Operating Performance Assessment of Smart Meters Based on Bayesian Networks and Convex Evidence Theory

Smart electricity meters have been widely installed in recent years. In order to automatically assess the operating performance of smart meters, in this paper we propose a method based on Bayesian network. Bayesian network is adopted to represent the casual relationship among attributes and operating performance of smart meter. Multiple Bayesian networks are trained from data with genetic algorithm and bagging sampling, then a subset of Bayesian networks are selected for assessment. The evidence theory is used to fuse the multiple results from Bayesian networks and generate final assessment results for smart meters. The experimental results show the effectiveness and efficiency of the proposed method.