Post-processing algorithmfor damped and step-change events detection in PMU signals

Abstract

The study presents an algorithm using signal processing techniques to screen PMUs data for events. The algorithm can be applied on multiple PMUs data for possible events detection and their display. The algorithm analyze the each moving window frame, i.e. segment of selected samples. The signal processing technique is based on decomposition of each segment via empirical mode decomposition followed by calculation of statistical indices on sum of intrinsic mode functions (IMFs)/square-root of spectral kurtosis computed on IMFs. The maximum energy content computed for each segment of PMUs signal signifies the presence of any event of damped transient, step-change/impulse or even normal condition.