Published Paper Details:

The main focus of this paper is on detection of arc fault and flash in DC photovoltaic (PV) system and DC grid which may causes fire hazard, personnel shock hazard and damage other system elements or appliances. A notable complication to their detection is that arc in dc system is not periodic, and thus may not have easily perceptible amplitude or frequency signature for pattern recognition based detection technique. Wavelet transform (WT) provides a time and frequency approach to analyze target signals with multiple resolutions. This paper proposes an effective method based on WT and support vector machine (SVM) for detection arc fault in DC PV systems. The process of detecting an arc fault involves signal analysis and then feature identification from system voltage signals. SVM is then used to identify arc fault location. Simulation results are synthesized to study and verify the accuracy of proposed methodology.