基于深度学习的“微型灯塔工厂”物理和虚拟交互平台设计

针对电子产品装配过程中存在的元器件精确定位难、细微缺陷检测慢、工程可视化程度低等技术难题，设计和搭建“微型灯塔工厂”工程示范厂房，采用Tecnomatix 平台搭建微型灯塔工厂数字孪生体，解决制造过程中可视化程度低的问题；采用LabVIEW平台及深度学习算法实现CO-WORK及人机工程的快速检测和诊断，解决了生产运维难的问题；通过融合多尺度注意力权重信息的智能检测识别算法，解决了微型电子电路板检测速度慢和识别率低的难题。

Design of a “Micro Lighthouse Factory” Physical and Virtual Interaction Platform Based on Deep Learning

In view of the technical challenges such as difficulty in precise positioning of components,slow detection speed for minor defects,and low engineering visualization level in the assembly process of electronic products,the “Micro Lighthouse Factory” demonstration factory was designed and built.The Tecnomatix platform was used to build a digital twin of a miniature lighthouse factory,to solve the problem of low visualization in the manufacturing process.The LabVIEW platform and deep learning algorithm were used to achieve rapid detection and diagnosis of CO-WORK and human-machine engineering,the problem of difficult production,operation and maintenance was solved.The intelligent detection and recognition algorithm integrating with multi-scale attention weight information was used to solve the problems of slow detection speed and low recognition rate in micro electronic circuit boards.