Power Quality Events Classification on Real-Time Voltage Waveform Using Short Time Fourier Transform and Bayes Classifier
Okelola, Muniru Olajide

Abstract
The increased use of non-linear loads and the occurrence of faults on the power systems have resulted in deterioration in the quality of power supplied to the end users. Power quality (PQ) events and other disturbances lead to either malfunctioning or complete failure of electrical /electronic equipment. Voltage dip, voltage swell, voltage interruption and harmonic distortions among others are the most common types of power quality (PQ) events. Accurate and timely detection of these PQ events is very essential for adequate corrective measures to be taken. This paper investigates such events by employing the use of Short Time Fourier Transform (STFT) for PQ events detection on a real-time voltage waveform, and Na�ve Bayes classifier for event classification. The Na�ve Bayes Classifier was trained using some generated synthesized parameters to model the common PQ events. The results revealed that the technique classifies voltage PQ events with high accuracy.

Full Text: PDF