Abstract: Based on the analysis of industrial Ethernet technology, this paper elucidates the working principle, modeling method, and message performance requirements of the IEC 61850 protocol, introduces the communication mechanism of applying the IEC 61850 standard in industrial Ethernet, and explores the design method of introducing the RTPS model into substation automation systems. Keywords: IEC61850; Ethernet; RTPS 1 Introduction The expansion of functions and scale of substation automation systems has accelerated the application of industrial Ethernet technology in power communication. The real-time performance, stability, versatility, and openness of industrial Ethernet technology, as well as its high transmission rate and large capacity, are unmatched by LonWorks fieldbus and CAN bus technologies. The expansion of functions and scale of substation automation systems, and the trend of substation automation technology developing towards high-voltage and ultra-high-voltage substation systems, have greatly accelerated the application of industrial Ethernet technology in power communication. The IEC 61850 protocol is the only communication standard system proposed by the International Electrotechnical Commission for future substation automation systems. Its application research in industrial Ethernet technology will have a profound impact on further improving the level of substation automation. 2 Analysis of Industrial Ethernet Technology Based on TCP/IP Protocol 2.1 Physical Layer and Data Link Layer Ethernet has become the mainstream technology and market, and with the gradual maturity of Fast Ethernet and Gigabit Ethernet technologies, network bandwidth is no longer a limiting factor for substation automation system applications, and the problem of transmission delay randomness caused by collisions has been mitigated [1]. Industrial Ethernet is a switched local area network based on the IEEE 802.3 standard. It uses the CAMA/CD protocol as the MAC and physical layer specification. Data transmission has the characteristics of "listening first, then detecting, and finally random retransmission". Therefore, some people believe that the randomness of the delay caused by Ethernet for "real-time" information transmission is unpredictable and cannot meet the needs of real-time systems. To this end, the Electric Power Research Institute (EPRI) conducted a performance comparison test on Ethernet and 12M token passing Profibus network. The results showed that 10M switched Ethernet can fully meet the "real-time" requirements of substation automation system network communication and is faster than 12M token passing Profibus network. 2.2 The network and transport layers use the de facto standard TCP/IP protocol as the high-level interface for the substation's Intelligent Equipment Devices (IEDs). This ensures that data transmission and reception by the IEDs within the substation can be performed using TCP/IP, guaranteeing good interoperability through standard data access methods. In this way, the monitoring master station or remote dispatch center can obtain data from the substation via a wide area network or even the Internet using the TCP/IP protocol. 3 Working Principle of IEC 61850 3.1 Advantages of IEC 61850 Protocol IEC 61850 protocol is the foundation of future seamless remote communication system and the only international standard for future substation automation system. It has outstanding advantages: (1) It uses object-oriented UML unified modeling technology; (2) It adopts a distributed and layered structure system; (3) It uses the Abstract Communication Service Interface (ACSI) and Specific Communication Service Mapping (SCSM) technology to separate communication service requirements from specific communication protocols, which is conducive to adapting to the continuous development of communication technology; (4) The independence of abstract modeling and specific implementation, the independence of service and communication network, and the applicability to multiple transmission protocols such as TCP and OSI; (5) It can realize interoperability between intelligent electronic devices; (6) It implements the Substation Configuration Description Language (SCL) based on XML technology; (7) Data and services are self-descriptive, which is convenient for system expansion. 3.2 Working Principle IEC 61850 works based on client/server mode. The actual equipment is modeled as data objects and services according to the IEC 61850 standard and represented by ACSI. The client sends a service request and receives a service confirmation that has been processed in the server; the client can also receive a report from the server. All service requests and responses are communicated by the communication protocol stack [2]. The working principle block diagram is shown in Figure 1 [align=center] Figure 1 Application layer information interaction principle block diagram[/align] IEC 61850 has two mechanisms for exchanging information between devices: push mechanism and query mechanism. The setting value or operation adopts the "push mechanism". In the case of the enhanced model, the event active reporting method is adopted, and the "read" value is selected from one or more data by name. The communication system provides means to prevent a single computer in the network from connecting to any device and to observe and modify all information of any device. There are multiple access methods to restrict the "visibility" of the device or device-specific data. The operator must not change specific settings. 3.3 Data and service modeling based on IEC 61850 protocol The IEC 61850 standard adopts object-oriented modeling technology to model each IED in the substation based on the client/server structure data model [3]. Each IED contains one or more servers, and each server itself contains one or more logical devices. Logical devices contain logical nodes, and logical nodes contain data objects. A logical node is an object composed of data and methods; it is the smallest callable sub-function and the smallest logical entity that exchanges data with other special logical entities. Data objects are named instances of common data classes composed of data attributes. From a communication perspective, the IED also acts as a client. Any client can communicate with the server through ACSI to access data objects, as shown in Figure 2. [align=center] Figure 2 Data Model Layered Structure[/align] ACSI provides system data objects and service objects, implementing a data and service-independent modeling approach. It establishes a model of the communication services that standard-compliant servers must provide, including server models, logical device models, logical node models, data models, and dataset models. Through ACSI, clients are mapped from the dedicated communication service mapping SCSM to the specific communication protocol stack used, such as the Manufacturing Message Specification (MMS), effectively resolving the contradiction between standard stability and future network technology development; that is, when network technology changes, only the SCSM needs to be modified. 4 IEC 61850 Message Performance Requirements PICOM (Piece of Information for Communication) is a description of exchanged data transmitted between two logical nodes through a defined logical path and with defined communication attributes [4]. PICOM can be divided into 7 message types, and their attribute range is composed of performance levels. Total transmission time is the total transmission time of the message, including the processing time required at both ends. It starts counting from the moment the sending end puts the data content on the top of the transmission stack until the moment the receiving end retrieves the data from the transmission stack. Fast messages: These messages typically contain simple binary codes of data, commands, or single messages. For example, "trip", "close", reclosing command, "start", "stop", "lock", and "unlock". Among them, "trip" is the most important fast message. For performance level P1, the total transmission time should be half a cycle sequence, set at 10ms; for performance level P2/3, the total transmission time should be less than 1/4 cycle sequence, set at 3ms. For other fast messages, the total transmission time should be less than or equal to 100ms for performance level P1; for performance levels P2/3, the total transmission time should be one cycle, set at 20ms. Medium-speed messages: Occurrence time is important, but transmission time is irrelevant. These messages have their own clock and contain a time stamp set by the transmitter. The receiver will react normally after a certain internal time delay and then perform calculations according to the time given by the time stamp. The total transmission time should be less than 100ms. Typical status information belongs to this category. Low-speed messages: May require time stamps and are generally used for low-speed automatic control, event logging, reading and writing setpoints, and system data descriptions. The total transmission time should be less than 500ms. Data transmission messages: Such as the output data of digital transmitters and digital instrument transformers. It includes a continuous synchronous data stream from IED outputs, interspersed with data from other IEDs. The total transmission time should be controlled within 10ms. File transfer messages: These messages are used to transmit large recorded data files and should generally be divided into finite-length message blocks for network transmission and use. The file type PICOM is generally greater than or equal to 512 bits, and the transmission time is not limited, but generally requires greater than or equal to 1000ms. Time synchronization message: synchronized with the internal clock of the IED, different levels of time synchronization are required according to the purpose, such as event time stamping or sampling accuracy of raw data. Command message with access control: this command can be issued by the local or remote human-machine interface HMI, which requires high security and must be accompanied by a password and authentication process. 5 RTPS model design Publisher/Subscriber model is a network data distributed model, which has been widely used in distributed systems, such as Web data publishing, network management, etc. It has the characteristics of information and device address independence, point-to-multipoint transmission and event-driven transmission [5]. Real-time distributed control system has a quantitative requirement for data transmission time, that is, the transmitted data must be transmitted to the destination within the specified time, otherwise it will cause system errors and functional failures. In addition, the reliability of connectionless mode is not strong, and it is necessary to introduce a control mechanism in the data stream to ensure reliable data transmission. The RTPS model (real-time publisher/subscriber model) is thus generated, as shown in Figure 3. [align=center]Figure 3 RTPS Model[/align] The RTPS model runs on a real-time operating system that supports multi-threaded priority processing. The Publisher output introduces three parameters: {SqNum, hTime, RtNum}, representing the data sequence number, data lifetime, and retransmission counter, respectively. The Subscriber input introduces a timeout period (TimeOut) and a data sequence number check. The retransmission interval after the Publisher publishes information is t = 4 + (1, n)R⁻¹, where 2 ≤ n ≤ 9. R represents the number of retransmissions. Data is retransmitted at a decreasing frequency until time t is greater than hTime. Simultaneously, the Subscriber receives data from the buffer sequentially. If it finds a disordered or incorrect data sequence number, it immediately initiates a query function, queries the Publisher for the unreceived data sequence, and returns a response. The following points must be noted when designing the RTPS model: (1) Use multi-threading mechanism to handle thread priority and access to shared data to improve the performance of the system itself and the real-time response performance of the service; (2) Design and manage the buffer structure reasonably to realize the parallel operation of the two buffer channels of input and output stream buffers without affecting each other; (3) Allocate the record buffer space reasonably according to system requirements and actual memory size; (4) Design the event-driven module scientifically to realize timely and effective collection and release of data; (5) Continuously optimize the model design to achieve the best utilization of memory and CPU resources. 6 Conclusion: Industrial Ethernet has become the main direction of control system network development. We should make full use of new and old information network technologies to improve the function of Ethernet. IEC61850 is the main direction of research on the application of substation automation communication network protocol. Its in-depth application research in industrial Ethernet will greatly promote the development of Ethernet technology. References: [1] Wu Zaijun, Hu Minqiang. Research on substation automation system based on IEC 61850 standard. Power Grid Technology [J]. 2003 (10). [2] Ru Feng, Xia Chengjun, Xu Yang. Discussion on the application of IEC 61850 standard in substation automation system [J]. Jiangsu Electrical Engineering. 2004 (3). [3] National Technical Committee on Standardization of Power System Control and Communication. IEC 61850 series of standards for substation communication networks and systems (translation compilation). November 2002. [4] Xu Lizi. Requirements of IEC 61850 for message performance of substation automation system. Power Grid Technology [J], 2002 (11). [5] Sun Junping, Sheng Wanxing, Wang Sunan. Research and implementation of real-time publisher/subscriber model based on Ethernet [J]. Journal of Xi'an Jiaotong University. 2002 (12).