Abstract: Focusing on the requirements of cloud-level mining and edge-level computing of real-time manufacturing big data generated from Industrial Internet of Things (IIoT) devices in smart factories,an IIoT architecture for smart factories based on cloud-edge interplay was proposed,together with the edge nodes configuration method and the cloud application services configuration method.Under this architecture,time-insensitive and time-sensitive manufacturing data computing tasks were executed by cloud server and IoT edge nodes respectively.Considering human workers,machine tools,physical objects,and other manufacturing resources in a smart factory were endowed with different levels of intelligence,the key technologies to realize the human-machine-product-symbiosis-based autonomous production control were researched from the perspectives of production planning,execution,monitoring,and optimization.From a case study,it was concluded that cloud-edge interplay-based IIoT architecture and autonomous production control technologies could not only make manufacturing data processing and analysis from both cloud-level and edge-level,but also achieve the real-time production status monitoring and autonomous production control of smart factories,which provided a basis for realizing smart and autonomous production in factories.

Key words: smart factory, industrial Internet of things, cloud-edge interplay, human-machine-product symbiosis, autonomous production control

摘要： 针对智能工厂底层大量物联设备产生的海量制造数据在云端处理分析和边缘端实时计算两个层面的需求,提出一种基于云—边协同的智能工厂工业物联网架构以及边缘节点、云端应用服务的配置方法。在该架构下,时间不敏感型与时间敏感型制造数据处理任务分别由工厂云服务器、工业物联网边缘节点完成。考虑智能工厂内人、机、物等生产主体均具有智能,进一步从生产计划、执行、监控、优化的逻辑视角,阐述了基于“人—机—物”共融的智能工厂自治生产管控关键技术。通过案例研究得出,所提出的智能工厂工业物联网架构可在云端与边缘端两个层面进行制造数据处理分析,并实现智能工厂实时生产状态监控与自治生产过程响应控制,为实现制造工厂智能化、自治化生产提供技术支撑。

关键词: 智能工厂, 工业物联网, 云&mdash, 边协同, 人&mdash, 机&mdash, 物共融, 自治生产管控