Load forecasting method based on data mining technology
ZHANG Min1, QIAN Shuangqiu1, WU Zhongqi2, WANG Ziyuan2
1. Nantong Power Supply Company of State Grid Jiangsu Electric Power Company, Nantong, Jiangsu 226000; 2. Tianjin Tiandianqingyuan Technology Co., Ltd, Tianjin 300000
Abstract：The power load prediction can guide and make decisions for the rational planning of power grid. This article uses the data mining technology for medium and long-term load forecasting in Nantong area. Firstly, cluster analysis is applied to classify all feeder data in Nantong, and the correlation analysis and grey correlation analysis are used to analyze the degree of influence of external factors on the load changes. Then these influential factors are used as the input of the neural network to establish a neural network model based on the clustering radial basis function (RBF), and the load forecasting results are obtained. Finally, compared with the prediction without considering clustering and only using RBF neural network model, the experimental results show that the method in this paper is more advanced, significantly improves the load prediction accuracy, and achieves the purpose of ensuring the reliability of power supply.
2021, Vol. 22 
