The Real-time Transformation Of Industrial Ethernet

With the industrial automation technology and information technology growing, the establishment of a unified and open communication protocols and network and the realization of information network and control network integration become pressing problems of network control system. Industrial Ethernet, which has a great advantage in applicability, feasibility and development of the technical and practical applications, is an effective way to solve the above problems and also becomes the direction of the development of networked control system. However, the uncertainty of the network and no synchronization of the system device clock become obstacles restricting network control system. In order to apply Ethernet-based network control system to industrial control field which has relatively high real-time requirements, network transmission delay and clock synchronization have become the key problems.This thesis analyzes the communication needs and model of network control system. On this basis, it discusses the culprit and solution of deterministic and real-time problem of Ethernet in the applications of network control system. In addition, through analysis of the maximum delay of shared Ethernet and switched Ethernet, using full-duplex switched Ethernet technology for industrial applications.Then this paper proposes a delay prediction method based on the combination of RBF neural network and weighted linear factors delay prediction, utilize IEEE1588 precision clock synchronization protocol to synchronize network clock and adopt compensation control to eliminate the negative impact delay on the system to the greatest extent. Finally, it is verified in the matlab / simulink / truetime simulation environment.