Electricity stealing identification method based on dynamic time adjustment and gradient boosting tree

Abstract

The invention discloses an electricity larceny identification method based on dynamic time adjustment and a gradient boosting tree. The problem that a traditional electricity larceny detection method cannot accurately analyze high-voltage users with high electricity consumption is effectively solved. According to the method, the power utilization information data in the high-voltage user is constructed into the input vector, the input vector is normalized to obtain the training sample data, the training sample data is utilized to train the gradient boosting tree model, and the verification set is utilized to improve the accuracy of the tree model in a gradient manner, so that the accuracy of the high-voltage user is improved. Therefore, the electricity stealing probability value of the high-voltage user is obtained, whether the high-voltage user is an electricity stealing user or not is finally judged, and the analysis accuracy of the high-voltage user with high electricity consumption is improved.

Description

technical field [0001] The invention relates to the field of electricity stealing detection, in particular to an electricity stealing identification method based on dynamic time adjustment and gradient boosting tree. Background technique [0002] The phenomenon of illegal electricity stealing has always existed in our country, which has seriously increased the burden on the country in the management of electricity consumption, and the economic losses caused by electricity stealing are also getting higher and higher. The losses are as high as billions or even tens of billions of yuan. Illegal electricity stealing not only damages the legitimate rights and interests of power supply companies, disrupts the normal order of electricity supply and consumption, but also threatens the security of the power grid. [0003] The ways of electricity stealing also emerge in an endless stream with the growth of time, and the traditional electricity stealing detection methods are mainly to...

AI Technical Summary

Problems solved by technology

[0006] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a method for identifying electricity theft based on dynamic time adjustment and gradient boosting tree, which effectively solves the problem that the traditional electricity theft detection method cannot identify high-voltage users with large power consumption. Questions for Accurate Analysis

Images