Design of Mechanical Fault Monitoring System Based on Deep Learning Algorithm

It provides a new idea for the condition monitoring of machine equipment. Firstly, based on the mechanical frequency domain convolutional neural network, a multi-modal data fusion algorithm is constructed. This method can get rid of the massive information processing and diagnosis, and realize the adaptive extraction and intelligent diagnosis of health status. It overcomes the traditional shallow modeling method which is difficult to depict the complicated corresponding pain points of signal and health state. The method of combining attention network and multi-level features is used to realize the effective identification of small defects. The accuracy of the system is improved by using series neural network for fault diagnosis. A new multi-stage diagnosis model based on neural network is proposed by combining theoretical and experimental research for different locations, types, working conditions and large samples.