关键词: 信息物理系统, 边缘智能体, 阀栓压力检测系统

Abstract: With the advancement of Industry 4.0 and smart manufacturing,Cyber-Physical Systems (CPS) have gradually become one of the core technologies in modern manufacturing. As an important component of CPS,edge agents establish a low-latency computing layer between field devices and control systems,effectively enhancing production efficiency and system response speed. To address the challenges of heterogeneous device integration,network bandwidth bottlenecks,and real-time control in industrial automation,a novel physical layer architecture for edge agents was proposed,and the new PAG series edge agent controllers were successfully developed. This architecture not only supports the integration of multiple industrial protocols but also incorporates semantic recognition technology to optimize data exchange and collaboration between devices. In practical applications,the system was successfully applied to the metal valve stress testing system at one company,where CPS transformation improved the intelligence and informatization of the production process. The system performs real-time data collection,analysis,and processing through the edge agent,enabling automatic monitoring of devices,data recording,and quality control,significantly improving the production efficiency and testing accuracy. Experimental results show that the transformed system reduced the production cycle by approximately 36 % and improved system stability. The study demonstrates that the proposed novel physical layer architecture of edge agents can effectively tackle various challenges in complex industrial environments and provide a feasible technical solution for the intelligent upgrade of industrial field devices.

Key words: cyber-physical systems, edge agents, valve stress testing system