Photovoltaic panels surface defect assessment based on vision-transformer

The penetration of photovoltaic (PV) power generation into the grid is increasing, but its intermittency and instability pose major challenges to grid operation. Among them, a crucial impacting the efficiency of PV power is the local shading resulting from contamination of PV panels, leading to power loss. Therefore, a method for detecting the degree of PV panel anomalies based on vision-transformer is proposed in this paper, the degree of PV panel anomalies is estimated using images captured by surveillance cameras. Specifically, in this paper, transforms the PV panel images captured by surveillance cameras into data. Secondly, constructs a vision-transformer PV panels surface defect assessment model to achieve the panel anomaly degree detection. Finally, the effectiveness of the method is illustrated by a case study. The test results show that the performance of the proposed method in this paper is 4.98% better than the F1 of the previous study.