Features extraction of complex dynamic electricity signals for testing dynamic errors in electricity meters

The paper addresses the issue of rapid amplitude changes in complex dynamic electricity (CDE) signals, which can induce substantial errors in electricity metering. Firstly, a discrete signal mathematical model is established for three types of typical dynamic loads within the power grid. Secondly, six important sensitive feature parameters are identified, along with their corresponding algorithms. Subsequently, six typical run waveform modes are extracted from the amplitude domain of current run waveforms using a combination of full traversal and Principal Component Analysis (PCA). Based on these waveform modes, a library for current feature waveforms is constructed. Finally, suggestions are provided for supplementing international electricity meter standards to address the insufficiency of test points for evaluating the dynamic errors of electricity meters.