Photovoltaic Panel Fault Detection and location System Based on Series Depth Mode

With the significant improvement in photovoltaic panel fault detection accuracy, researchers have proposed many models to locate the detected faults on photovoltaic panels. At present, faults can be located in photovoltaic panels, but the robustness of fault detection methods in most models is not high. For example, different fault contrast of the same photovoltaic panel in different pictures and incomplete solar reflection filtering is likely to lead to false detection. For this reason, this paper proposes a photovoltaic panel fault detection and location system based on the series depth learning model, which introduces data augmentation and changes the solar reflection algorithm for photovoltaic panel sequence to the solar reflection filtering algorithm for the single photovoltaic panel. At the same time, the faulted edge in the PV panel data set determines the type of fault, so we further considered and decided to use the edge-sensitive Mask R-CNN for fault detection. The simulation results show that the new model reduces the previous false detection and improves the accuracy of fault detection based on fault location.