Electrical Pulsed Infrared Thermography and supervised learning for PV cells defects detection

As the most basic elements of photovoltaic (PV) module and power station, the defects in PV cells can affect the overall performance of the module and the operation status of the power station. Therefore, it is very important to carry out defects detection of PV cells. In this work, we have built an Electrical Pulsed Infrared Thermography (EPIT) experimental system to detect PV cells with different types of defects, such as broken gate, hidden crack, scratch and hot spot. Then, thermography sequence information of PV cells is captured by an infrared camera. Linear Discriminant Analysis (LDA) and Quadratic Discriminant Analysis (QDA), the two supervised learning (SL) algorithms, are used to process the thermography sequence and recognize the defects. In addition, the identification effect of the two algorithms is quantitatively evaluated by signal-to-noise ratio (SNR). The experiment results show that the EPIT method can realize the PV cells defects detection. Under different bias excitation, QDA algorithm is superior to LDA in SNR index, which can identify various defects of PV cells effectively.