Open AccessPower Load Forecasting Using BiLSTM-Attention
Author(s) -
Jie Du,
Yingying Cheng,
Quan Zhou,
Jiaming Zhang,
Xiaoyong Zhang,
Gang Li
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/440/3/032115
Subject(s) - computer science , robustness (evolution) , electric power system , smart grid , artificial neural network , long short term memory , power grid , big data , artificial intelligence , data mining , key (lock) , time series , scheduling (production processes) , machine learning , recurrent neural network , power (physics) , engineering , biochemistry , chemistry , physics , computer security , operations management , quantum mechanics , electrical engineering , gene
With the development of big data and artificial intelligence, the applications of smart grid have received extensive attention. Specifically, accurate power system load forecasting plays an important role in the safety and stability of the power system production scheduling process. Due to the limitations of traditional load forecasting methods in dealing with large scale nonlinear time series data, in this paper, we proposed an Attention-BiLSTM (Attention based Bidirectional Long Short-Term Memory, Attention-BiLSTM) network to do the accurate short-term power load forecasting. This model is based on BiLSTM recurrent neural network which has high robustness on time series data modeling, and the Attention mechanism which can highlight the key features playing key roles in load forecasting in input data. The verification experiments with real data in a certain area show that the proposed model outperforms other models in terms of prediction accuracy and algorithm robustness.