Electricity stealing analysis method and device

Abstract

The invention discloses an electricity stealing analysis method and device. The method comprises steps that electricity behavior data of a user and archive data of the user are acquired; the acquired data is made to be sample data, the sample data comprises training data, test data and prediction data, the training data and the test data have labels, and the labels are made through marking historical electricity stealing records of the user one by one; the training data and the test data are loaded to a machine learning model established based on an Xgboost algorithm, and the model is trained and tested; the prediction data is loaded to the trained model, and the electricity stealing behavior analysis result of the user is acquired. The method is advantaged in that high-efficiency and accurate electricity stealing suspicion analysis can be realized.

Description

technical field [0001] The invention relates to the technical field of electric power management, in particular to a power stealing analysis method and device. Background technique [0002] Stealing electricity is an illegal act that has always existed since the beginning of the power supply. The act of stealing electricity has brought great harm to the power department and people's lives. It can damage low-voltage electrical equipment and damage the property interests of power companies. Leading to large-scale network power outages and even electric shock casualties due to electricity theft, threatening the personal safety of others. In order to combat electricity theft, power supply companies have been committed to the work of anti-theft. However, with the development of science and technology, the means of stealing electricity are becoming more and more "sophisticated", the behavior is becoming more and more hidden, and the difficulty of anti-theft work is also increasing...

AI Technical Summary

Problems solved by technology

[0007] Although these methods have achieved good results, generally speaking, their shortcomings are: low efficiency, high equipment and manpower costs, and now the means of stealing electricity are high-tech and concealed. The traditional anti-stealing methods are no longer No matter how suitable it is, a new and more powerful and effective anti-stealing method is needed

[0009] But its disadvantage is: BP neural network algorithm will easily fall into the problem of insufficient local minimum points and long iteration time, so the effect is not very ideal

Images