Load temperature decoupling method and system based on user feature analysis

Abstract

The application discloses a load temperature decoupling method and system based on user feature analysis, relates to the technical field of power distribution network load decoupling, and comprises the following steps: acquiring a general decoupling model, performing load temperature decoupling based on current temperature and current user information, acquiring a preliminary decoupling result, further acquiring a decoupling residual error, correcting the decoupling residual error by means of an industry temperature sensitivity parameter and an electricity capacity fluctuation rate of the current user, and acquiring a decoupling effect parameter; adding a user with a decoupling effect parameter smaller than a decoupling effect threshold to a difficult decoupling user group, selecting a best general decoupling sub-model based on user features, historical temperature and historical load of the difficult decoupling user group for incremental learning, acquiring a special decoupling sub-model, and performing load temperature decoupling on the difficult decoupling user group by combining the special decoupling sub-model and the general decoupling model except the special decoupling sub-model, so as to acquire a load temperature decoupling result. The application solves the technical problem of insufficient load temperature decoupling accuracy in the prior art.

Images