The Research Of Restraining Electricity Stealing On Various Energy Incorporating Into The User Side

With the development of new energy technologies, energy generation and more clientswill be directly incorporated into the distribution network, this can save fossil fuels andreduce carbon emissions, but it will also greatly affect the distribution network operation,control and planning. Among them, as more and more clients and more energy into thedistribution grid, for Stealing technology requirements will also be higher.As in the power transmission line loss rate remains high, there is a phenomenon oftenoccurs stealing, it will cause huge economic losses. High-speed communicationstechnology-based management and stealing electricity metering automation systems for theimplementation of the project to improve power supply enterprise management level andeconomic benefits of a great help. Since a large number of users of electricity accumulation ofhistorical data, you want fast, comprehensive and effective analysis of large powerconsumption data and historical information, and stealing behavior, it is a powerful datamining tool.This article will introduce metering automation systems, CCS marketing managementsystem, stealing models and data mining techniques, and then reveals the essence oftampering with the traditional Stealing Technology in a multi-energy into the client when thelack of data mining techniques and some algorithms principle was proposed based on datamining technology Stealing model, clustering techniques and classification technique ismainly used in this study, data mining techniques. Stealing model can be divided into twosub-problems: a group of users to use electricity characteristic attribute clustering analysis andthe user’s normal use of electricity with the user’s the main difference between non-normal useof electricity, to find that the user can conduct electricity the distinguishing features. Followedby classification tree method to be able to represent the difference between the user’selectricity generation determine the main characteristics of the suspect stealing the user’sclassification rules, and then follow this classification rule to predict the user’s electricity type,and select a power line in Longgang Power Supply Bureau area to check and verification, thepredictive result is quite close to the actual condition.