Energy Load Forecasting Using a Dual-Stage Attention-Based Recurrent Neural Network

被引:15
作者
Ozcan, Alper [1 ]
Catal, Cagatay [2 ]
Kasif, Ahmet [3 ]
机构
[1] Akdeniz Univ, Dept Comp Engn, TR-07070 Antalya, Turkey
[2] Qatar Univ, Dept Comp Sci & Engn, Doha 2713, Qatar
[3] Bursa Tech Univ, Dept Comp Engn, TR-16330 Bursa, Turkey
关键词
dual-stage attention-based recurrent neural network; time series forecasting; energy consumption prediction; CNN-LSTM;
D O I
10.3390/s21217115
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
Providing a stable, low-price, and safe supply of energy to end-users is a challenging task. The energy service providers are affected by several events such as weather, volatility, and special events. As such, the prediction of these events and having a time window for taking preventive measures are crucial for service providers. Electrical load forecasting can be modeled as a time series prediction problem. One solution is to capture spatial correlations, spatial-temporal relations, and time-dependency of such temporal networks in the time series. Previously, different machine learning methods have been used for time series prediction tasks; however, there is still a need for new research to improve the performance of short-term load forecasting models. In this article, we propose a novel deep learning model to predict electric load consumption using Dual-Stage Attention-Based Recurrent Neural Networks in which the attention mechanism is used in both encoder and decoder stages. The encoder attention layer identifies important features from the input vector, whereas the decoder attention layer is used to overcome the limitations of using a fixed context vector and provides a much longer memory capacity. The proposed model improves the performance for short-term load forecasting (STLF) in terms of the Mean Absolute Error (MAE) and Root Mean Squared Errors (RMSE) scores. To evaluate the predictive performance of the proposed model, the UCI household electric power consumption (HEPC) dataset has been used during the experiments. Experimental results demonstrate that the proposed approach outperforms the previously adopted techniques.
引用
收藏
页数:15
相关论文
共 33 条
[1]   Reliable Solar Irradiance Forecasting Approach Based on Choquet Integral and Deep LSTMs [J].
Abdel-Nasser, Mohamed ;
Mahmoud, Karar ;
Lehtonen, Matti .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2021, 17 (03) :1873-1881
[2]   Accurate photovoltaic power forecasting models using deep LSTM-RNN [J].
Abdel-Nasser, Mohamed ;
Mahmoud, Karar .
NEURAL COMPUTING & APPLICATIONS, 2019, 31 (07) :2727-2740
[3]   Short-term self consumption PV plant power production forecasts based on hybrid CNN-LSTM, ConvLSTM models [J].
Agga, Ali ;
Abbou, Ahmed ;
Labbadi, Moussa ;
El Houm, Yassine .
RENEWABLE ENERGY, 2021, 177 :101-112
[4]  
Boden M., 2002, DALLAS PROJECT
[5]   Multi-Sequence LSTM-RNN Deep Learning and Metaheuristics for Electric Load Forecasting [J].
Bouktif, Salah ;
Fiaz, Ali ;
Ouni, Ali ;
Serhani, Mohamed Adel .
ENERGIES, 2020, 13 (02)
[6]   Optimal Deep Learning LSTM Model for Electric Load Forecasting using Feature Selection and Genetic Algorithm: Comparison with Machine Learning Approaches [J].
Bouktif, Salah ;
Fiaz, Ali ;
Ouni, Ali ;
Serhani, Mohamed Adel .
ENERGIES, 2018, 11 (07)
[7]   Time Series Forecasting with Multi-Headed Attention-Based Deep Learning for Residential Energy Consumption [J].
Bu, Seok-Jun ;
Cho, Sung-Bae .
ENERGIES, 2020, 13 (18)
[8]   Short-Term Load Forecasting: Similar Day-Based Wavelet Neural Networks [J].
Chen, Ying ;
Luh, Peter B. ;
Guan, Che ;
Zhao, Yige ;
Michel, Laurent D. ;
Coolbeth, Matthew A. ;
Friedland, Peter B. ;
Rourke, Stephen J. .
IEEE TRANSACTIONS ON POWER SYSTEMS, 2010, 25 (01) :322-330
[9]  
Cho K., 2014, P SSST 8 8 WORKSH SY, P103, DOI [10.3115/v1/w14-4012, DOI 10.3115/V1/W14-4012, DOI 10.3115/V1/D14-1179]
[10]  
Chung J, 2014, NIPS 2014 WORKSH DEE