Abstract
Recognition of power quality events by analyzing voltage waveform disturbances is a very important task for power system monitoring. This paper presents a hybrid intelligent scheme for the classification of power quality disturbances. The proposed algorithm is realized through three main steps: feature extraction, feature selection and feature classification. The feature vectors are extracted using S-transform (ST) and Wavelet transform (WT) which are very powerful time-frequency analysis tools. In order to avoid large dimension of feature vector, three different approaches are applied for feature selection step, namely Sequential Forward Selection (SFS), Sequential Backward Selection (SBS) and Genetic Algorithm (GA). In the next step, the most meaningful features are applied to Probabilistic Neural Network (PNN) as classifier core. Various transient events, such as voltage sag, swell, interruption, harmonics, transient, sag with harmonics, swell with harmonics, and flicker, are tested. Sensitivity of the proposed algorithm under different noisy conditions is investigated in this article. Results show that the classifier can detect and classify different power quality signals, even under noisy conditions, with high accuracy.