault locations in an unbranched power distribution line using deep learning algorithm |
| Sub Title (in English) | |
| Keyword(1) | TDR method |
| Keyword(2) | distribution system |
| Keyword(3) | accident point probe |
| Keyword(4) | deep learning |
| Keyword(5) | fundamental matrix |
| 1st Author's Name | Daiki Nagata |
| 1st Author's Affiliation | Kyushu Institute of Technology(Kyushu Inst of Tech) |
| 2nd Author's Name | Tohlu Matsushima |
| 2nd Author's Affiliation | Kyushu Institute of Technology(Kyushu Inst of Tech) |
| 3rd Author's Name | Yuki Fukumoto |
| 3rd Author's Affiliation | Kyushu Institute of Technology(Kyushu Inst of Tech) |
| 4th Author's Name | Hideaki Kawano |
| 4th Author's Affiliation | Kyushu Institute of Technology(Kyushu Inst of Tech) |
| 5th Author's Name | Shunya Fujioka |
| 5th Author's Affiliation | Kyushu Institute of Technology(Kyushu Inst of Tech) |
| Date | 2022-01-21 |
| Paper # | EMCJ2021-61 |
| Volume (vol) | vol.121 |
| Number (no) | EMCJ-339 |
| Page | pp.pp.1-6(EMCJ), |
| #Pages | 6 |
| Date of Issue | 2022-01-14 (EMCJ) |