Power distribution area load prediction method and system based on multi-model hierarchical learning

Abstract

The invention provides a power distribution area load prediction method based on multi-model hierarchical learning. The method comprises the following steps: acquiring related data of a power distribution area; and importing the related data of the power distribution area into a pre-trained multi-model hierarchical learning model for prediction to obtain a load value required to be predicted by the power distribution area, wherein the multi-model hierarchical learning model comprises an upper-layer prediction model and a lower-layer prediction model which are connected in sequence; k algorithm models are preset in the upper prediction model; an algorithm model is preset in the lower-layer prediction model, the input of the algorithm model comes from the output results of the K algorithm models in the upper-layer prediction model, and the output of the algorithm model is the load value required to be predicted. The invention also provides a power distribution area load prediction system based on multi-model hierarchical learning. By implementing the load prediction method, the generalization ability of the load prediction model is improved by constructing the multi-model hierarchical learning model by utilizing the difference of various algorithms, so that the problem of relatively large prediction deviation in the prior art is solved.

Description

technical field [0001] The invention relates to the technical field of electric power system data processing, in particular to a method and system for load forecasting in a distribution station area based on multi-model layered learning. Background technique [0002] In recent years, as an important work of a series of departments such as power system distribution network planning, dispatching, operation and maintenance, the distribution network load has been widely valued. Accurate load forecasting will effectively improve the level of dispatching and safe operation of the distribution network, help guide the consumption of photovoltaics in low-voltage stations, and optimize the power flow operation mode of the distribution network. [0003] With the gradual maturity of online monitoring equipment in distribution stations and the comprehensive advancement of digital grid construction, it is no longer difficult to obtain massive and stored power load data information and rel...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the embodiments of the present invention is to provide a load forecasting method and system for a distribution station area based on multi-model hierarchical learning. By using the differences of various algorithms, a multi-model hierarchical learning model is constructed to improve the load. The generalization ability of the prediction model, so as to solve the problem of large prediction deviation in the existing technology

Images