Automatic classification of power quality events using S-Transform

This paper presents a new method for automatic classification of power quality events, developed based on S-transform. This transform provides high resolution time-frequency representations used to calculate power quality indices. Supplementary information about detected events (the magnitude, duration and frequency spectrum of the identified disturbances) are extracted in order to characterize the disturbances. The proposed method is analyzed using the main types of single disturbances. The results show the S-transform capabilities to detect, localize and classify power quality events.