A Real-time PV Forecast using Extreme Learning Machine

カテゴリ: 部門大会

論文No: 247

グループ名: 【B】平成29年電気学会電力・エネルギー部門大会

発行日: 2017/09/05

タイトル(英語): A Real-time PV Forecast using Extreme Learning Machine

著者名: Imam Wahyudi Farad(広島大学),佐々木 豊(広島大学),餘利野 直人(広島大学),造賀 芳文(広島大学)

著者名(英語): Imam Wahyudi Farid|Yutaka Sasaki|Naoto Yorino|Yoshifumi Zoka

キーワード: 発電量予測|太陽光発電|ニューラルネットワーク|極端学習機械,Output Forecast,Photovoltaic Power Generation,Neural Network,Extreme Learning Machine

要約(日本語): In this paper, we focused on the real-time PV forecast for energy management operation. We use the public historical insolation and weather data that can be easily obtained from the meteorological agency website of Japan. We propose the novel method using these data to predict the real-time PV output. By considering the correlation between the target and the neighboring areas, a 5-minute ahead high-accurate prediction for energy management operation can be achieved. By using the correlation coefficients between several areas, we can analysis and compare the autoregressive, proposed solar radiation correlation analysis, and using extreme learning machine.

PDFファイルサイズ: 248 Kバイト