Online crack detection on photovoltaic devices using a dynamic response analysis

This study presents a method to detect cracks in solar photovoltaic modules by analyzing their dynamic electrical response without interrupting operation. The approach evaluates indicators like settling time and damping coefficient using dynamic current and voltage measurements. Baseline assessments rely on electroluminescence imaging and I–V curve analysis. A DC/DC converter generates transient responses, and outdoor tests under stable irradiance confirm the method’s reliability, achieving a correlation coefficient above 0.89. Results show that cracks affect the damping coefficient in both current and voltage. Cracked modules exhibit a damping coefficient notably different from healthy ones. A linear dynamic electrical model supports this, showing healthy modules have a more oscillatory response. This method enables real-time, non-intrusive fault detection in PV modules, offering a practical solution for continuous health monitoring in solar energy systems. Its effectiveness across varying temperatures and irradiance suggests broad applicability in real-world conditions. Future research should address nonlinear aspects of the transient response, extend testing to diverse conditions, and integrate this method with current diagnostic techniques to improve accuracy. Additionally, incorporating advanced signal processing and machine learning could further enhance its ability to identify faults.