Load Forecasting Method based on Long Short-term Memory Network and Business Expansion Installation
DOI:
https://doi.org/10.62051/ijepes.v3n1.05Keywords:
Monthly Load Forecast, Business Expansion and Installation, Long-term and Short-term Memory NetworkAbstract
In view of the problem that the traditional monthly load forecasting method lacks considering the internal load influencing factors, this paper puts forward the monthly load forecasting method considering the industry expansion. On the basis of the similar load development trend, the benchmark value of the planning annual forecast is determined. Then, the basic data of uncertainty and selection factors were collected, and the correlation model affecting the bias of load size factors and the percentage of load deviation was established based on the long-and short-term memory network, and then the benchmark value of monthly load prediction value was corrected. The comparison of the forecast results of considering the actual expansion, the uncorrected expansion and the industry expansion shows that the actual expansion has an important impact on the monthly load and can effectively improve the prediction accuracy.
References
[1] Biswal B ,Deb S ,Datta S , et al.Review on smart grid load forecasting for smart energy management using machine learning and deep learning techniques[J].Energy Reports,2024,123654-3670.
[2] Jiang Mengyang, Cheng Haozhong, Wu Zhen, et al. Monthly Load Forecasting Method Using Vector Machines Considering Business Expansion and Installation [J]. Proceedings of the CSU-EPSA, 2017, 29 (07): 1-6.
[3] Zhou C ,Guo D ,Hu H , et al.TCN Multi-time-scale Transformation and Temporal-Attention Neural Network for Monthly Electricity Consumption Forecasting[J].Recent Advances in Electrical & Electronic Engineering,2023, 16(8):872-883.
[4] Sing C L ,Zhenyao M ,Ting W , et al.Load forecasting based on deep neural network and historical data augmentation[J].IET Generation, Transmission & Distribution,2020,14(24):5927-5934.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.