Power quality disturbance recognition using stransforms and FCM-based decision tree

This paper presents a new approach for recognizing nonstationary signals in power quality (PQ) disturbances. Meanwhile the new approach includes the most types of PQ disturbance, such as voltage sags, swells, interruptions, transients and harmonics. The new model mainly includes two steps. Firstly, S-transform is used to analyze power system disturbance signals, and two most distinguishing features are extracted. In this process based on these two features, 2D feature vectors are clustered using hierarchical Fuzzy C-means algorithm (FCM). Secondly, a binary decision tree is constructed from FCM cluster centers to automatic recognize disturbance patterns. Finally the simulation results show the validity and efficiency of the proposed model.