Construction Method Of Digital Twin Dynamic Consistency Simulation Model Of Manufacturing System Based On Cloud Fog Edge Cooperation

Modeling and simulation are important methods for humans to understand and change the world,which has greatly promoted the process of human industrialization.Currently,with the development of manufacturing industry to information and intelligence,requirements for the application breadth,depth and accuracy of modeling becomes more critical.However,in the existing modeling and simulation technology,most models lack dynamic update ability with the change of equipment in the life cycle,which limits the adaptability of the model in different life cycle stages.To solve the above problems,this paper proposed a cloud-fog-edge collaborative digital twin dynamic consistency simulation model maintenance method of manufacturing system based on the concept of digital twins.This method uses deep learning algorithm for disturbance identification,and contradiction between large amount of computing burden and real-time ability is coordinated by cloud-fog-edge framework.At the same time,the consistency of the simulation model is maintained according to the recognition results.The specific research contents are as follows:According to the functional requirements,it is divided into three parts:cloud,fog and edge,The cloud has sufficient computing power,The fog has personalized adaptability and real-time processing ability.Edge end can collect and control of equipment status at the edge.At the cloud end,the public data set is used to train the general recognition model for general context,but the accuracy is not high for specific applications.Aiming at this shortcoming,perceptual data is used to transfer the general model,and the personalized recognition model for specific application is derived.At the fog end,a distributed data storage system is built based on Hadoop for the storage of collected data.the personalized recognition model obtained from the cloud is deployed,and real-time disturbance recognition is carried out based on the data collected at the edge.Finally,according to the recognition results,the disturbance is updated into the model using XML language and simulation solver K file.The edge end is mainly responsible for controlling the device and collecting status data and uploading it to the fog end.An experimental platform for dynamic disturbance is built.The bearings with different fault types are loaded on the experimental platform,and the data collected at the edge end are uploaded to the fog end.Meanwhile,a distributed storage server and cloud high-performance computing service platform are built at the fog end.Finally,the experimental verification is carried out.Firstly,the feasibility and accuracy of the proposed dynamic disturbance identification algorithm model under the framework of cloudfog-edge cooperation are verified.Then,through the simulation analysis of the consistency updated model,simulation data is acquired,then time and frequency domain comparation between simulation data and and collected data is carried out.According to the results,the real-time ability,accuracy and feasiblity of the proposed consistency maintenance method are verified.