Research On Optimization Of Smart Meter Reliability Model Based On Data Fusion

As one of the most basic components of smart grid,the reliability of smart meters has a crucial impact on the economic and stable operation of the whole grid.Reliability models can directly reflect product reliability,and it is important to obtain accurate smart meter reliability models to evaluate smart meter reliability and improve it in a targeted manner.Accelerated life test is the most commonly used method to establish smart meter reliability model,which is a mature system and can cover the whole life cycle of smart meters,but due to the large difference in stress type and size between the test stress of accelerated life test and the environmental stress under field operation,the smart meter reliability model obtained by accelerated life test only often has certain errors;field data can best reflect the realistic reliability of smart meters,but due to the difference in the field data,the reliability of smart meters can be improved.The reliability model obtained from the field data cannot cover the whole life cycle of smart meters because the statistical analysis of field data started late,which makes the field data less or the existing field data are not comprehensive.If the reliability acceleration test data can be fused with the field data to get the optimized smart meter reliability model,it will be of great significance to improve the reliability of smart meters.To address the above problems,this paper relies on the support of the Southern Power Grid Company,designs a smart meter reliability accelerated life test program,and carries out accelerated life test on smart grid,processes the test data with minimum unbiased estimation,and obtains a smart meter reliability model based on accelerated life test;investigates the smart meter field data in a city under the jurisdiction of Southern Power Grid,filters and classifies the field data after processing The posterior distribution of the accelerated life test data was used as the prior distribution of the field data,and the joint posterior distribution of the field data was constructed by combining the likelihood function of the field data.The unknown parameters of the joint a posteriori distribution are obtained through sampling simulation of the joint a posteriori distribution by MCMC method,and the reliability model of smart meter based on data fusion is obtained.The reliability models of the three methods are compared and analyzed,and it is found that the data fusion-based smart meter reliability model can retain the realistic and accurate first half of the field data-based reliability model while combining the advantages of the accelerated life test-based reliability model that can cover the whole life cycle of smart meters,so that the fusion model can accurately evaluate the reliability of smart meters in the whole life cycle.The research results of this paper provide a reference basis for using smart meter field data and accelerated life test data at the same time,provide a case reference for conducting accelerated life test and data analysis of smart meters,and have a positive effect on improving the reliability of smart meters.