New Combined S-transform and Logistic Model Tree Technique for Recognition and Classification of Power Quality Disturbances

This article presents certain features and a new classification method for recognition of power quality disturbances. The proposed method uses S-transform to extract the most severable features of the power quality waveforms. The disturbances decision-making scheme is performed using the logistic model tree. This new algorithm requires fewer features compared to the wavelet based approach for the identification of power quality events. Thus, the required memory space and learning logistic model tree time are substantially reduced. The most common types of disturbances, including sag, swell, interruption, harmonics, transient, and flicker, are studied. Disturbances consisting of both sag and harmonics, as well as both swell and harmonics, are also considered. The result shows that the classifier can detect and classify different power quality events correctly. Sensitivity of the proposed algorithm under noisy condition is also investigated.