Automatic power quality recognition and analysis system using wavelet-based neural network

This paper introduces a novel approach to detect power quality disturbance of distribution power system combing complex wavelet transform based on radial basis function neural network. Firstly, this paper tries to explain to design complex supported orthogonal wavelets by compactly supported orthogonal real wavelets. Secondly, the extraction of disturbance signal to obtain the feature information is explored. Finally several novel wavelets combined information to analyze the disturbance signal is proposed and is superior to real wavelet analysis result. For power quality disturbance pattern recognition the feature obtained from wavelet transform coefficients are inputted into radial basis function network. The network structure and parameter identification is fulfilled by establishing the power quality disturbance recognition model and using genetic algorithm. When signal representing fault is inputted to the trained network, the type of disturbance can be obtained. The simulation result show that the complex wavelet transform combined with radial basis function network is more sensitive to signal singularity and has significant improvement over current methods in real-time detection and better noise proof ability.