Simulation of Multilevel Depth Filtering of Energy Metering Information Under Dynamic Iterative Sampling

The amount of data generated by power metering systems is huge and complex, including power consumption records, power quality parameters, equipment status, etc., so it is difficult to filter large-scale power metering data effectively. Therefore, a multilevel depth filtering method for energy metering information under dynamic iterative sampling is proposed and verified by simulation. Based on dynamic iterative sampling architecture, the characteristics of redundant points of energy metering information are obtained by sampling time series of energy metering information. Based on this, voltage anomaly data, current anomaly data, and power factor anomaly data are extracted. The mean-shift transfer function is established to realize multilevel depth filtering of redundant points of energy metering information. The experimental results show that under different learning rates and pretraining iterations, the filtering accuracy of the basic and new types of energy metering information is higher, and the application efficiency is higher.