Design And Implementation Of Industrial Equipment Health Monitoring System Based On Deep Learning

With the development of Internet,Internet of things,wireless communication and intelligent manufacturing technology,the ability of industry to collect and store data is increasing,and the amount of data collected is increasing exponentially,which brings a new perspective of processing and mining information for industry.In the field of fault diagnosis,modern industry can get the health status of equipment by analyzing the operation data collected.However,the data acquisition speed is usually faster than the analysis speed of the diagnostic staff.How to extract available features from industrial big data efficiently and accurately identify the corresponding health status has become an urgent research topic.In this paper,combined with the actual application scenario of industrial Internet of things,big data processing technology and fault diagnosis algorithm based on deep learning are studied,and a health monitoring system for industrial equipment is developed for the problem of vertical double bearing equipment.The main work of this paper includes:(1)A fault diagnosis algorithm based on deep learning is proposed.In this paper,the deep neural network and convolutional neural network technology are studied,and a fault diagnosis algorithm based on multi-dimensional feature fusion is designed and implemented.The fault diagnosis algorithm does not need to carry out feature engineering on the collected data.It can capture the correlation of multiple types of original sensor data under different sampling frequencies in parallel,and can complete the task of multi classification diagnosis to achieve hierarchical diagnosis.This paper analyzes the performance of the fault diagnosis algorithm from the perspective of input data dimension change and model comparison using the actual equipment operation data.The experimental results show that the accuracy of the algorithm has been significantly improved.(2)A health monitoring system for industrial equipment is designed and implemented based on the proposed algorithm.This paper analyzes the requirements of industrial equipment health monitoring system,designs the overall framework of health monitoring system software,including 6 functional modules of data collection,data storage,data processing,data calculation,data interaction and user management,and implements the health monitoring system with big data and machine learning technology.This system not only has the basic functions of data storage,data display,fault diagnosis and user management,but also can obtain the real equipment operation status feedback data through the online learning mechanism designed in this paper,so as to optimize the fault diagnosis model in the operation process.(3)In this paper,the function test of each module of the industrial equipment health monitoring system is carried out on the spot.In addition,the performance test of fault diagnosis function is carried out aiming at the bearing center inconsistent problem of large vertical dual-bearings rotating machinery.The test results show that the system runs stably and can diagnose the equipment fault in real time.