Research On Electricity Inspection Based On Big Data

Electricity inspection is an important part of electric power marketing,which is of great significance to the safe and reliable operation of power grid and even the economic benefits of power supply enterprises.With the continuous development of intelligent distribution network and advanced metering infrastructure,power distribution and utilization data gradually presents the characteristics of 5Vs(Volume,Variety,Velocity,Value and Veracity).Electricity abnormal detection has been mainly based on traditional anti-stealing technology with a lot of human and material resources,a lot of false alarms,a long consuming time and low efficiency.With the application and development of big data technology in power system,the efficiency and accuracy of electricity inspection is expected to be improved.Research and application of big data technology for electricity inspection is of great significance to improve the power service quality,reduce non-technical losses of power grid,save human and material resources and reduce operating costs.This thesis focuses on the framework of electricity inspection and methods,designs a framework of electricity inspection based on big data technology and presents a customer electricity abnormal detection approach,develops an electricity abnormal detection module and applies it into practice.The details are as follows:(1)The businesses,characteristics,data sources and the data processing technology of electricity inspection is analyzed.The scenarios of applying big data processing technology in electricity inspection are studied and a framework of electricity inspection using big data technology is designed.The framework is illustrated by five steps: data resource,data classification,analysis model,function module and application.(2)A two-stage electricity abnormal detection approach based on k-means clustering algorithm is proposed by extracting customer load patterns and considering the influence of temperature and humidity on load using power load curves,and its validity is verified by a real case of electricity detection.(3)The detailed flow chart of electricity inspection business using two-stage electricity abnormal detection approach is presented,a corresponding function module is developed by Visual Studio 2010 and its application is demonstrated.The work of this thesis provides a reference for electricity inspection business based on big data technology.The developed module can provide a tool for power supply enterprises to improve the accuracy and efficiency of electricity inspection.Thus,the management and economic benefits for power supply enterprises is reinforced.