王健全, 马彰超, 孙雷, 张超一, 李卫. 工业网络体系架构的演进、关键技术及未来展望[J]. 工程科学学报, 2023, 45(8): 1376-1389. DOI: 10.13374/j.issn2095-9389.2022.08.01.003
引用本文: 王健全, 马彰超, 孙雷, 张超一, 李卫. 工业网络体系架构的演进、关键技术及未来展望[J]. 工程科学学报, 2023, 45(8): 1376-1389. DOI: 10.13374/j.issn2095-9389.2022.08.01.003
WANG Jian-quan, MA Zhang-chao, SUN Lei, ZHANG Chao-yi, LI Wei. Evolution, key technology, prospects, and applications of industrial network architecture[J]. Chinese Journal of Engineering, 2023, 45(8): 1376-1389. DOI: 10.13374/j.issn2095-9389.2022.08.01.003
Citation: WANG Jian-quan, MA Zhang-chao, SUN Lei, ZHANG Chao-yi, LI Wei. Evolution, key technology, prospects, and applications of industrial network architecture[J]. Chinese Journal of Engineering, 2023, 45(8): 1376-1389. DOI: 10.13374/j.issn2095-9389.2022.08.01.003

工业网络体系架构的演进、关键技术及未来展望

Evolution, key technology, prospects, and applications of industrial network architecture

  • 摘要: 基于工业互联网应具备的特征,结合现有工业互联网总体现状,分析总结传统工业自动化封闭式五层架构存在的问题。首先,提出支持数据高效流转的云、边端新型工业网络协同架构,架构的变革对现有技术提出挑战,同时也为传统自动化系统提供了新的机遇。其次,在总体架构的基础上,提出适配新型工业网络基础架构的两项关键技术。5G–时间敏感网络(Time-sensitive networking, TSN)协同传输技术,包括5G–TSN异构网络融合架构、网络时钟适配机制以及基于软件定义网络(Software defined network,SDN)的融合管控和资源调度三部分技术内容;基于确定性网络的云化可编程逻辑控制器(Programmable logic controller,PLC)技术包括工业控制虚拟化和5G云化工业控制技术两部分技术内容。基于此,提出自主设计面向实时运动控制的5G云化PLC与EtherCAT融合系统,以及面向实时运动控制的EtherCAT与TSN融合系统试验平台,并验证了新型工业网络架构的科学性和合理性。最后,对未来网络、控制、计算一体化工业自动化系统中的高效性、可靠性和安全性之间的融合问题及潜在解决方案进行了探讨。

     

    Abstract: Based on a summary of the characteristics of the industrial Internet, combined with the overall status of the existing industrial Internet, this paper analyzes the problems of the traditional industrial automation closed five-tier architecture and concludes that the current industrial Internet remains in a stage of technological development and maturation, which restricts the promotion and standardization of China’s intelligent manufacturing level to some extent. In the future, the industrial Internet will break the original data hierarchy structure, break the data barrier, and realize the development of intelligent manufacturing toward intelligent, flattening, lightweight, and green. First, this paper proposes a new industrial network collaboration architecture on the cloud side that supports efficient data flow. The proposed architecture is a flat, platform-based structure that realizes cloud-side collaboration and the integration of control, computing power, and network. Second, on the basis of the overall architecture of the industrial network, two key technologies are proposed to adapt the new industrial network infrastructure. One is 5G–time-sensitive networking (TSN) collaborative transmission technology. TSN realizes the interconnection and integration of heterogeneous networks in the industrial field based on the data link layer, and 5G–TSN collaborative transmission has become an important evolution trend of the intelligent factory network. Three key technical contents are introduced: 5G–TSN heterogeneous network convergence architecture, network clock adaptation mechanism, and software defined network (SDN)-based integration management and resource scheduling. The other key technology is cloud PLC technology based on a deterministic network, and the virtualization and cloud control system is the basis for breaking the original closed industrial control system. On the one hand, cloud-based hardware resources can be used to achieve one-to-many control, saving a large amount of industrial control equipment deployment investment costs. More importantly, the centralized control system can achieve unified control and optimization of global resources. This paper introduces the virtualization 5G cloud chemical industry control technology with two parts: technical content and technical route. Third, this paper proposes a cloud-based allocation of control resources and a cloud–network integration collaboration scenario and designs the 5G cloud-based programmable logic controller (PLC) and EtherCAT fusion system and the EtherCAT and TSN fusion system for real-time motion control. Through the testing of the end-to-end delay and cross-network time synchronization accuracy of the actual system, the current end-to-end delay of network transmission is less than 5 ms, the cross-network time synchronization error is between 100–400 us, and the accuracy is less than 100 ns These performance indicators reach the current industry-leading level. The test platform verifies the scientificalness and rationality of the new industrial network architecture. Finally, the integration problems and potential solutions of efficiency, reliability, and security are discussed in the future industrial automation system integrating network, control, and computing.

     

/

返回文章
返回