Fault Diagnosis of Photovoltaic Array Based on BP Neural Network and SSTDR

The photovoltaic(PV) array in PV system is always exposed to the harsh outdoor environment. The weather-beaten PV arrays are easy to appear accelerated aging and faults, such as grounding, short circuits, open circuits, etc., which will lead to reduced output efficiency and safety threats to PV system. A SSTDR based faults detection and diagnosis method is proposed in this paper, which can be tested online by sending high frequency test signal to PV array branches and analyzing the cross correlation between the injected and reflected signals. It is superior to conventional electrical detection methods needing no additional sensors to detect the PV array’s output voltage, current, irradiance, temperature and other parameters. Faults diagnosis can also be realized combined with BP neural network. The feasibilities and validities of SSTDR based PV array fault detection and diagnoses were evaluated both by simulations and experiment tests.