Automatic disturbance signal monitoring method for on-line detection and recognition

Based on wavelet transform with neural network, a novel approach is put forward to detect and classify power quality disturbances in distributed power system. The wavelet transform provides such a framework for the analysis of transient signal that can locate energy in both the time and scale domain. Thus, the multiresolution analysis based on wavelet transform is an excellent tool in providing spatial-frequency decomposition, employing the supported orthogonal wavelet. The application of statistics-based signal denoising is brought forward to determine the threshold of each order of wavelet space, and an effective method is proposed to determine the decomposition adaptively, increasing the signal-noise-ratio. The feature information obtained from wavelet decomposition coefficients are used as input variables of neural network for power quality disturbance pattern classification. The power quality disturbance classification model is established and the proper training algorithm is used to calculate network parameters with good convergence. The method incorporates the advantages of wavelet neural network to extract the feature information of transient signal meanwhile restraining various noises. The effectiveness of the proposed method is verified with the simulation results.