基于数字孪生的激光加工零件表面温度监控系统的构建

Construction of surface temperature monitoring system for laser machining parts based on digital twin

数字孪生是一个旨在物理空间与虚拟空间之间建立实时映射的概念。为拓展数字孪生技术在激光加工领域的应用,以纳秒激光打标机加工7075铝合金为例,阐述了有限元仿真在激光加工领域的重要作用,但由于有限元仿真计算量较大且目前计算机的运算水平有限,存在仿真计算时间较长的问题,无法满足数字孪生技术所需的实时映射。为此,提出了利用CGAN(conditional generative adversarial network,条件生成对抗网络)模型代替有限元仿真的方法。该方法先利用图像化处理后的工况图像和由有限元仿真获得的温度云图对CGAN模型进行训练,随后将训练好的CGAN模型封装,用于构建基于数字孪生的激光加工零件表面温度监控系统。在完成数字孪生温度监控系统虚拟端的搭建后,使用MQTT(message queuing telemetry transport,消息队列遥测传输)通信协议与实体端进行数据交互,实现数字孪生系统的远程监控与操作。基于数字孪生的激光加工零件表面温度监控系统实现了零件表面温度的快速仿真计算,解决了有限元仿真计算时间长、无法实现实时映射的问题,基本能够满足激光加工零件表面温度的监控与预测,在激光加工温度监测领域具有一定的参考价值。

Digital twin is a concept that aims to establish a real-time mapping between physical space and virtual space.In order to expand the application of digital twin technology in laser machining,taking the processing of 7075 aluminum alloy by nanosecond laser marking machine as an example,the important role of finite element simulation in laser machining was expounded.However,due to the large computational amount of finite element simulation and the limited computing level of current computers,the simulation calculation time was long,which could not meet the real-time mapping required by digital twin technology.Therefore,a method of replacing finite element simulation with CGAN(conditional generative adversarial network)model was proposed.In this method,the CGAN model was trained by using the working condition images after image processing and the temperature cloud maps obtained by finite element simulation.Then,the trained CGAN model was packaged for building the surface temperature monitoring system for laser machining parts based on the digital twin.After completing the construction of the virtual end of the digital twin temperature monitoring system,the MQTT(message queuing telemetry transport)communication protocol was used to interact with the physical end to realize the remote monitoring and operation of the digital twin system.The surface temperature monitoring system for laser machining parts based on digital twin achieves the rapid simulation calculation of the surface temperature of parts,which solves the problems of long calculation time and inability to achieve real-time mapping by the finite element simulation.It can basically meet the monitoring and prediction of the surface temperature of laser machining parts,and has certain reference value in the field of laser machining temperature monitoring.