Multi-technology fused electricity stealing behavior identification method and system

Abstract

The invention discloses a multi-technology fused electricity stealing behavior identification method and system, and relates to the technical field of electricity utilization detection. The method comprises the following steps: acquiring power consumption data of a plurality of known users, taking one part of the power consumption data as a training sample of a detection model, and taking the other part of the power consumption data as a test sample of the detection model; carrying out dimension reduction processing on the power consumption data to obtain dimension-reduced data; performing unsupervised pre-training on the dimension reduction data by using a network model to obtain a trained training sample and a network parameter; selecting a trained training sample to carry out fine tuning on network parameters, and obtaining the output of a weak classifier; weighting the output of the weak classifier to obtain an anomaly detection threshold value of the to-be-detected user; and calculating an optimal detection threshold value of the detection model, and judging whether the power consumption behavior of the user is normal or not by comparing the abnormal detection threshold value with the optimal detection threshold value. The power consumption behavior can be judged according to the power consumption data of the user, and the power supply safety and reliability are enhanced.

Description

Technical field [0001] The present invention relates to electrical detection technology, and particularly relates to stealing behavior recognition method and system for multi-technology integration. Background technique [0002] With the rapid economic development, a substantial increase in electricity consumption, growing all kinds of stealing behavior. Stealing behavior with concealment, dispersion and diversification, etc., before stealing are mostly private, collective organization mainly stealing behavior now start showing a trend of diversification, high energy consumption enterprises, some state institutions, etc. also appeared on theft, stealing a number of growth caused by increased line loss, serious damage to the legitimate interests of power companies, disrupted the normal order for electricity. For stealing the problems associated power staff a lot of experimental studies, such as the use of physical methods, by optimizing the structure of the metering tank, the mete...

AI Technical Summary

Problems solved by technology

Stealing electricity has the characteristics of concealment, decentralization, and diversification. In the past, most of the stealers were private and collective organizations. Now the stealing behavior has begun to show a trend of diversified development. High-energy-consuming enterprises, some state-owned institutions, etc. Electricity theft has also occurred, and the increase in the number of stolen electricity has caused increased line losses, which has seriously damaged the legitimate rights and interests of power supply companies and disrupted the normal order of power supply and consumption.

Aiming at the problem of stealing electricity, relevant electric power workers have carried out a lot of experimental research, such as using physical methods, by optimizing the structure of the metering box and changing the material of the metering box, to a certain extent weakened the high-frequency signal and strong magnetic signal band. However, it has a limited effect on improving power loss, and the management level of power theft prevention has not been effectively improved.

The anti-stealing devices launched on the market can monitor customers' electricity consumption behavior, but the analysis and judgment of electricity theft behavior cannot achieve double protection, and power inspectors cannot obtain strong criteria to substantially prevent electricity theft behavior

Images