A singular point detection and localization of power quality disturbance using filter bank and the residuals of adaptive filter

It is very important to detect power quality disturbances effectively in power quality monitoring. In industrial and commercial power systems, voltage sags, voltage swells, transients, and interruptions are main disturbances which affect sensitive equipments. This paper presents a signal processing technique which combines filter bank system and adaptive filter for power quality disturbances detection and localization. In this method, the filter bank system which has binary tree structure is designed for decomposing input power signal to subband signal components. In each subband, adaptive filters are operated as predictor, and the residual values of each adaptive predictor are used to detect power quality disturbances. The usefulness of the proposed method is demonstrated by synthesized signals harmonically.