Remote Diagnosis System of Robot Intelligent Flexible Production Line Based on Convolutional Neural Network

Abstract

In order to improve the intelligent control level and automatic monitoring ability of robot intelligent flexible production line, a remote diagnosis system of robot intelligent flexible production line based on convolutional neural network was designed. Firstly, the dynamic and kinematic models of intelligent flexible production line were established. Secondly, the resonance fault parameters of the robot intelligent flexible production line were detected, and the resonance fault parameters of the robot intelligent flexible production line were measured. Then, the convolutional neural network control algorithm was used for fault feedback, and the fault feedback tracking model for remote diagnosis of robot intelligent flexible production line was obtained. Finally, the resonant fault error of the robot intelligent flexible production line was compensated by the self-feedback dynamic adjustment mechanism, the remote resonant fault diagnosis of the robot intelligent flexible production line was realized, and the integrated design was realized under the embedded control platform. The simulation results show that the designed oscillation control fitting curve is basically consistent with the actual oscillation control fitting curve and has good stability. In addition, the remote diagnosis accuracy of the design system is close to 100%, and the load error range is small. At the same time, the control time of the design system is short, which can effectively realize the remote diagnosis of the robot intelligent flexible production line and improve the output stability and robustness of the robot intelligent flexible production line.