Power consumer value analysis method and system based on payment behavior

Abstract

The invention relates to the technical field of smart power grids, and provides a power consumer value analysis method and system based on a payment behavior in allusion to the problem of accurate division and identification of customer groups with different value levels under a new standard of smart power grid power marketing development. The analysis process mainly comprises three links of user payment behavior historical data acquisition and preprocessing, power user value group division and value type matching of users to be identified. The feature vectors constructed by the method can effectively express key information of power consumer value characteristics; the constructed power consumer payment behavior training data network not only can record the payment behavior state, but also effectively expresses the complex correlation contained among power consumers, and is beneficial to obtaining a division result with stronger robustness and interpretability; a power consumer value group division model is established by using a variational graph auto-encoder structure, so that the model has a certain generation capability, and the value type identification of power consumers has a stronger generalization capability.

Description

technical field [0001] The invention relates to the technical field of smart grids, in particular to a method and system for analyzing the value of electric power users based on payment behavior. Background technique [0002] Under the new direction of smart grid power marketing development, power companies must accurately locate high-quality customers, change the original thinking mode, scientifically allocate service resources, and look at marketing from a new perspective. Therefore, research and analyze the behavioral data of power users, accurately locate the needs, consumption habits, behavior trends and psychological changes of users, open up the electricity sales market for various power companies, improve the service quality of domestic users, and enhance the core competitiveness of domestic and overseas markets is of great significance. The analysis of user electricity consumption behavior is to use data mining and other means to conduct statistics, analysis, and p...

AI Technical Summary

Problems solved by technology

[0003] At present, the processing and analysis of power data by domestic and foreign power companies is in its infancy, and most of them focus on theoretical research or are limited to a certain area, and have not realized large-scale applications. The research model of user behavior data analysis is relatively single, and the mining and analysis of power payment data is limited. , the lack of an efficient and accurate clustering model, the value model helps the power enterprise department to understand users in more detail and clearly, and cannot further open up the electricity sales side for the power department and users, and provide a basis for optimizing the formulation of marketing strategies, which limits the user's experience in using electricity. Adapt to the modern power marketing that is market-oriented and customer demand-centered

Images