A Novel Method for Analysis of Power Quality Transient Disturbances

Several methods have been revealed in the reference for detection and localization of power quality (PQ) transient disturbances utilizing S-transform, Fourier transforms, Gabor-Wigner, Gabor transform, Hilbert transform and families of the wavelet transform. However, this paper presents an alternative approach to detect and locate the power quality (PQ) transient disturbances based on singular spectrum analysis (SSA). SSA is a non-parametric technique, which does not require any supposition to generate the observed signal, and affords an effective way to recognize weak transient PQ signal. It has the capability to decompose the PQ transient disturbance signals into a sum of a small number of detectable oscillatory components, removing noise and reconstructing the original signal. Based on the proposed method, transient signals are decomposed into approximate and detail signal. The experiment is performed using PQ long-and-short duration transient disturbances such as high and low-frequency oscillations, impulsive transient and voltage sag. The results of the experiment are compared with multi-resolution db8 wavelet transform. As shown in the simulation experiment results, the proposed SSA technique can be used effectively to detect and locate the transient disturbance. The proposed technique is efficient and alternative technique comparatively wavelet transform and SSA works well even in noisy environment.