Research on Abnormal Diagnosis Method of Electric Energy Metering Device Based on BSO-BPNN

The traditional electric energy metering device anomaly detection mainly relies on manual detection, the electric network staff need to analyze the data of metering device collected by the electric information system periodically. Aiming at the problems of missing report, false report and low accuracy in manual judgment. A back propagation neural network algorithm (BPNN) based on beetle swarm optimization algorithm (BSO) is proposed to construct an anomaly diagnosis model for power metering devices. BSO is based on particle swarm optimization algorithm (PSO), combined with the merits of beetle antennae search algorithm (BAS). And it improved the PSO algorithm easy to fall into the local minimum problem. Both PSO and BSO improve the performance of BPNN by optimizing the weight and threshold of BPNN. The experimental results show that BSO-BPNN has higher accuracy than BPNN and PSO-BPNN in the abnormal diagnosis of metering device.