The Research On The Household Variable Relation And Line Loss Anomaly Diagnosis In Courts Based On The Electricity Information Acquisition System

The stable operation of the Electricity Information Acquisition System can effectively respond to the growth of the electri city demand of the power users.With the non-standard and inadequate management of the power users,the fault of the Electricity Information Acquisition System is increasingly prominent.Due to the lack of diagnostic methods,when a fault occurs,the failure quickly troubleshoot will bring great obstacles to the operation and maintenance of the whole system.In particular,the abnormalities of household variable relation and line loss in courts seriously affect the power consumption experience of power users and directly bring huge economic losses to power grid companies.Existing single fault diagnosis technology researching did not make full use of electricity data of Electricity Information Acquisition System itself including the event data,traffic data,the main operating data to deepen the analysis and research.How to use the big data of the Electricity Information Acquisition System and propose the methods to reasonably diagnose the abnormalities of household variable relation and line loss in courts is the main content of the present study.This thesis takes the Electricity Information Acquisition System as the research object.Firstly,The research background and significance of this thesis are expouded.Then the structure and function modules of the Electricity Information Acquisition System are introduced in detail.According to the nature of the fault system fault is classified.The causes of failure and damage are described.The abnormalities of household variable relation and line loss in courts are researchd emphatically.In order to solve the problem of abnormal household variable relation in the Electricity Information Acquisition System,this thesis proposes an abnormal diagnosis method of household variable relation based on improved K-means clustering and improved Pearson correlation coefficient.This thesis first through the principal component analysis of GIS system for courts total table and voltage meter data reduces dimension.Setting up to improve voltage data extracted K-means clustering characteristics,the improved Pearson correlation coefficient algorithm analysis of the user to be detected,accordingly based on improved K-means clustering and Pearson correlation coefficient between abnormal change diagnosis method,realize the abnormal area user belongs to correct diagnosi s.The analysis results of actual calculation examples show that the algorithm proposed in this thesis can effectively realize the accurate detection and analysis of abnormal users in the case of identifying one or more abnormal users in the same platform area and multiple abnormal users in different platform area.Compared with the traditional detection methods,the implementation is simple and the accuracy is higher.To solve the problem of abnormal line loss in the Electricity Information Acquisition System,this thesis proposes a line loss diagnosis method based on deep confidence network and improved support vector regression.Due to the complicated power grid structure and electric index data feature extraction difficulty results in poor area line loss calculating accuracy.This thesis first uses the maximum related-minimum redundancy method filter closely related to the line loss of the electrical characteristics of the index.Then extracts fully electric characteristic index data multilayer complex feature extracting through the DBN and then replaces DBN top network with a strong ability of nonlinear regression SVR network.The Particle Swarm Optimization to optimize the Support Vector Regression model of SVR parameter optimal solution is introduced.According to the data of Electricity Information Acquisition System verifies the calculation results of this method and realizes the abnormal diagnosis of line loss in high loss station.The analysis results of actual calculation examples show that the relative errors of the line loss rate calculated by the method proposed in this thesis are all within 5%,and the diagnosis rate of the abnormal causes of line loss in high loss stations can reach 98%.Compared with the existing line loss calculating methods,the method proposed in this thesis has higher accuracy,which provides an effective guarantee for the abnormal diagnosis of line loss in high loss station area.Based on the approaches proposed in this thesis,the Web page based on virtual instrument technology in electric Electricity Information Acquisition System fault diagnosis unit is exploited.The overall system design of the abnormalities diagnosis unit of household variable relation and line loss in courts is provided.The functions such as data reading,fault diagnosis,results query of line loss exception,results query of the abnormalities diagnosis of household variable relation and line loss in courts besed on the Electricity Information Acquisition System is programming realized.The system test and analysis by the measured data are accomplished.The effectiveness and accuracy of the m ethod are further verified.