Fault early warning model of power metering automatic verification assembly line based on residual neural network

Abstract

In order to solve the problem of low accuracy of traditional fault early warning model for automatic verification of electric power measurement, a new fault early warning model is designed based on residual neural network. The model is based on residual neural network, and uses the methods of feature extraction, data training and transition classification to get the fault early warning model for the target system. The residual neural network is introduced into the verification system of automatic assembly line to realize the integration of fault early warning and diagnosis. Based on the deep learning neural network, the fault classification model for automatic assembly line is constructed, and the model early warning effect is tested and verified with the actual system operation data. The results show that this method has an ideal fault early warning effect, and provides more accurate information for automatic assembly line fault detection and prevention.