E3S Web Conf.
Volume 185, 20202020 International Conference on Energy, Environment and Bioengineering (ICEEB 2020)
|Number of page(s)||6|
|Section||Energy Engineering and Power System|
|Published online||01 September 2020|
A short term load forecasting of integrated energy system based on CNN-LSTM
1 Anhui Provincial Laboratory of Renewable Energy Utilization and Energy Saving, Hefei University of Technology, Hefei, Anhui Province,
2 School of Electrical and Automatic Engineering, Hefei University of Technology, Hefei, Anhui Province, 230009, China
* Corresponding author’s e-mail: firstname.lastname@example.org
The accurate forecast of integrated energy loads, which has important practical significance, is the premise of the design, operation, scheduling and management of integrated energy systems. In order to make full use of the coupling characteristics of electricity, cooling and heating loads which is difficult to deal with by traditional methods, this paper proposes a new forecast model of integrated energy system loads based on the combination of convolutional neural network (CNN) and long short term memory (LSTM). Firstly, the Pearson correlation coefficients among the electricity, cooling and heating load series of the integrated energy system are calculated, and the results show that there is a strong coupling relationship between the loads of an integrated energy system. Then, the CNN-LSTM composite model is constructed, and CNN is used to extract the characteristic quantity which reflects the load coupling characteristics of the integrated energy system. Then, the characteristic quantity is converted into the time series input to LSTM, and the excellent time series processing ability of LSTM is used for load forecasting. The results show that the CNN-LSTM composite model proposed in this paper has higher prediction accuracy than the wavelet neural network model, CNN model and LSTM model.
© The Authors, published by EDP Sciences, 2020
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.