ニューラルネットワークによる配電用ポリマーがいしの漏れ電流発生量予測手法の提案

カテゴリ: 論文誌(論文単位)

グループ名: 【B】電力・エネルギー部門

発行日: 2023/05/01

タイトル(英語): Proposal of Prediction Method of Leakage Current Generation of Polymeric Insulators for Distribution Lines using Neural Network

著者名: 石関　智哉（東北大学），辻　知耀（東北大学），八島　政史（東北大学），佐藤　智之（東北電力ネットワーク（株）），渡部　卓也（東北電力ネットワーク（株））

著者名(英語): Tomoya Ishizeki (Tohoku University), Kazuaki Tsuji (Tohoku University), Masafumi Yashima (Tohoku University), Tomoyuki Sato (Tohoku Electric Power Network Co., Inc.), Takuya Watanabe (Tohoku Electric Power Network Co., Inc.)

キーワード: ポリマーがいし，磁器がいし，汚損，課電曝露試験，漏れ電流，気象条件　　polymeric insulator，porcelain insulator，contamination，voltage applied exposure test，leakage current，meteorological conditions

要約(英語): Compared to conventional porcelain insulators, polymeric insulators are easier to work because of their lightness, and have better withstand voltage performance under discharge, rain, and contaminated conditions due to their higher hydrophobicity. Therefore, their application is expected to reduce the size and cost of power distribution facilities, so the introduction of polymeric insulators is being promoted in Japan. On the other hand, the use of polymeric materials for the outer coating raises concerns about their deterioration over time, and clarifying the long-term performance in the contaminated and humid environment unique to Japan is an important issue for practical application. To evaluate the long-term weather resistance of polymeric insulators for distribution lines, Tohoku Electric Power Network Co., Inc. has conducted field tests in the coastal areas along the Sea of Japan. In this study, we developed prediction models of leakage current generation based on meteorological conditions using feed-forward neural networks and verified the effectiveness of the models. And we also evaluated meteorological factors which affect the prediction results with the neural network models.

本誌: 電気学会論文誌B（電力・エネルギー部門誌） Vol.143 No.5 （2023）

本誌掲載ページ: 282-289 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejpes/143/5/143_282/_article/-char/ja/