Research On Anti-stealing Technology Based On Electricity Acquisition System

Power theft is now a common occurrence driven by profit,and with the development of technology in the field,the ways of stealing electricity have become diversified and complicated.In order to prevent the occurrence of stealing electricity,has gradually become the most important power companies.In this paper,the research on anti-stealing of electric power acquisition system is taken as the object of discussion.In view of the current development status and existing problems of electric power acquisition system and anti-stealing of electric power,the following aspects are mainly studied:First of all,based on the on-site operation experience and many domestic and foreign papers consulted,the current electricity acquisition system and anti-electricity stealing subject related to intelligent electricity stealing diagnosis method are investigated,and the practical application of K-means algorithm and Soft Max classifier in the engineering field is deeply studied and summarized.Secondly,the logical structure and physical platform of the system are studied.This paper introduces the system function positioning,technical index setting and local remote communication layer module control.Through the deployment of user electricity information acquisition system,including information acquisition,concentrator,front machine,system master station,collector and other related content and technology introduction of equipment,six types of user data for different user business demand,user power nature and different requirements of collection and measurement.Then the present situation of stealing electricity in a power supply company is studied in detail,including the basic situation and main ways of stealing electricity and corresponding detection methods.The main stealing modes adopted by users include under-voltage stealing,under-current stealing,phase-shifting stealing,diffused stealing and no phenotype stealing,and the detection characteristics of different stealing modes are analyzed.Then,the principle and characteristics of this kind of learning method of K-means clustering algorithm are introduced and analyzed,the quadratic iteration k-means clustering algorithm is used to reduce the dimensionality of the collected data,and then the Decision model of Soft Max classifier is established.Finally,A case study was carried out based on the data samples of the customers in area A of A company with high line loss rate and large stolen electricity,and the data diagnosis and algorithm accuracy diagnosis were carried out to verify the effectiveness of the anti-stolen electricity model based on electricity acquisition system and quadratic iteration K-means clustering algorithm.