JAUE2020-066: A hybrid model for power consumption forecasting using EEMD-based long short-term memory neural network

Abstract

Energy consumption prediction is a popular research field in computational intelligence. However, traditional energy consumption prediction methods usually ignore the importance of the selection of the modeling sample, which often results in poor forecasting performance. To address this gap, this paper proposes a hybrid forecasting method including data preprocessing and forecasting modules. For data preprocessing, the ensemble empirical mode decomposition (EEMD) technique is used to reduce the influence of noise within the raw data series to obtain a more stable sequence. In the forecasting module, long short-term memory (LSTM) algorithm is used to predict power consumption. In order to verify the reliability of EEMD-LSTM, empirical mode decomposition (EMD) algorithm is selected as the reference model. Compared with the traditional methods, like Artificial Neural Network (ANN) and long short-term memory (LSTM) neural network, the results indicate that the hybrid forecasting method outperforms the discussed traditional forecasting models in forecasting accuracy.