Design Of Non-intrusive Anti-theft Method And Automatic Identification System

Under the new era,the prosperous and beautiful world is independent from the security of electricity.Enhancing electricity security is regarded as a necessary part of the energy planning of the "13th Five-Year Plan",as well as one of the most important parts considering China's rapid economic development.China is a country with poor electricity resource.However,as China's economy takes off and the problem of electricity theft becomes more and more serious,the reason is that capitalists in order to reduce the cost of electricity to obtain huge economic profits.At the same time,with the hidden way of stealing electricity,high-tech,China's annual power grid loss is much higher than other developing countries.According to the survey,private enterprise processing plants,electrolytic metals,catering,collective enterprises and other traditional industries theft of electricity is serious,and with the combination of finance and IT Internet derived from such as blockchain,Bitcoin "mining",futures trading and other lucrative industries,these industries account for a largr part of electricity theft.These industries cause serious loss of power resources to the country,but also tests the safety of the power grid.Traditional anti-stealing methods are generally divided into measures to prevent damage to the metering end device and identification of electricity theft by analyzing electricity theft data,to prevent the destruction of the metering device measures generally have the metermeter to seal,regulate the installation of the meter wiring,strengthen auxiliary metering equipment and other measures.The collection of stolen electricity data is generally obtained by manual checking and intrusive power detection system,this anti-theft method in the system cost,complexity,intelligence,stability is insufficient and lacking.Based on the research on non-intrusive power detection and anti-theft power in China and abroad,this paper puts forward the application of non-intrusive power detection algorithm in anti-theft technology.Because this application uses a non-intrusive device,it changes the original need to decentralize the collection and sensing system on the electricity meter of each user's home.Only by measuring and analyzing the load characteristics of the current,voltage,power and harmonics at the bus port,Solved the problem of simplified and intelligentized collection and recognition.Themain contribution of this paper mainly is introduced as follows.1.Summarize the traditional stealing electricity methods and the methods of high-tech theft,such as high-frequency interference and photovoltaic electricity theft,and also makes a comparative study and analysis of the characteristics of electricity theft behavior,and extracts the parameters of conventional electricity consumption,including current,voltage,power,phase,line loss,etc.2.Analyze and study the non-intrusive anti-theft algorithm,mainly involving the non-intrusive characteristic parameter harmonic detection algorithm,including the use of non-intrusive harmonic feature extraction,the main component analysis of dimensional,fuzzy C mean cluster outlier detection analysis,the introduction of the concept of membership and fuzzyness analysis harmonic abnormal data.3.According to the previous analysis and research,a set of non-intrusive anti-stealing automatic identification system is designed.This system can extract current,voltage,power,phase and other conventional electrical characteristic parameters,and use the non-intrusive anti-stealing algorithm to harmonize This special characteristic parameter of wave is collected and processed.Mainly collect and extract current,voltage,power,phase,harmonics and other electrical characteristic parameters through transformers and ATT7022 B measurement circuit,and then use simplified,dimensionality reduction,wavelet and other methods to preprocess the characteristic data to construct a set of recognizable Analytical model of electricity theft.Finally,according to the electricity stealing identification model system,the abnormal electricity consumption result of the user is identified and diagnosed to judge the suspected electricity stealing user.4.Debug the core components of the system and the ATT7022 B metering chip,and use tools such as NI Multisim and Matlab to simulate and analyze the current and voltage characteristics and fuzzy C-means clustering harmonic outliers detection analysis.In this paper,through the study of non-intrusive anti-stealing algorithms and systems,it has changed the mode that only manual meter reading and intrusive power detection systems can obtain electricity stealing data in the past,effectively applying non-invasive power monitoring to anti-stealing electricity,saving the cost of anti-stealing electricity and manpower and material resources,improving the stability and reliability of the grid measurement system,and improving the prediction accuracy of electricity theft identification.It will provide a good reference for identifying some common power abnormalities through non-invasive power monitoring in the future.