Analysis of Power Load Characteristics and Research of Load Forecasting Model in a Certain Area Under the Background of New Power Reform

• Published in: 2023 5th International Conference on Smart Power & Internet Energy Systems (SPIES) pp. 100 - 104
• Main Authors: Boning, Yu, Junwei, Qin, Zhenli, Deng, Xingwu, Guo, Jianan, Si
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2023
• Subjects: Analytical models; Load characteristic; Load forecasting; Maximum temperature; Neural networks; Power supplies; Prediction model; Predictive models; Temperature control; Temperature distribution
• DOI: 10.1109/SPIES60658.2023.10474724

Along with the gradual realization of "double carbon" development goal in our country, the demand for electric power has increased exponentially, and the continuous expansion of the electric load has become more and more challenging to the power system in our country. Based on the existing research results, this paper analyzes the power load characteristics from three aspects: regional load characteristics, typical user load characteristics and distributed power supply load characteristics. Secondly, based on the characteristics of regional load and the research results, the influence factors of regional power load, namely power supply, maximum temperature and output value ratio, are determined, and the regional power load forecasting model is established by using BP neural network method. Finally, a power load is selected for an example analysis. The results show that, first, the error is controlled within 5% compared with the actual maximum load when using neural network method to forecast the data over the years, and the prediction accuracy is higher; Second, with the increasing of maximum temperature and output value ratio, the regional maximum load increases, which has a positive impact on it.