Research And Implementation Of Virtual Monitoring System For Machine Tools Process Based On Digital Twin

With the gradual integration of traditional manufacturing and Internet information technology,traditional manufacturing has shifted to intelligent manufacturing.The basis of intelligent manufacturing is the interconnection of physical space and digital space,and digital twin is the best way to achieve the integration of physical space and digital space.Based on the digital twin of CNC machine tools,this paper studies the visual monitoring technology of multi-source heterogeneous data in the machining process and the fusion method of the physical rules of digital twin,and finally implements a virtual monitoring system for the machining process of machine tools based on digital twin.The main research contents of this article are as follows:(1)The overall design of the virtual monitoring system: Put forward the system design goals,principles,and development process;analyze the problems of the virtual monitoring system,and conclude that the core goal of the virtual monitoring system is to improve the performance of the geometric level mapping and physical level mapping of the digital twin,and then realize virtual monitoring based on digital twin.Design a virtual monitoring system structure including five demand modules,and a five-layer structure of the virtual monitoring system network architecture.(2)Design and implementation of machine tool visual monitoring based on multisource heterogeneous data: Analyze and collect multi-source heterogeneous data of machine tool processing,design and implement data integration based on Json;design and implement asynchronous data transmission method based on Redis and RabbitMQ,develop a device data management cloud platform.The results show that the asynchronous transmission method improves the reading and writing performance by 5.4 times and the throughput by 1.64 times.Design and implement mixed programming of C # and Python,and analyze the processing of machine tool operating data.Research and implementation of data visualization mapping and display performance optimization,develop visualization components based on Unity3 D,and use scene elimination technology to optimize the display performance of the system.(3)Research on the physical simulation of the digital twin model of machine tools based on deep learning: In the mapping of the physical level of digital twin,research on physical rule abstraction based on simulation data regression calculation modeling and physical rule packaging based on deep learning.Research and realize the image processing of simulation working condition data and simulation result data,establish a regression calculation model based on cGAN,and implement based on TensorFlow.The VMC0656 e machine tool is used as the object to realize the fusion of the transient thermodynamics and physical rules of the spindle system.With small errors,the solution time and the scale of the data generated by the calculation are significantly reduced.(4)Integration test of virtual monitoring system: Explain the development environment of virtual monitoring system,design the operating process and interface of the system;Test the virtual monitoring system client,the results shows that the system has low hardware resource occupancy rate and good display performance;The VMC0656 e machine tool is used as an example to test the virtual monitoring system.The experiment shows that the virtual monitoring system has good real-time performance and small error in the generated results.