Research on Creating Multi-Attribute Power Consumption Behavior Portraits for Massive Users

With the gradual opening of electricity sales markets in China and continuous development of integrated energy systems, problems such as diversified and improper user energy consumption modes are becoming more prominent, and user behavior analyses under market environments are apparently insufficient. Multi-attribute power consumption behavior portraits including user characteristics, pattern, group character and individuality should be established to provide a theoretical basis for intelligent power consumption technologies based on coordinated “source-network-load” interactions. In this paper, based on massive user data from various industries, a pattern of user behavior is extracted from three perspectives: temperature sensitivity, electricity price sensitivity and power consumption stationarity. Characteristic indexes are created and quantified for comprehensive evaluation. User power consumption behavior tags are created. Finally, based on a fuzzy c-means (FCM) clustering algorithm, a multi-attribute user power consumption behavior portrait with quantified, classified and combined tags is created.