Academic Journal of Engineering and Technology Science, 2022, 5(5); doi: 10.25236/AJETS.2022.050506.
Zhaozhengyang Li, Haoqian Wu
Automated Institute, Chongqing University of Posts and Telecommunications, Chongqing, China
In recent years, as the electrical energy of distributed storage has gradually increased, the randomness of load demand has increased. This makes it more difficult to rationally dispatch and store loads. How to quickly and accurately dig out more effective information and objective laws from the massive power load data, effectively clarify the instability and timing of load changes, and then reduce the consumption and accidents in dispatching is of great significance. This paper uses wavelet analysis and stochastic sparrow search algorithm to optimize LSTM neural networks to construct long- and short-term power load forecasting models, in Singapore as of December 2021 Empirical studies were conducted on the electricity load from 24 to January 23, 2022, as well as the corresponding peak-to-valley electricity prices, meteorological data and other related data. The results show that the long-term and short-term power load prediction model based on WNR-SSA-LSTM is compared with that of traditional LSTM In terms of neural networks, RMSE and MAPE were reduced by 4.000% and 1.7871%, model goodness-of-fit R2 increased by 0.0196, It is more suitable for short-term load forecasting.
Short- and long-term power load, Wavelet noise reduction, LSTM neural network, Sparrow search algorithm
Zhaozhengyang Li, Haoqian Wu. Short-term Power Load Forecasting Based on WNR-LSTM — Take the Singapore Region as an Example. Academic Journal of Engineering and Technology Science (2022) Vol. 5, Issue 5: 28-33. https://doi.org/10.25236/AJETS.2022.050506.
[1] Zhang Lin, Wang Tinghua, Zhou Huiying. A Multi-strategy Improved Sparrow Search Algorithm [J/OL]. Computer Engineering and Applications:1-10 [2022-03-06].
[2] Wang Xiaohui, Deng Weiwei, Qi Wang. Power Load Forecasting Model Based on PSO-LSTM [J]. Shanghai Energy Conservation, 2022(02): 164-169. DOI: 10.13770/j.cnki.issn2095-705x.2022.02.007.
[3] Lu Jixiang, Zhang Qipei, Yang Zhihong, Tu Mengfu, Lu Jinjun, Peng Hui. Short-term load forecasting method based on CNN-LSTM hybrid neural network model [J]. Power System Automation, 2019, 43(08): 131-137.
[4] Pei Chao, Hooper. The Influence of Linguistic Prior Knowledge on the Natural Language Processing Task of Neural Network Models [J]. Journal of Chinese Information Science, 2017, 31(06): 10-17.
[5] Liu Ye, Cheng Shan, Wang Rui, Zuo Xianwang, Xu Jingwei. Classification method of power quality disturbance based on wavelet noise reduction and deep learning [J/OL]. Proceedings of the Society of Power Systems and Automation: 1-7[2022-03-06]. DOI: 10.19635/j.cnki.csu-epsa.000940.
[6] Kang Yi, Shi Liu Jun, GUO Gang. Short-term load forecast of power system based on WT-IPSO-BPNN [J]. Electrical Technology, 2021, 22(01): 23-28+62.]
[7] Zhang Kaikai, Yan Ping. Application of Time Series and Kalman Filtering in Short-term Load Forecasting of Power System [J]. Electrical Switches, 2016, 54(02): 91-96.]
[8] Liao Qingling, Dou Zhenhai, Sun Kai, Zhu Yaling. Optimization of load prediction of support vector machine based on adaptive particle swarm algorithm [J]. Modern Electronics Technology, 2022, 45(03): 125-129. DOI: 10.16652/j.issn.1004-373x.2022.03.023.
[9] Study on short-term load probability density prediction method based on regression analysis [D]. North China Electric Power University (Beijing), 2018
[10] Zhu Jian'an, Wei Yunbing, Zhu Pengjie, Jiang Chengcheng, Zhu Chengming. Medium- and long-term load prediction based on optimized grey Fourier residual correction [J]. Electronic Science and Technology, 2021, 34(12): 49-55. DOI: 10.16180/j.cnki.issn1007-7820.2021.12.009.