Electricity consumption behavior portrait of power users considering seasonal load characteristics

This work suggests a big data mining-based approach to assess power users’ power consumption behavior while taking seasonal load characteristics into account, with the goal of improving our understanding of power enterprises’ power consumption behavior. he original load data is cleaned and preprocessed, and the feature is filtered by variance filtering; Second, the seasonal base load and easy to disturb load correlation coefficients are gradually clustered, producing the bicluster label findings. Third, the processed load data is split into basic seasonal load and easy to disrupt load, accounting for the seasonal independence of base load. Ultimately, the profile of the user’s electricity consumption behavior created by the two kinds of labels is acquired.