K-means clustering and correlation coefficient based methods for detection of flicker sources in non-radial power system

Abstract

Nowadays electric power quality problems such as flicker (voltage fluctuation) are major concern of electric companies and industrial consumers. Identification of flicker sources is an important stage in customers convincing and in flicker reduction process. For equal fluctuation frequency of flicker sources, detection of coupling points of these sources is not an easy task. In this paper, voltage envelope is extracted by Enhanced Phase-Locked Loop (EPLL), which acts as a nonlinear adaptive filter, and then the amplitude of flicker tone is used as a flicker index. The power system with arbitrary flicker sources is assumed and simulated to generate a set of training data. Then two methods based on K-means clustering algorithm and correlation coefficient are introduced to identify the location of flicker sources in a non-radial power system. Using proposed methods location of flicker sources with similar fluctuation frequency in non-radial power system is determined. In a typical test system, several simulations have been conducted. The simulation results show the performance of the proposed method is very promising and flicker sources can be detected tolerably.