Abstract
Most of the factories still employ traditional equipment without interfaces of status parameters transmission and standard communication protocol. The inspection of traditional equipment has become a difficult and key point in the development of manufacture. The present study aimed to count printing, inspect status and identify types of traditional printing devices. A smart sensor-based system is proposed in this work. Variational mode decomposition (VMD) is utilized to extract feature values, and back-propagation neural network (BPNN) models are built for type identification and status inspection of the printing devices. Experiment results show that: the accuracy of print counting, status inspection and type identification of printing devices were 100%, 86.6% and 93.3% respectively. The methodologies and system proposed in the present study can be adapted to other industrial equipment identification and inspection, which can significantly facilitate the development of industrial technologies.