Wavelet transform based fuzzy logic for power quality classification

This paper proposes a wavelet-based fuzzy-expert system to power quality disturbance classification. The wavelet transform and Parseval´s theorem are introduced in this paper. To extract features in power-quality disturbance, the energy distribution of wavelet at each decomposition level is calculated. A fuzzy-expert system is proposed to classify the power quality disturbances. Based on the energy distribution patterns the membership functions of input and rule base are generated. Finally, the fuzzy set of the output variable is converted to a crisp number, in terms of which the disturbances are classified. The types of disturbances concerned include voltage sag/swell, interruption, harmonic, and flickers.