ELECTRONIC DEVICE, XGBoost-BASED POWER DATA ANOMALY PREDICTION METHOD, AND STORAGE MEDIUM

Abstract

The invention provides an electronic device, a XGBoost-BASED power data abnormality prediction method and a storage medium. The method comprises the following steps: acquiring power data of a preset building in a first predefined period of time; analyzing the obtained electric power data using a predetermined outlier determination method to obtain a first anomaly-free electric power data set and afirst anomaly-free electric power data set; Substituting the first abnormal power data set into a pre-trained outlier prediction model for outlier prediction to predict abnormal power data in the first abnormal power data set. The invention can improve the accuracy of forecasting abnormal data in electric power data, and provide reference data for the construction and promotion of smart grid.

Description

technical field [0001] The invention relates to the field of power data processing, in particular to an electronic device, an XGBoost-based power data abnormality prediction method and a storage medium. Background technique [0002] Power data (referring to the historical records of electricity consumption of various buildings collected in the power system) is a key reference factor in the construction and promotion of smart grids. At present, power data is mainly read by meter readers of power supply companies after daily meter reading. Then enter the data management system through professionals. On the one hand, due to the fact that a large amount of manpower is used to complete the entire data entry process, abnormal power data is prone to occur. On the other hand, during the data entry process, there may be system failures, line Interruption and other abnormalities in the input data caused by the failure of the equipment itself; and the abnormality of power data will cau...

AI Technical Summary

Problems solved by technology

[0002] Power data (referring to the historical records of electricity consumption of various buildings collected in the power system) is a key reference factor in the construction and promotion of smart grids. At present, power data is mainly read by meter readers of power supply companies after daily meter reading. Then enter the data management system through professionals. On the one hand, due to the fact that a large amount of manpower is used to complete the entire data entry process, abnormal power data is prone to occur. On the other hand, during the data entry process, there may be system failures, line Interruption and other abnormalities in the input data caused by the failure of the equipment itself; and the abnormality of power data will cause unexpected accidents in the construction and application of the smart grid. If the building has power outages or circuit burnouts, it may even lead to accidents.

Images