Fault Diagnosis Technology of Photovoltaic Device Based on Computer Prediction

This paper presents a fault diagnosis method for electronic circuits based on power equipment and applies tolerance fault diagnosis simulation experiments in analog circuits. The experimental results show that the fault diagnosis accuracy of the proposed method is more than 99% in analog circuits. Compared with the SVM method, it is found that the proposed method is insensitive to parameter selection, and RF (random forest) model training time is short, which indicates that the proposed method has good performance in the tolerance fault diagnosis of analog circuits. It further demonstrates the reliability of the proposed method. Experiments show that the proposed method is efficient, stable, and more suitable for engineering applications.