Research On Key Technology Of Equipment Management System Driven By Field Data

With the coming of new industrial revolution,the competition in manufacturing industry is becoming more and more intense.All countries have issued a series of policies suitable for the development of domestic manufacturing industry along the road of manufacturing informatization development.Each manufacturing enterprise also began to integrate modern information technology with traditional manufacturing industry,which is used to improve production efficiency and reduce production cost.As an important part of manufacturing enterprise production,equipment management has been paid attention by the enterprise managers.At present,many domestic manufacturing enterprises have basically built equipment management system in equipment management,which uploads equipment production information to the server in real time,and completes real-time supervision of equipment by setting fixed threshold judgment rules on the server side.But there are still some places to be improved in the domestic equipment management system,such as in some specific use scenarios,the real-time and stability of some equipment management systems are not high,the equipment operation and diagnosis technology is difficult to meet the needs of production management,and preventive equipment maintenance functions are not perfect.In view of these problems,this dissertation optimizes the structure and improves the function of the existing equipment management system.The main work of this dissertation includes:(1)This dissertation analyzes the existing problems of the existing equipment operation status diagnosis measures,and proposes a method of equipment operation state diagnosis based on cart algorithm.The method establishes the real-time diagnosis model of equipment operation state by calculating Gini coefficient.At the same time,the promotion method is introduced to solve the problem that single cart tree is prone to have fitting phenomenon and the classification accuracy is not high.By constructing many cart trees and combining them,the classification accuracy of diagnostic model can be improved.Finally,the test results show that the model of equipment operation state diagnosis after additive combination has high real-time and accuracy in equipment operation fault diagnosis.(2)By using the three parameter Weibull distribution to analyze the equipment history maintenance records,the reliability function of the equipment is obtained.The migration factor is added to the reliability function to simulate the change trend of the reliability function in the actual use of the equipment.Finally,the optimal maintenance period of preventive maintenance is obtained for reasonable production scheduling.(3)By changing the existing equipment management system structure and introducing the working mechanism of cloud edge collaboration,some core functions of equipment management are transferred to the edge end,so as to improve the real-time and reliability of equipment management system.Finally,taking the PVC rolling equipment management system of a new material Co.,Ltd.in Foshan,Guangzhou as an example,the equipment management system of the enterprise is reconstructed.While redesigning the functional structure of the or iginal equipment management system,the equipment operation state diagnosis model and the equipment preventive maintenance cycle prediction model are applied to the equipment management system.Finally,a equipment management system driven by fi eld data is realized,which has the functions of real-time diagnosis of equipment oper ation state and equipment preventive maintenance cycle prediction.