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: 2022/11/01タイトル(英語): Insolation Forecasting by using Image-processed Satellite Images Data and Forecasting of Cloud Imageries著者名: Shoji Kawasaki (Department of Electronics and Bioinformatics, School of Science and Technology, Meiji University), Qingkun Wang (Department of Electronics and Bioinformatics, School of Science and Technology, Meiji University)著者名(英語): Shoji Kawasaki (Department of Electronics and Bioinformatics, School of Science and Technology, Meiji University), Qingkun Wang (Department of Electronics and Bioinformatics, School of Science and Technology, Meiji University)キーワード: insolation forecasting，photovoltaic generation，satellite image，neural network，ConvLSTM要約(英語): Recently, there is a movement aimed at realizing distributed energy systems, especially photovoltaic power generation (PV) is expected in Japan, and it is expected that the number of interconnections of PV will increase in the future. Insolation is easily affected by the weather and the supply of electricity becomes unstable. Therefore, it is important to forecast the amount of insolation. In this study, the main purpose of this research is to develop a highly accurate and practical method for insolation forecasting, based on satellite image data provided by the Himawari-8 satellite and past actual measurement data released by the Japan Meteorological Agency. In this paper, the authors propose an amount of insolation forecasting method by using neural network and carry out the accuracy verification by case study. In addition, the authors also use the ConvLSTM (Convolutional Long Short-Term Memory) model to forecast the cloud imageries for the insolation forecasting. In order to verify the usefulness of the proposed method, the forecast results by the proposed method were compared with the forecast results without using satellite image data and the forecast results with latest satellite image data. In conclusion, by using the satellite image data and forecasting the cloud imageries, it was possible to grasp the characteristics of the insolation, prevent erroneous prediction, and improve the prediction accuracy.本誌: 電気学会論文誌B（電力・エネルギー部門誌） Vol.142 No.11 （2022） 特集：気象情報の利活用に基づく新しい電力系統運用技術本誌掲載ページ: 525-532 p原稿種別: 論文／英語電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejpes/142/11/142_525/_article/-char/ja/
