Fault Monitoring Method of Electric Energy Meter Verification Assembly Line Based on Deep Learning
Abstract
In order to realize the whole process monitoring and fault warning of the electric energy meter's verification assembly line, the deep learning method is adopted to monitor its faults. First, for complex and multi-source monitoring heterogeneous collected data, VQ-VAE is used to construct a feature extraction model to achieve data dimensionality reduction, while retaining sample information. Subsequently, in order to ensure the mutual correlation and influence between the entities, the Sperman rank correlation coefficient is used to calculate the correlation between the features. CGN is used to construct a fault warning model to realize the fault prediction of the deterioration state of the equipment. Finally, experimental comparative analysis shows the effectiveness of this method. Compared with traditional methods, it improves the real-time and accuracy of monitoring, and provides more accurate information for troubleshooting and prevention of automated assembly line.