Industrial network systems
Industrial network systems are multi-dimensional dynamic systems that integrate industrial control and information communication. These systems highly integrate advanced technologies such as automatic control, computer technology, and communication network technology, and achieve synergy between information systems and industrial physical processes through networks. This results in optimized production, simplified processes, and maximized efficiency, making them crucial for promoting the digital, networked, and intelligent integrated development of industrial manufacturing.
Industrial network systems integrate ubiquitous sensing, adaptive transmission, and collaborative control, possessing significant advantages such as networked structure, on-site control, and decentralized functions. They are the core for realizing the intelligence and interconnectivity of industrial cyber-physical systems. However, the integrated design of sensing-communication-control faces numerous challenges. For example:
1) Perception: Heterogeneous terminals are difficult to integrate under resource-constrained conditions;
2) Transmission: The communication environment is complex and ever-changing, requiring high time determinism and transmission reliability;
3) Control: Information and control are interactively coupled in a network environment;
...
To meet these challenges, the analysis and design of the system must meet the industrial system’s requirements for real-time, reliable, ubiquitous communication and agile, precise, and collaborative control. By comprehensively utilizing control, communication, sensing, and computing theories, and combining control optimization theory with communication network design methods, a new generation of industrial network systems with adaptive capabilities to system dynamics and networks can be formed.
To achieve integrated design, it is essential to clearly represent the complementary and mutually restrictive coupling relationship among sensing, transmission, and control. This lays the foundation for revealing the interactions among these three elements and improving the overall performance of industrial network systems. As shown in Figure 1, this paper, based on the integrated framework, briefly describes the connotation and main characteristics of industrial network systems, and analyzes the challenges and key issues faced by "sensing-transmission-control integration." The paper reviews the current research status and progress both domestically and internationally from three aspects: distributed fusion estimation of heterogeneous networks under non-ideal communication, adaptive transmission for sensing and control, and collaborative control of complex systems in network environments.
Figure 1. Integrated sensing-transmission-control framework of an industrial network system.
How to achieve joint design of perception, transmission, and control in industrial network systems? Traditional control theory typically considers how to utilize feedback information to achieve specific control objectives, assuming perfect communication and information transmission or conforming to certain transmission models; while communication theory mainly focuses on how to reliably and quickly transmit information from the source to the terminal, without paying much attention to the content and purpose of the transmitted information. This has led to the independent design of control and communication mechanisms and algorithms in the past, constraining the improvement of the overall performance of the network system. However, harsh industrial environments lead to unpredictable system states, loss or timeouts of some information during transmission, and incomplete information affecting control performance. To solve this problem, joint design is an effective approach.
This paper explores an integrated design of sensing, communication, and control, proposing a layered architecture for an industrial network system as shown in Figure 2. The bottom layer deploys edge estimation terminals, responsible for preprocessing and forwarding raw sensing data. The aim is to reduce the energy and communication resources consumed in directly transmitting sensing data to the fusion center, decrease conflicts and collisions, and thus improve the reliability and real-time performance of information interaction. Since the sensing range of the terminals is limited and the sensing information of adjacent terminals is highly correlated, the terminals are clustered according to geographical location, and one edge estimation terminal is deployed in each cluster. Furthermore, all edge estimation terminals constitute the middle layer, utilizing edge computing technology to filter, extract, and fuse the received raw sensing data to remove redundant information, improve sensing accuracy, and transmit local estimates to the fusion center. Based on this architecture, the total cost of sensing, transmission, and control in the industrial network system can be minimized, enabling the joint design of resource-constrained adaptive network resource scheduling and control laws.
Figure 2. Layered architecture of industrial network system
The integrated design of sensing, communication and control enables the sensing process to provide information support for control, the transmission process to realize real-time and reliable interaction of sensing information, and the control process to provide control decisions to ensure the stable and efficient operation of the system, thereby improving the collaborative sensing/control capabilities of industrial network systems.
Currently, research on collaborative design is still in the exploratory stage. With the rapid development of control science, communication science, and computing science, interdisciplinary research is making continuous progress, and industrial network systems will also experience rapid development. New "sensing-transmission-control integration" paradigms will continue to emerge, and the methods and implementations of collaborative optimization control will become increasingly diverse, with their application scope constantly expanding. It is foreseeable that achieving optimized control of industrial systems through the optimization and coordination of various links in network control systems will become an important research direction, and industrial network systems will play an increasingly important role in the intelligentization and informatization of production processes!
Disclaimer: This article is a reprint. If it involves copyright issues, please contact us promptly for deletion (QQ: 2737591964 ) . We apologize for any inconvenience.
