Research On Condition Monitoring Technology Of Energy Metering Device

Electric energy measurement is an important part of the electric power production process.Electric energy metering device is an indispensable metering device in Chinese industrial production and daily life.Its running condition is directly related to the accuracy and reliability of energy metering.Traditional statistics and analysis work for energy data is based on manual on-site meter monitoring,which comes staffing complex,or low efficiency and prones to human error.The research and application of the condition monitoring technology for the electric energy metering device will change this unfavorable situation.As an important part of smart grid technology,the realization of the aboved technology will help promote the development of smart grid process.This paper studies the realization method of using metering automation system to solve the problem of condition monitoring in electric energy metering device.That is,to employ electricity load data with the collection of electrical parameters system on site and apply the principle of electrical engineering,data analysis and Artificial Intelligence technology to achieve the operation of the energy metering device analysis and evaluation.Therefore,the abnormal measurement of energy metering device could be timely found out that ensures the accuracy of measurement.Based on the research results and the latest development of equipment condition monitoring technology at home and abroad,this paper introduces the basic architecture of energy metering automation system and electric energy metering device,and summarizes some existing state monitoring technologies and their applicabilities.Then,from the perspective of electrical theory,this paper analyzes the abnormal sources that may occur in the operation management of the electric energy metering device and finds out the topological relations which obtains the theoretical or empirical criterion.More profoundly,the principles,techniques and methods of the operation condition monitoring for the electric energy metering device are analyzed and designed in detail;which makes deeper researchs and developments in the existing electrical parameter information from the point of statistical mathematics.In this case,some practical methods are given an example,including Distance measurement method,Similarity function method and Partial least squares method(PLS).Ultimatily,on the basis of Information Fusion,the BP neural network,the Analytic hierarchy process(AHP)and the Fuzzy theory widely used in the Artificial Intelligence technology system are studied to analyze and apply the information of different dimensions,so as to carry on the condition monitoring process and the processing technology of the electric energy metering device.With the support of existing theories and related technologies,the data collected by the Metrological automation system put into operation by the power supply bureaus of China Southern Power Grid are taken as the basic data sources.And the simulations of these methods are carried out by filtering the suspects in the data source.The simulation results show that the above methods can effectively monitor the condition of each specific object,which can realize the remote condition evaluation of the electric energy metering device.