Method for Early Warning of Faults of Solar Cells Based on Convolutional Neural Networks

Solar cells are a renewable energy source, and their efficiency and quantity are constantly increasing. Over time, battery performance has decreased, which may lead to malfunctions. Traditional solar cell fault warning has problems such as low prediction accuracy, low efficiency, and weak adaptability. This article proposed a Convolutional Neural Network (CNN) model based on Dropout and ResNet-50 technologies for early warning of solar cell failures. Among them, CNN extracted the input image from the original input, performed classification and prediction, and identified faults based on the differences between the input image and the output image. The experimental results showed that the solar cell fault warning method relying on convolutional neural networks could effectively improve the prediction accuracy, with the highest accuracy reaching 97.9%. In addition, it could also improve the efficiency of fault diagnosis and had high robustness.