An adaptive window-based denoising scheme to enhance power quality monitoring systems

Wavelets transform (WT) based monitoring of power quality (PQ) has been proven to be very effective means of detecting and localizing the PQ disturbances. However the PQ signals under investigation are often corrupted by noise. The presence of noise not only degrades the detection capability of wavelet based PQ monitoring system but also results in increased false alarm rate. To enhance the detection capability of WT based monitoring system and to avoid the false alarm, this paper proposes an adaptive window based denoising scheme of WT coefficients in noisy environment. The technique used here, exploits the local structure of wavelet coefficients, namely the intrascale dependencies of WT coefficients representing PQ disturbances, in an adaptive manner. Despite the simplicity of the proposed method, both in its concept and implementation, the results obtained after denoising of the PQ signals are among the best reported in the literature. (5 pages)