An Alternative Approach for Periodicity Detection in Various Time Series Data of Electrical Power Systems

Abstract

Proper analysis of time series data is an important task that should be considered in numerous engineering applications. Examining the periodic nature of such data has a greater impact in some fields such as dynamic analysis of power systems or analyzing the electrical energy consumption behavior of residential users. In this study, three types of time series data are handled, which are transient electrical current signal in a simulated power system, annual energy consumption data of real residential users, and synthetically created electrical power signal containing harmonic distortion with various high-frequency components. In this study, an alternative approach for the identification of time series characteristics is also proposed, which is based on statistical analysis with a different structure of the sampling window, and relevant results of periodicity detection analysis carried out for time series data are revealed. In addition to that, autocorrelation function of the same data is also calculated for comparison purposes. In conclusion, it is demonstrated that the proposed approach has significant properties in the context of periodicity detection when compared to the well-known autocorrelation function. It is a promising result for the proposed approach when considering performance in periodicity detection analysis realized for time series data. © 2025 IEEE.