Design And Implementation Of Digital Twin Virtual Monitoring System For Laser Processing Equipment

With the rapid development of the Internet of Things,Cloud Computing,Artificial Intelligence and other new generation information technologies,it has gradually become possible to realize the digital twin of the complete body.Digital Twin is considered as one of the general purpose technologies and core technologies of the fourth industrial revolution,a comprehensive technology system supporting the interconnection of everything,the basis for the development of the digital economy,and the information infrastructure of the future intelligent era.In this paper,designing and implementing a digital twin virtual monitoring system for laser processing equipment in the ultra-clean room with low digitalization level,poor visualization performance,and contaminated processing environment due to high personnel mobility,and study the visualization monitoring and remote control of laser processing equipment at the geometric level and physical level during the working process.The main research contents of this paper are as follows:(1)The overall framework of the digital twin virtual monitoring system for laser processing equipment is designed: the objectives,principles and processes of the system design are formulated;the key problems in the system development process are analyzed,and preliminary solutions to the corresponding problems are proposed;to improve the stability and reliability of the system,the modular design of the digital twin system is carried out to ensure that the failure of some functions does not affect the operation of other modules;finally,the overall architecture of the digital twin system is determined,and the design logic of the system is further clarified.(2)The visualization of laser processing equipment data is realized: a variety of laser processing equipment is introduced,and the data sources of the equipment are analyzed,and the method of using PLC to connect to the data sources is proposed to collect the equipment data;the acquired device data is transmitted by building the MQTT communication protocol,the carrier model for data visualization is drawn,and the model is optimized using parametric simple surface method to reduce the performance overhead during system operation;the laser parallel processing equipment with light guide scattering dot is used as an example for data acquisition,transmission and kinematic analysis of the virtual model to realize the visualization of the geometric level of the laser processing equipment digital twin virtual monitoring system.(3)Physical rule-based surface temperature monitoring of laser processing workpiece using deep learning technology: A complete digital twin virtual monitoring system should include both geometric and physical level simulation,and physical level simulation usually requires a large number of simulation calculations,which is a timeconsuming process,so proposing to use deep learning technology to accelerate the finite element simulation calculation;by using the finite element simulation software to perform the finite element solution of the transient temperature field of the workpiece surface under some working conditions,the solution data and the working condition data are image processed using Open CV to construct a deep learning dataset of the working condition data and the result data;the deep learning CGAN model is selected and trained using the dataset,and the training process is tuned to the model;the nanosecond laser processing system is used as the research object.The trained CGAN model is used to perform fast simulation of the workpiece surface temperature.Finally,the finite element simulation technique and the deep learning model are tested in terms of solution performance.The results show that the deep learning model greatly improves the temperature cloud generation speed with small solution error compared to the finite element simulation.(4)The system was built and tested: some important functions of the digital twin system were developed,and for the visualization of laser subtractive manufacturing,an octree algorithm was used to reconstruct the surface mesh of the workpiece to realize the visualization of high-precision material removal with low performance consumption;the user interface of the system was built using UGUI components to integrate the simulation of geometric level and physical level in(2)(3);the digital twin virtual monitoring system was tested by example using the femtosecond laser 5-axis machining platform,and the test results showed that the system runs with low hardware resource consumption,good display performance,and excellent real-time monitoring performance to meet the design and use requirements.