Line-to-Line Faults Detection for Photovoltaic Arrays Based on I-V Curve Using Pattern Recognition

Abstract

Fault detection plays a crucial role in reliability and safety of photovoltaic systems. However, the fault detection by the conventional protection devices is always difficult due to nonlinear characteristics of PV systems, Maximum Power Point Tracking (MPPT), low irradiation, and high fault impedance. In addition, it may lead to the power losses, efficiency reduction and even fire hazard. This paper proposes an innovative fault detection method based on the pattern recognition techniques and extraction of the essential features from the current-voltage (I-V) characteristics. The main benefit of this method is using less data to detect faults while improving accuracy. The primary results demonstrate that the proposed method is accurate, effective and reliable for detecting line-line faults in PV systems.