面向智慧企业的工业认知网络体系架构、设计方法与应用验证

Industrial Cognitive Network (ICN) as a next generation automation system is focused on knowledge automation for the process industry, which is based on Internet of Things (IoTs), cloud computing, big data technologies with the properties of self-awareness, self-calculation, self-regulation, self-organization and self-execution. In order to meet the needs of global optimization and overall lifecycle management, and solve the problems of cross-layer coordination and cross-domain integration in process industry, research on the theory and methodology of ICN for process industry is urgent. According to the characteristics of mass semantic information, knowledge updating automation and contextualized organization and operation in ICN, three scientific issues are resolved, which includes “interoperability of heterogeneous knowledge resources ”, “cooperative computation under hybrid computation environment”, “hybrid transportation of control-flow, management-flow and knowledge-flow under heterogeneous dynamic network environment ”. Methods of IP based heterogeneous interconnection, hybrid flows transportation guarantee, semantic integration, distributed cooperative computation and self-organization service based on knowledge reasoning are proposed, respectively. The architecture of ICN and protocols including data connection layer, transport control layer, knowledge management layer and knowledge application layer are developed. A series of prototype design tools for ICN are proposed. A system of ICN for the process industry is constructed and the effectiveness is justified in real process industry.

工业认知网络是基于物联网、云计算、大数据等技术构建的具有自感知、自计算、自调节、自组织和自执行等流程工业知识自动化功能的新一代自动化系统。为满足我国流程工业对全局优化和全生命周期管理的需求，解决跨层协调和跨域集成困难的问题，研究面向流程工业的工业认知网络理论和方法体系。针对工业认知网络海量信息语义化、知识更新自动化、组织运行情境化等特征，在解决“流程工业多维、异构、时空多尺度知识互操作”、“混杂计算环境下知识网络分布式协同计算”、“异构动态网络环境下的控制、管理和知识流混合传输” 三项核心科学问题的基础上，突破基于IP的异构网络互联、混流传输保障、语义化数据集成、分布式协同计算、面向知识推理的服务自组织等技术难点，提出工业认知网络体系架构，设计包括数据链接层、传输控制层、知识管理层和知识应用层的协议体系；研发工业认知网络设计工具原型；建设面向流程工业过程的工业认知网络系统并开展工业化验证。

流程工业对知识自动化需求迫切，工业认知网络是实现流程工业知识自动化的新一代自动化系统。为满足我国流程工业对全局优化和全生命周期管理的需求，项目组重点从“体系架构”、“设计方法”、“原型系统及应用验证”三个层面深入且系统地对工业认知网络进行了研究，具体突破了工业网络（异构网络互联、无线组网、混流业务传输）、可重构计算（资源管控、实时调度、计算平台）、知识自动化（语义建模、标注、推理）等技术难点，提出包括互联、感知和组织的三层工业认知网络体系架构，建设面向流程工业过程的工业认知网络原型系统并开展工业化验证，最终形成了一套面向智慧企业的工业认知网络关键技术体系。.综上，项目所取得的研究成果覆盖了项目任务书的全部内容，并完成既定的项目指标。

内容获取失败，请点击重试