Electricity Consumption Forecasting using Deep Neural Network Model

カテゴリ: 部門大会

論文No: SS3-5

グループ名: 【C】平成30年電気学会電子・情報・システム部門大会プログラム

発行日: 2018/09/05

タイトル(英語): Electricity Consumption Forecasting using Deep Neural Network Model

著者名: Chandramitasari Widyaning(Waseda University),Kurniawan Bobby(Waseda University),Fujimura Shigeru(Waseda University)

著者名(英語): Widyaning Chandramitasari|Bobby Kurniawan|Shigeru Fujimura

キーワード: Deep Neural Network|Stacked Long Short-Term Memory|Electricity Forecasting|Feed-Forward Neural Network

要約(日本語): Electricity has a main role in development of standard living of society. Since electricity energy is difficult to store, it is better to have the good energy management system in power supply company. Electricity consumption forecasting helps the power supply company to keep the balance of electricity supply and demand. In this work, we proposed combination of deep neural network model to forecast the electricity consumption in manufacturing company. We forecasted the electricity consumption for the next day using combination of Stacked Long Short-Term Memory (SLSTM) and Feed-Forward Neural Network (FNN). To show the performance, we compared the experimental result of proposed method with normal LSTM. We evaluated the performance by calculating the Root Mean Squared Error (RMSE) score.

PDFファイルサイズ: 782 Kバイト