Skip to Main Content
To improve the precision of power quality disturbance detection and recognition in distributed power system, a novel method based on complex transform wavelet transform is presented. Due to the property that instantaneous amplitudes of voltages and currents as well as instantaneous phase differences can be obtained, the combined information with time and frequency localization properties are defined. These have advantages over Fourier transform expression in the frequency domain when significant distortions are present in the signals, causing the periodicity to be lost. The signal containing noise is de-noised by wavelet transform to obtain a signal with higher signal-to-noise ratio. The feature obtained from wavelet transform coefficients are inputted into wavelet network for power quality disturbance pattern recognition. By means of enough samples to train the network, the synthesized approach of recursive orthogonal least squares algorithm with improved Givens transform is used to fulfill the network parameter identification. The simulation results demonstrate that the combined information with wavelet network achieve more useful signal features, and improve detection and classification accuracy.