Research on User Electricity Behavior Analysis Based on Clustering and Deep Learning

With the development of smart grids and the increasing importance of power big data, accurate analysis of user electricity consumption behavior has become particularly important. This paper addresses the demand for big data analysis of electricity in smart grids. The K-means algorithm and LSTM model are used to cluster analyze residential electricity consumption behavior and predict electricity consumption patterns. Data decomposition and random matrix theory are used to detect abnormal situations in the electricity consumption data of smart meter users. The research results indicate that the proposed LSTM model performs well in prediction accuracy and can effectively support power consumption pattern recognition and anomaly detection, providing strong support for policy formulation and economic benefits improvement in the power system.