Real-time model base fault diagnosis of PV panels using statistical signal processing

This paper proposes new method of monitoring and fault detection in photovoltaic systems, based mainly on the analysis of the power losses of the photovoltaic system (PV) by using statistical signal processing. Firstly, real time new universal circuit based model of photovoltaic panels is presented. Then, the development of software fault detection on a real installation is performed under the MATLAB/Simulink environment. With model based fault diagnosis analysis, residual signal from comparing Simulink and real model is generated. To observe clear alarm signal from arbitrary data captured, Wald test technic is applied on residual signal. A model residual based on Sequential Probability Ratio Test (WSPRT) framework for electrical fault diagnosis in PV system is introduced.