Abstract: In the data-driven intelligent manufacturing workshop production environment,data processing,storage and mining become more and more important.Traditional single-flow data processing architecture has serious delays lack of real-time control and inefficient invocation,which cannot adapt to characteristics of aviation structural parts processing such as more,faster and better.Based on the characteristics of aviation manufacturing environment,a new Phi data processing architecture combining edge computing and feedback control was proposed.By integrating streaming data and batch data processing,the real-time online monitoring and historical data traceability functions were satisfied;thought the edge computing unit in the equipment layer and relay layer,the traffic monitoring and fast feedback could be supported.In virtue of Hadoop ecosystem and microservice development model,the proposed data processing architecture could meet requirements of massive data storage,access,and flexible calls in different manufacturing scenarios better.The verification was carried out in the simulation workshop.The experimental results showed that the new Phi architecture had been effectively applied in the scenarios of data monitoring,temperature compensation and low-latency control,and the stability and scalability of the system were improved,which provided an important platform support for promoting the development of aviation manufacturing digital workshop to high quality,high efficiency and high stability.

Key words: intelligent manufacturing, digital workshop, multi-scenario application, data processing architecture, edge computing, aviation digital workshop

摘要： 在数据驱动的智能制造车间生产环境下,数据的处理、存储及挖掘变得愈发重要。传统单一流向的数据处理架构延时严重、缺乏实时控制、调用效率低,无法适应航空加工多、快又好的制造特点。鉴于此,基于航空制造的环境特点,提出一种结合边缘计算和反馈控制的新型Phi数据处理架构。通过融合数据流处理和批处理,满足信息的实时监控和历史数据追溯功能;通过布置在设备层和中继层的边缘计算单元,满足生产现场数据的流量监控、快速反馈等需求;更进一步,借助Hadoop生态技术和微服务开发模式,使数据处理平台更适应于海量工业数据的存储、访问及不同场景下的柔性调用。最后,在实验室搭建的模拟车间进行了验证,结果表明新型Phi架构在数据监控、温度补偿和低延时控制等场景中均得到了有效应用,并提升了系统的稳定性和扩展性,为推动航空制造数字化车间向高质量、高效率和高稳定性发展提供了重要的平台支撑。

关键词: 智能制造, 数字化车间, 多场景应用, 数据处理架构, 边缘计算, 航空数字化车间