A method and system for fast communication between devices in a flexible DC control and protection system

By inserting the shortest Ethernet protection frame for fault information into a gigabit point-to-point Ethernet link in a flexible DC transmission system, the problem of untimely communication between the valve protection device and the valve interface device is solved, enabling rapid fault identification and interlocking, and improving the real-time performance and accuracy of the system.

CN123122032APending Publication Date: 2026-11-17SIEYUAN QINGNENG ELECTRICAL & ELECTRONICS CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611387780.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-08
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In flexible DC transmission systems, the fault diagnosis and interlocking command transmission between valve protection devices and valve interface devices suffer from multiple transmission levels and untimely communication, making it difficult to transmit phase-specific temporary interlocking and valve system rapid interlocking commands in a timely manner.

Method used

Using a gigabit point-to-point Ethernet link, the shortest Ethernet protection frame carrying fault phase information and fault arm number information is inserted at the end of the regular service frame through a field-programmable gate array, reducing the transmission layers caused by independent contacts, cables and additional interfaces, and achieving fast communication.

Benefits of technology

It improves the real-time performance and accuracy of protection communication between devices, simplifies communication interfaces, and ensures rapid triggering and recovery of fault blocking and reset.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122032A_ABST
    Figure CN123122032A_ABST
Patent Text Reader

Abstract

This invention provides a method and system for rapid communication between devices in a flexible DC transmission control and protection system, relating to the field of flexible DC transmission technology. The method includes: when a temporary overcurrent fault persists, appending the shortest Ethernet protection frame after each regular service frame is sent; stopping the transmission of the shortest Ethernet protection frame after the temporary overcurrent fault returns to normal; the receiving device's field-programmable gate array verifies the received shortest Ethernet protection frame, and if the verification is valid, parses the fault phase information and the fault arm number information, driving the corresponding arm to perform a blocking protection action based on the parsing result; based on the reception status of the shortest Ethernet protection frame within a continuous device communication cycle, when the number of consecutive unreceived cycles reaches the fault return determination cycle number, determining that the temporary overcurrent fault has disappeared and canceling the fault blocking status of the corresponding arm. This invention improves the real-time performance and accuracy of protection communication between devices.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of flexible DC transmission technology, and in particular to a method and system for rapid communication between devices in a flexible DC control and protection system. Background Technology

[0002] In flexible DC transmission systems, the converter valves consist of numerous insulated-gate bipolar transistor (IGBT) power modules connected in series. When an internal valve fault such as overcurrent occurs in a bridge arm, the fault current rises rapidly. The valve protection system, the 3-out-of-2 valve protection device, and the valve interface device need to complete fault identification, logic voting, and interlock command transmission within a short timeframe. Existing valve protection devices typically have two sets of independent output contacts for each 3-out-of-2 valve protection device. Each set of contacts is connected to the digital input module of the 3-out-of-2 valve protection device via an independent cable, while digital action signals are transmitted via fiber optic channels. The 3-out-of-2 valve protection device also needs to identify the phases of the six bridge arms and generate corresponding output logic for different bridge arm interface devices. Multiple sets of hardwired connections, fiber optic channels, and bridge arm output logic collectively increase the inter-device interfaces and transmission layers, making it difficult to transmit phase-specific temporary interlocking and rapid valve system interlocking commands in a timely manner. Summary of the Invention

[0003] This invention provides a method and system for fast communication between devices in a flexible DC control and protection system. It inserts the shortest Ethernet protection frame carrying fault phase information and fault arm number information at the end position of the current regular service frame. Therefore, regular service data and protection data can be transmitted in a unified communication interface without adding external hardware circuits, reducing the transmission layers caused by independent contacts, cables and additional interfaces.

[0004] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: In a first aspect, a method for fast communication between devices in a flexible DC control and protection system is provided, applied to a transmitting device and a receiving device connected via a gigabit point-to-point Ethernet link, the method comprising: The transmitting device transmits regular service frames carrying device operating conditions and cross-device heartbeats through the gigabit point-to-point Ethernet link according to the device communication cycle that matches the device operation cycle. The field-programmable gate array of the transmitting device acquires the overcurrent fault level of the bridge arm, determines the temporary overcurrent fault based on the overcurrent fault level of the bridge arm, and obtains the fault phase information and the fault bridge arm number information. Based on the fault phase information and the fault arm number information, after the current regular service frame is sent, the shortest Ethernet protection frame carrying the fault phase information and the fault arm number information is immediately sent through the same gigabit point-to-point Ethernet link. While the temporary overcurrent fault persists, the shortest Ethernet protection frame is sent after each regular service frame is sent. The sending of the shortest Ethernet protection frame stops after the temporary overcurrent fault is resolved. The field-programmable gate array of the receiving device verifies the shortest Ethernet protection frame received. When the verification is valid, it parses the fault phase information and the fault arm number information, and drives the corresponding arm to perform a blocking protection action based on the parsing result. Based on the reception status of the shortest Ethernet protection frame within the continuous device communication cycle, when the number of consecutive non-reception cycles reaches the fault reset determination cycle number, it is determined that the temporary overcurrent fault has disappeared and the fault blocking status of the corresponding bridge arm is revoked.

[0005] Secondly, a fast communication system between devices in a flexible DC control and protection system includes a valve protection device, a valve protection three-out-of-two device, a valve interface device, and a gigabit point-to-point Ethernet link connecting the valve protection device, the valve protection three-out-of-two device, and the valve interface device. The valve protection device includes a transmitting field-programmable gate array (FPGA). The transmitting FPGA is used to send regular service frames according to the device communication cycle that matches the device's operation cycle, obtain the overcurrent fault level of the bridge arm, obtain the fault phase information and fault bridge arm number information based on the overcurrent fault level of the bridge arm, and send the shortest Ethernet protection frame carrying the fault phase information and fault bridge arm number information after the current regular service frame is sent. The valve protection three-out-of-two device includes a receiver field-programmable gate array (FPGA), which is used to verify and parse the shortest Ethernet protection frame. The valve protection three-out-of-two device is used to complete logical voting based on the parsing result, issue a corresponding bridge arm locking command to the valve interface device, and determine whether the temporary overcurrent fault has disappeared based on the reception status of the shortest Ethernet protection frame within the continuous device communication cycle. The valve interface device is used to drive the corresponding bridge arm to perform a lockout protection action according to the lockout command, and to cancel the fault lockout state of the corresponding bridge arm when the temporary overcurrent fault disappears.

[0006] The above-described solution of the present invention has at least the following beneficial effects: This application uses the same gigabit point-to-point Ethernet link between the transmitting and receiving devices. It uses Ethernet II regular service frames to carry device operating conditions and cross-device heartbeats. After the field-programmable gate array (FPGA) determines a temporary overcurrent fault in a bridge arm, it inserts the shortest Ethernet protection frame carrying the fault phase information and the faulty bridge arm number information at the end of the current regular service frame. Therefore, regular service data and protection data can be transmitted within a unified communication interface without the need for additional external hardware circuits, reducing the transmission layers caused by independent contacts, cables, and additional interfaces. The shortest Ethernet protection frame is sent frame by frame during the fault duration. After the receiving end completes the validity verification, it directly parses the bridge arm position and drives the corresponding bridge arm to lock, so that the protection information can be transmitted in the direction of the faulty bridge arm.

[0007] After fault recovery, frame interpolation stops. The receiving end uses the absence of a protection short frame for two consecutive device communication cycles as the criterion for revoking the blocking, ensuring that fast triggering and fast recovery use the same protection short frame for closed-loop connection. This collectively improves the real-time performance and accuracy of protection communication between devices and simplifies the communication interface. Attached Figure Description

[0008] Figure 1 The architecture of a flexible DC converter valve protection system is shown.

[0009] Figure 2 The timing of the interpolated frame data between device A and device B is shown. Detailed Implementation

[0010] To make the technical solution of this application clearer, the specific embodiments are described in detail below with reference to the accompanying drawings. The terms "first" and "second" are used to distinguish similar objects and do not indicate the order or importance of the objects; the term "including" is an open-ended expression; the accompanying drawings are used to illustrate the relationships between devices, communication timing, and implementation effects, and do not represent actual component dimensions or signal ratios.

[0011] This embodiment is applied to a flexible DC transmission control and protection system. Flexible DC transmission, also known as VSC-HVDC, uses modular multilevel converter valves (MMC valves) as the converter valves. Valve protection devices are used to detect internal valve faults such as overcurrent in the bridge arm. A two-out-of-three valve protection device performs logical voting on protection information from the valve protection device. A valve interface device receives interlocking commands and controls the submodule drive unit to execute interlocking of the corresponding bridge arm. Field-programmable gate arrays (FPGAs) are installed in both the transmitting and receiving devices to perform fault detection, message transmission and reception, message verification, and protection linkage on the hardware side.

[0012] Figure 1The architecture of a flexible DC converter valve protection system is shown. This architecture includes a protection host A, a protection host B, and a protection host C; two sets of valve protection 3-out-of-2 devices, designated 3-out-of-2 A and 3-out-of-2 B respectively; the architecture also includes a valve control host A, a valve control host B, a bridge arm control A, a bridge arm control B, and odd-interface and even-interface chassis housing the core board. The protection host is connected to the two sets of valve protection 3-out-of-2 devices, which are respectively connected to the valve control host and the core board within the interface chassis. The valve control host is connected to the corresponding bridge arm control. Figure 1 The device names and connection relationships are used to describe the communication scenarios between the valve protection, the valve protection three-out-of-two, and the valve interfaces.

[0013] The existing valve protection device is equipped with two sets of independent output contacts for each valve protection three-out-of-two device. Each set of contacts is connected to the digital input module of the valve protection three-out-of-two device via an independent cable, and transmits digital action signals through a fiber optic channel, forming a dual verification of hard-wired signals and digital action signals. Each valve protection three-out-of-two device is also equipped with two sets of parallel trip output contacts for the two valve control systems, so that commands can continue to be transmitted when any contact fails.

[0014] After the bridge arm overcurrent or phase-separated temporary blocking protection is triggered, the valve protection device first completes the fault identification, the valve protection three-out-of-two device then completes the logic voting and identifies the six bridge arm phases, and then issues the zone blocking command to the corresponding valve interface device. The valve interface device finally drives the submodule to execute the blocking. In this link, multiple sets of contacts, cables, fiber optic channels and output logic set for different bridge arms increase the number of interfaces and transmission layers, which is not conducive to the timely transmission of phase-separated temporary blocking and valve system rapid blocking information.

[0015] This embodiment establishes a gigabit point-to-point Ethernet link between the flexible DC control and protection devices, using 1000BASE-X or 1000BASE-T communication modes. The media access control layer frame structure is compatible with 100Mbps and 10Mbps Ethernet, with differences in physical layer timing and half-duplex conflict compensation mechanisms. Gigabit fiber optic communication between the flexible DC polarity control device and the valve control device uniformly adopts the Ethernet II standard frame. Substation event messages, sampled value messages, and manufacturing message specifications under the IEC 61850 system can all be transmitted based on this frame format. Since the protection short frames and regular service frames share existing gigabit links and the device's native FPGA resources, no additional external hardware circuitry is required during implementation, reducing the need for additional communication interfaces while retaining existing service configurations.

[0016] A complete Ethernet II line message includes a physical layer synchronization prefix and a media access control (MAC) data frame body. The physical layer synchronization prefix consists of a seven-byte preamble and a one-byte start-of-frame (FOF) character. All seven preamble bytes have a value of 0x55, and the FFO character has a value of 0xD5. The MAC data frame body consists of a six-byte destination MAC address, a six-byte source MAC address, a two-byte type field, a payload of 46 to 1500 bytes, and a four-byte CRC32 checksum. The preamble and FFO character are used for line synchronization; the destination and source addresses identify the two ends of the communication; the type field distinguishes the carried content; the payload carries device status, cross-device heartbeats, or protection points; and the CRC32 checksum is used by the receiving end to check frame integrity.

[0017] The transmission cycle of regular service frames matches the device's operation cycle and can be configured to 50 microseconds or 100 microseconds. In the given configuration implementation, each cabinet group uses a 100-microsecond communication cycle, periodically sending heartbeat messages and status interaction messages under steady-state conditions without changing the original service communication application logic. A regular service frame carries 64 groups of 32-bit basic interaction data, with a complete transmission time of 2.256 microseconds on the line. This cycle configuration provides an insertion position for protection short frames outside of regular service data transmission.

[0018] The FPGA of the valve protection device acquires the fault level used to characterize the overcurrent state of the bridge arm and performs temporary overcurrent fault identification based on the fault level. When a temporary overcurrent fault occurs in any bridge arm, the FPGA determines the phase and bridge arm number to which the fault belongs, forming fault phase information and fault bridge arm number information respectively. The two pieces of information jointly identify the bridge arm that needs to be locked, enabling the subsequent valve protection three-out-of-two device to identify the faulty bridge arm and form the corresponding output logic.

[0019] The transmitting end sets up a protection short frame interpolation channel within the FPGA and configures the transmission priority of protection short frames to be higher than that of regular service frames. While a regular service frame is still being transmitted, the transmitting end does not truncate it. Once the regular service frame has finished transmitting, the protection short frame interpolation channel is immediately opened, and the fault phase information and fault arm number information are written into the shortest Ethernet protection frame and transmitted. The protection short frame data channel is isolated from regular service data. Regular service frames continue to carry operating conditions and heartbeats, while protection short frames only carry the fault arm location, thus avoiding the mixing of content carried by long and short frames.

[0020] According to IEEE 802.3, the minimum effective media access control frame for Gigabit Ethernet is 64 bytes, and the minimum service payload is 46 bytes. When this shortest protection frame includes the physical layer synchronization prefix, the complete transmission time on the line is 0.576 microseconds. At the most unfavorable post-interpolation position, the sending end needs to first complete a regular service frame with a transmission time of 2.256 microseconds, and then send the shortest protection frame with a transmission time of 0.576 microseconds. The maximum transmission time of the two frames combined is 2.832 microseconds. This time is obtained by adding the transmission times of the two types of frames and is used to describe the line occupancy time after the protection short frame is inserted into the regular service frame.

[0021] During the fault duration, the FPGA of the valve protection device maintains the detected overcurrent fault level of the bridge arm. After each regular service frame is sent, an additional Ethernet shortest protection frame is sent. Therefore, the valve protection 3-out-of-2 device can repeatedly obtain the same faulty bridge arm location within a continuous communication cycle. After the fault level completely returns to normal, the transmitting end closes the protection short frame interpolation channel and stops sending protection short frames, while regular service frames are still sent according to the original communication cycle. The start and stop of protection short frames with the fault ensures that the fault status is directly reflected in whether protection short frames are continuously received, providing the same input for the receiving end's latch-up maintenance and reset determination.

[0022] After receiving the shortest Ethernet protection frame, the receiving FPGA first performs a frame integrity check based on the CRC32 check field in the frame. If the check fails, the frame is not used as the basis for bridge arm blocking. If the check succeeds, the protection short frame bypasses the regular service parsing thread, and the FPGA directly reads the fault phase information and fault bridge arm number information from the service payload. This receiving path reduces the number of protection points waiting for regular service parsing, while using frame integrity check to exclude invalid protection short frames.

[0023] In the link consisting of the valve protection device, the valve protection 3-out-of-2 device, and the valve interface device, the valve protection device sends a protection short frame. The valve protection 3-out-of-2 device receives the fault phase information and the fault bridge arm number information, completes logical voting, and then sends the corresponding bridge arm's partitioned interlocking command to the valve interface device. The FPGA of the valve interface device parses the received bridge arm position, and the control submodule driver unit executes the interlocking of the corresponding bridge arm. Since the interlocking command corresponds to the fault bridge arm number in the protection short frame, the receiving side can perform partitioned interlocking on the faulty bridge arm, avoiding the treatment of a single bridge arm fault as a global interlocking.

[0024] Figure 2The timing diagram illustrates the frame insertion data between device A and device B. After device A finishes transmitting the regular service frame at the time of the fault trigger, it appends a protection short frame. Device B receives the protection short frame and continues processing according to its communication rhythm. After the fault disappears, device A stops appending protection short frames, while device B continues to receive regular service frames. This timing diagram corresponds to the processing order of complete transmission of regular service frames, post-insertion of protection short frames, and cessation of frame insertion after fault recovery.

[0025] The receiving end uses the device communication cycle as the reset timing unit. In each communication cycle, it checks whether a valid shortest Ethernet protection frame has been received. If a protection frame is received in the current cycle, the corresponding bridge arm is locked according to the faulty bridge arm location, and the count of missed cycles is not incremented. If no protection frame is received in the current cycle, a missed cycle is recorded. If no protection frame is received for two consecutive device communication cycles, the receiving end determines that the temporary overcurrent fault of the bridge arm has disappeared, cancels the fault locking status of the corresponding bridge arm, and releases the phase-specific temporary locking.

[0026] When the device communication cycle is configured to 50 microseconds, the receiver checks two consecutive 50-microsecond cycles; when the device communication cycle is configured to 100 microseconds, the receiver checks two consecutive 100-microsecond cycles. The cycle used in the recovery rule is consistent with the actual communication cycle of the regular service frame, and no independent timing parameter is introduced. During the fault duration, the transmitter repeatedly sends a protection short frame after each regular service frame, and stops sending after the fault is recovered. Therefore, the receiver can distinguish between fault persistence and fault disappearance based on the reception status of two consecutive cycles.

[0027] The short frame protection mechanism does not change the original communication disconnection criteria of the flexible DC control and protection system. The system continues to perform three-layer disconnection discrimination: the physical layer identifies the physical connection status based on the link hardware detection results; the data link layer identifies the message transmission status based on the frame integrity verification results; and the application layer identifies the service communication status based on the timing verification results of cross-device heartbeats. The three-layer discrimination results are combined with a two-out-of-three redundancy voting mechanism, and the communication disconnection judgment time is no more than 10 milliseconds.

[0028] The valve protection 3-out-of-2 device is compatible with parsing both regular service frames and the shortest Ethernet protection frames. The heartbeat and status information in the regular service frames continue to serve the original communication disconnection judgment, while the fault phase information and fault arm number information in the protection short frame serve the handling of arm power faults. The receiving end distinguishes between communication disconnection faults and arm power faults and sends corresponding alarm information accordingly, so that the start and stop of the protection short frame will not replace the disconnection judgment of the physical layer, data link layer, and application layer.

[0029] Step S101: The flexible DC control and protection system unifies the gigabit communication parameters of the polar control device, valve control device, and valve interface device, adopts the 1000BASE-X fiber gigabit communication mode and Ethernet II standard frame, configures the device communication cycle of 50 microseconds or 100 microseconds according to the cabinet group, and maintains the original sending and receiving logic of heartbeat message and status interaction message; thus forming a gigabit point-to-point Ethernet link shared by regular service frames and protection short frames.

[0030] In step S102, the FPGA of the valve protection device detects the overcurrent fault level of the bridge arm. When no temporary overcurrent fault occurs, the valve protection device only sends regular service frames according to the device communication cycle. When a temporary overcurrent fault occurs, the FPGA determines the fault phase information and the fault bridge arm number information, and generates the shortest Ethernet protection frame carrying the two point information.

[0031] In step S103, the sending end waits for the current regular service frame to be completely transmitted. After the regular service frame ends, it immediately opens the protection short frame interpolation channel and sends the shortest Ethernet protection frame. If the fault still exists, it continues to send protection short frames after each subsequent regular service frame. If the fault returns, it closes the interpolation channel and stops sending protection short frames.

[0032] In step S104, the valve protection three-out-of-two device receives the shortest protection frame from the Ethernet. The FPGA completes the CRC32 check and parses the fault phase information and fault bridge arm number information when the check is successful. The valve protection three-out-of-two device completes the logical voting based on the parsing result and sends the corresponding bridge arm partition interlocking command to the valve interface device. The valve interface device drives the corresponding bridge arm to perform the interlocking protection action.

[0033] In step S105, the receiving end checks the reception status of the protection short frame every communication cycle; when the protection short frame is received continuously, the corresponding bridge arm is locked; when the protection short frame is not received for two consecutive device communication cycles, the fault is determined to have disappeared and the phase-separated temporary lockout is released, while the regular service frame is still transmitted periodically.

[0034] In step S106, the system performs physical layer link hardware detection, data link layer frame integrity verification, and application layer heartbeat timing verification in parallel, and determines communication disconnection by combining a three-out-of-two redundancy voting mechanism; the system processes communication disconnection alarms and bridge arm power fault alarms respectively, without allowing the reset determination of protection short frames to replace the original communication disconnection criteria.

[0035] This embodiment distinguishes between regular service frames and protection short frames within the same gigabit point-to-point Ethernet link, retaining the original operating conditions, heartbeat, and disconnection judgment logic, and inserting the shortest protection frame at the end of the regular service frame via FPGA. This relationship reduces the need for additional external hardware interfaces for fast protection. The protection short frame only carries the fault phase and bridge arm number. After verification by the receiving end, it performs zone blocking according to the bridge arm location. This relationship ensures that the protection action object corresponds to the faulty bridge arm determined by the sending end. After fault recovery, the sending end stops inserting frames, and the receiving end releases the blocking when it does not receive a protection short frame for two consecutive communication cycles. This relationship allows protection triggering, maintenance, and recovery to use the same short frame state connection.

[0036] Corresponding to the above method, the fast communication system between devices in the flexible DC control and protection system includes a valve protection device, a valve protection 3-out-of-2 device, a valve interface device, and a gigabit point-to-point Ethernet link. The valve protection device is configured with a transmitting FPGA to execute steps S101 to S103; the valve protection 3-out-of-2 device is configured with a receiving FPGA and message receiving, CRC32 verification, bridge arm position parsing, and logical voting functions to execute the receiving and voting in step S104 and the fault return determination in step S105; the valve interface device is configured with a latching output corresponding to the submodule drive unit to execute the latching drive in step S104 and the latching release in step S105; each device continues to execute the three-layer communication disconnection judgment described in step S106.

[0037] exist Figure 1 In the architecture shown, three protection hosts provide protection information to the two-out-of-three (3-out-of-two) A and two-out-of-three (3-out-of-two) B devices, respectively. Two sets of valve protection 3-out-of-two devices are connected to the valve control host and the core boards in the odd-interface and even-interface chassis, respectively. This architecture is used to implement the communication relationship between the valve protection devices, the two-out-of-three valve protection devices, and the valve interface devices. The system's regular service frames, shortest Ethernet protection frames, device communication cycles, fault phase information, fault bridge arm number information, reset criteria, and disconnection criteria are all the same as in the aforementioned implementation method.

[0038] In the absence of logical conflicts, the above steps and system components can be implemented according to the recorded 50 microsecond or 100 microsecond communication cycle. The 1000BASE-X or 1000BASE-T communication mode, Ethernet II frame format, regular service frame payload, protection shortest frame, continuous fault retransmission, fault reset and stop transmission, release of blocking for two consecutive cycles, and the original three-layer disconnection discrimination relationship remain unchanged.

[0039] The transmitting device refers to the device that generates regular service frames during inter-device communication and generates the shortest Ethernet protection frame after detecting a temporary overcurrent fault in a bridge arm. In this embodiment, it corresponds to the valve protection device. The receiving device refers to the device that receives and verifies the shortest Ethernet protection frame, obtains fault phase information and fault bridge arm number information, and participates in the interlocking process. In this embodiment, it corresponds to the valve protection 3-out-of-2 device. The valve interface device is located at the downstream stage of the receiving-side protection link and is used to apply the interlocking command generated by the valve protection 3-out-of-2 device to the corresponding bridge arm. Therefore, sending, receiving, voting, and execution each have a clear subject. The role played by the same device in different communication directions can be determined according to the transmission direction of the frame, but it does not change the data relationship in this embodiment.

[0040] Regular service frames are standard Ethernet frames continuously transmitted according to the device communication cycle. Their service payloads are used to carry device operating conditions, cross-device heartbeats, and basic interactive data. The shortest Ethernet protection frame is the minimum valid media access control frame that is appended to the regular service frame during the existence of a temporary overcurrent fault. Its service payload is used to carry fault phase information and fault arm number information. The term "shortest" means that the frame meets the requirements of the minimum valid media access control frame for Gigabit Ethernet, rather than omitting the destination address, source address, type field, or cyclic redundancy check field. The protection frame still has complete Ethernet frame fields and is subject to integrity checks corresponding to its frame format.

[0041] The device operation cycle is the cycle corresponding to the flexible DC control and protection device performing one existing operation. The device communication cycle is the cycle on which the transmitting end sends adjacent regular service frames. The two are matched during configuration. In this embodiment, no separate protection short frame transmission cycle is set. The protection short frame uses the end event of the regular service frame as the transmission trigger position. When the fault continues, each occurrence of the end event of the regular service frame provides an opportunity to insert a protection short frame. Therefore, the repeated transmission rhythm of the protection short frame naturally follows the existing device communication rhythm and will not change the original operating conditions and heartbeat data transmission order due to the separate setting of an independent rhythm.

[0042] The overcurrent fault level of the bridge arm is a fault status quantity provided by the existing overcurrent protection logic of the bridge arm inside the valve protection device to the field programmable gate array. When the fault level is in the active state, it indicates that a temporary overcurrent fault exists in the corresponding bridge arm. When the fault level returns to the reset state, it indicates that the temporary overcurrent fault of the corresponding bridge arm has disappeared. In this embodiment, the fault level is used to trigger the generation and termination of protection short frames. No separate bridge arm current threshold calculation model is established, nor is the judgment criterion of the bridge arm overcurrent protection itself changed. Therefore, the input boundary of the protection short frame communication processing begins with the fault level and its corresponding point given by the existing protection logic.

[0043] The fault phase information indicates the phase in which a temporary overcurrent fault occurred, and the fault arm number information indicates the arm in which the temporary overcurrent fault occurred under that phase. Both pieces of information are obtained from the protection point to which the fault level belongs. The corresponding arm refers to the arm that the fault phase information and the fault arm number information both point to. The receiving end only sends the parsing result to the logic voting and blocking command formation process after the protection short frame is valid and both pieces of information have been parsed. Since the phase and the arm number together limit the action object, the valve interface device can apply the blocking to the arm indicated by the protected short frame.

[0044] At the start of communication configuration, the configuration personnel obtain the communication port parameters corresponding to the target point-to-point connection from the extreme control device, valve control device, valve protection device, valve protection three-out-of-two device, and valve interface device. The communication port parameters are at least used to indicate the gigabit fiber optic communication mode, Ethernet standard frame format, and device communication cycle. In accordance with the principle that the sending and receiving ends of the same point-to-point link use the same communication parameters, the ports at both ends are set to the recorded gigabit communication mode, and the regular service frame format is set to Ethernet II format. Since the two ends of the link use the same physical communication mode and media access control frame format, the receiving end can identify the line data according to the same field boundaries as the sending end.

[0045] When determining the device communication cycle, first obtain the device operation cycle currently used by the target cabinet group, then select the cycle that matches the device operation cycle from the recorded communication cycle configuration, and write the cycle into the regular service frame transmission configuration and reception check configuration at both ends of the same communication link respectively; the transmitting end forms a regular service frame in each device communication cycle accordingly, and the receiving end divides adjacent protection short frame reception check intervals accordingly. Since the transmission interval of the transmitting end and the check interval of the receiving end have the same cycle basis, the fault continuous retransmission and fault reset count can be connected around the same device communication beat.

[0046] After configuration, the link parameters are checked through the periodic sending and receiving of regular service frames. The sending end writes the existing device status, cross-device heartbeat, and basic interaction data into the regular service frame according to the order of the service fields. The receiving end reads the data according to the same field order and continues to execute the original heartbeat and status interaction logic. If the receiving end cannot obtain a complete regular service frame, it locates the communication status according to the original physical layer link hardware detection, data link layer frame integrity verification, and application layer heartbeat timing verification, instead of replacing the link configuration check with the absence of a protection short frame.

[0047] In this embodiment, point-to-point refers to a direct communication relationship between the transmitting and receiving interfaces of a gigabit Ethernet link, protecting short frames from transmission along the link already used by regular service frames. Protecting short frames does not require adding independent cables, independent fiber optic channels, or independent digital input modules. The transmitting end field-programmable gate array adds a protective short frame data channel within the existing transmitting link, and the receiving end field-programmable gate array adds a protective short frame identification path within the existing receiving link. Since the new processing is located within the device's native field-programmable gate array and the existing link, no new external hardware circuitry is required.

[0048] When forming a regular service frame, the sending end first obtains the device operating status, cross-device heartbeat, and basic interaction data that need to be exchanged in this communication cycle from the device internally, and maintains the data sequence relationship specified by the original service logic. Then, the sending end writes the media access control address of the destination device into the destination media access control address field, writes the media access control address of its own device into the source media access control address field, writes the type information used to indicate the type of regular service frame into the type field, and then writes the basic interaction data into the service payload in sequence. When the service payload is less than the minimum Ethernet payload length, it maintains the effective frame length according to the Ethernet standard frame requirements. When the service payload reaches the recorded amount of regular service data, the frame body is formed according to that amount of data.

[0049] The basic interaction data of the recorded regular service frame consists of 64 groups of 32 bits each. When forming the service payload, the order of the basic interaction data in the existing service configuration is used as the writing order. First, the 32 bits of the current group are obtained, and then appended to the previous group of data that has been written. This process of obtaining and appending is repeated until all 64 groups of data are included in the service payload. Here, "composition" refers to the sequential splicing relationship. No summation, averaging, or compression operations are performed on each group of data. The receiving end retrieves each group of data in sequence according to the same group order and the same 32-bit boundary of the group, thus maintaining the semantics of the original basic interaction data.

[0050] After the Media Access Control (MAC) data frame body is formed, the sending end performs CRC32 operations on each field of data participating in Cyclic Redundancy Check (CRC) according to the Ethernet frame check order, and writes the resulting verification result into the CRC field. The physical layer then sends a seven-byte preamble and a one-byte start-of-frame (FOF) character before the MAC data frame body. The preamble is used for receiving line establishment bit synchronization, and the FFO character is used to indicate the start of the MAC data frame body. The destination address, source address, type field, service payload, and CRC field are then sent continuously on the line according to the already formed field order.

[0051] The end of regular service frame transmission means that the last line field of the frame has been transmitted, not just that the service payload has been written. The transmitting end uses the complete frame transmission end signal as a prerequisite for inserting protection short frames. When the protection short frame transmission request arrives during the transmission of regular service frames, only the request and the corresponding fault point are maintained, and the regular service frame being transmitted is not truncated. Only after the complete frame transmission end signal appears will the transmission of protection short frames begin. Since the transmission object is switched at the boundary of the complete frame, it is possible to avoid inserting protection data into the regular service frame and thus destroying the integrity of both types of frames.

[0052] The process for determining frame length and line transmission duration is as follows: To determine the transmission duration of a regular service frame, first obtain the number of bytes for the physical layer synchronization prefix, destination media access control address, source media access control address, type field, regular service payload, and cyclic redundancy check field, respectively. Then, add the byte counts of each part in sequence to obtain the total number of bytes of the regular service frame on the line. Next, multiply the total number of bytes by the number of bits per byte to obtain the total number of bits of the regular service frame. Finally, divide the total number of bits of the regular service frame by the number of bits that a gigabit point-to-point Ethernet link can transmit per second. The resulting time is the complete transmission duration of the regular service frame from the preamble to the end of the cyclic redundancy check field.

[0053] The conventional service frame described in this application contains 64 groups of basic interactive data, each group consisting of 32 bits. When calculating the number of service payload bytes, the number of groups is first multiplied by the number of bits in each group to obtain the total number of bits of basic interactive data. Then, the total number of bits of basic interactive data is divided by the number of bits per byte to obtain the number of service payload bytes. This number of service payload bytes is added to the number of bytes of the address, type field, cyclic redundancy check field, and physical layer synchronization prefix. The complete transmission duration is then determined according to the ratio of the total number of bits to the link transmission rate. According to the described frame structure, the complete line transmission duration of the conventional service frame is 2.256 microseconds.

[0054] When determining the line transmission duration of the shortest Ethernet protection frame, the number of bytes in the minimum effective Media Access Control (MAC) frame of Gigabit Ethernet is used as the body length of the MAC data frame. This is then increased by adding a seven-byte preamble and a one-byte start-of-frame character to obtain the total number of bytes in the protection frame, including the physical layer synchronization prefix. This total number of bytes is then converted to the total number of bits in the protection frame. Finally, the total number of bits in the protection frame is divided by the number of bits that a Gigabit point-to-point Ethernet link can transmit per second to obtain the complete line transmission duration of the protection frame. Based on the recorded shortest frame length, this duration is 0.576 microseconds.

[0055] When determining the combined transmission duration of the subsequent interpolated frame, the waiting boundary is set at the time when the protection short frame request appears after the current regular service frame has started to be sent. The sending end first completes the current regular service frame and then sends an Ethernet minimum protection frame. The complete line transmission duration of the aforementioned regular service frame is directly added to the complete line transmission duration of the protection frame. The result is the combined line occupancy time of the regular service frame and the subsequent protection short frame. According to the sum of the two recorded durations, the maximum combined transmission duration is 2.832 microseconds. This determination process only reflects the continuous transmission relationship of the frame on the line and does not include the device operation time, fault judgment time or subsequent blocking execution time in the frame transmission duration.

[0056] When it is necessary to verify the calculation results, the check can be performed in the reverse direction. First, multiply the obtained transmission duration by the number of bits transmitted per second of the gigabit link to get the number of bits that should be transmitted on the line. Then, divide the number of bits by the number of bits contained in each byte to get the total number of bytes that should be there. Compare this number with the total number of bytes obtained by adding the number of bytes in each field. If the two are consistent, it indicates that the field count, bit conversion, and rate conversion correspond to each other. If the two are inconsistent, it is necessary to go back and check whether the physical layer synchronization prefix is ​​missing or whether the media access control frame body and the service payload are mixed up as the same length.

[0057] The specific process of fault location acquisition and protection short frame generation is as follows: The field-programmable gate array (FPGA) of the valve protection device uses the bridge arm overcurrent fault level as the fault state input. Each bridge arm point that can trigger temporary overcurrent protection has an input relationship corresponding to its phase and bridge arm identifier. When the FPGA reads the fault level, it also retains the source point of the fault level. When the fault level is in the action state, it obtains the phase from the source point and forms the fault phase information, and obtains the bridge arm identifier from the same source point and forms the fault bridge arm number information. Therefore, the two pieces of information come from the same fault point and are not obtained by recalculating the threshold by the current sampling value or reordering multiple bridge arms.

[0058] In this communication method, fault identification specifically means distinguishing the fault level output by the existing bridge arm overcurrent protection logic into an active state and a reset state. The field-programmable gate array generates a protection short frame transmission request in the active state and cancels the protection short frame transmission request in the reset state. The generation of the active state and the reset state is still the responsibility of the original bridge arm overcurrent protection function of the flexible DC control and protection system. This implementation does not add protection settings, change overcurrent criteria, or perform smoothing, fitting, or prediction processing on the fault level. Therefore, the communication method only processes the already identified fault state and fault location.

[0059] When generating a protection short frame, the sending end first writes the destination MT control address and source MT control address that are consistent with those of the receiving end. Then, it writes type information in the type field to distinguish the protection short frame from regular service frames. Subsequently, according to the protection payload field order used by both the sending and receiving ends, the fault phase information is written to the phase position, the fault arm number information is written to the arm number position, and the protection frame is made to reach the minimum effective MT control frame length of Ethernet. Finally, CRC32 operation is performed on the frame data participating in the cyclic redundancy check and the check result is written to obtain a complete protection short frame that can be sent along a gigabit point-to-point Ethernet link.

[0060] The field position correspondence is directly determined by the protection short frame load layout in the device configuration. The transmitting end writes the information according to the relationship that "phase position corresponds to fault phase information and bridge arm number position corresponds to fault bridge arm number information". The receiving end reads the information according to the same position relationship. This correspondence does not require the use of additional curves, lookup tables or fitting functions. If described by a data channel, the phase output terminal of the protection short frame data channel is connected to the phase position in the load, and the bridge arm number output terminal is connected to the bridge arm number position in the load. The receiving side recovers the two pieces of information from the two positions in the opposite direction.

[0061] The remaining load positions required for the protection short frame to reach the minimum effective frame length are used to maintain the effective length of the standard frame, and undefined positions are not resolved as new protection action information; the receiving end only forms the corresponding bridge arm position based on the recorded fault phase information and fault bridge arm number information, and any data not defined as fault phase position or bridge arm number position in this embodiment is not used to increase the fault category, protection setting value or action range.

[0062] The post-frame insertion scheduling and continuous transmission process are as follows: The transmitting field-programmable gate array (FPGA) maintains a transmission state for both the regular service frame transmission channel and the protection short frame data channel. The regular service frame transmission channel submits a transmission request according to the device communication cycle, while the protection short frame data channel submits a transmission request when the bridge arm overcurrent fault level is in the active state. When only a regular service frame transmission request is received, the FPGA transmits the regular service frame according to the original cycle. When a protection short frame transmission request occurs during the transmission of a regular service frame, the FPGA latches the request, its fault phase information, and the fault bridge arm number information, and waits for the current regular service frame to be completely transmitted.

[0063] After the current regular service frame is fully transmitted, the field-programmable gate array (FPGA) checks whether the protection short frame transmission request still corresponds to the acquired temporary overcurrent fault. If the fault is still active, the protection short frame data channel is selected, and the already formed shortest Ethernet protection frame is transmitted to the same gigabit point-to-point Ethernet link. After the protection short frame transmission is completed, the FPGA returns to the regular service frame cycle transmission process. Since the channel selection occurs at the boundary of the complete frame and the protection short frame is located after the current regular service frame, this scheduling is a post-interpolation frame and does not change the order of fields within the current regular service frame.

[0064] The continuous fault transmission process is maintained on the condition that the overcurrent fault level of the bridge arm remains in the active state. Whenever a subsequent regular service frame ends, the field-programmable gate array (FPGA) obtains the current protection short frame transmission request again. If the request is valid, a protection short frame is appended after the regular service frame. This process does not take the successful transmission of the first protection short frame as the termination condition, nor does it require waiting for the receiver to return new confirmation information. During the fault state, the protection short frame recurs with the regular service frame cycle, so that the receiver can continuously obtain the bridge arm fault location and maintain the blocking state.

[0065] When a fault is reset, the existing bridge arm overcurrent protection logic restores the fault level to the reset state, and the field-programmable gate array cancels the protection short frame transmission request accordingly. If no protection short frame is currently being transmitted, the subsequent regular service frame will not be transferred to the protection short frame channel when it ends. If the current protection short frame has started being transmitted, the integrity of the frame will be maintained and no subsequent protection short frames will be generated after it ends. Regular service frames continue to be transmitted according to the device communication cycle. Therefore, stopping frame insertion will not interrupt regular service communication, nor will it generate incomplete protection frames that only send part of the fields.

[0066] The so-called priority of protection short frames over regular service frames means that when a regular service frame has been completely transmitted and the request to send a protection short frame is valid, the protection short frame is sent first, and then the normal transmission process of the next regular service frame begins. This priority relationship does not mean that the protection short frame is allowed to preempt the regular service frame being transmitted, nor does it mean that the regular service frame is discarded. The sending end still uses complete frames as the scheduling unit. Therefore, fast protection data and existing service data can be transmitted sequentially in the same link according to clear boundaries.

[0067] The specific calculations and validity checks for cyclic redundancy check are as follows: When the sending end performs CRC32 calculation, it first obtains the destination MAC address, source MAC address, type field, and service payload according to the order of the fields participating in cyclic redundancy check in the Ethernet II frame, and connects them to obtain the bit sequence to be checked. Then, it updates the check register bit by bit according to the CRC32 initial state, input bit order, and generator polynomial relationship specified by the Ethernet standard. For each input bit, it first XORs the input bit with the end of the check register used for feedback to obtain the feedback state, and then moves the check register one bit position in the specified direction. If the feedback state is valid, it performs bit-by-bit XOR with the coefficient bits corresponding to the standard generator polynomial. If the feedback state is invalid, it retains the shift result.

[0068] After all the bits to be checked are input, the sending end obtains the contents of the check register according to the output processing order specified by the Ethernet standard, forms the CRC32 check result of the sending end, and writes the result into the cyclic redundancy check field of the protection short frame; the preamble and start-of-frame character are used for physical layer synchronization and are not confused with the fields of the Media Access Control data frame body. The sending end and the receiving end keep the range of fields participating in the check consistent. Therefore, the receiving end can repeat this operation for the same frame data received.

[0069] After receiving the protection short frame, the receiving end first determines the start position of the media access control data frame body according to the frame start symbol, and then obtains the frame data to be checked and the verification result of the sending end based on the frame structure of destination address, source address, type field, service payload and cyclic redundancy check field. The receiving end adopts the same initial state, input order, register movement direction, generator polynomial relationship and output processing order as the sending end for the frame data to be checked, and performs feedback XOR and register update bit by bit to obtain the CRC32 verification result of the receiving end.

[0070] During the validity comparison, the receiving end compares the CRC32 check result of the receiving end with the CRC32 check result of the sending end carried in the cyclic redundancy check field of the protection short frame bit by bit. When all corresponding bits are consistent, a valid check result is formed, allowing subsequent type identification and fault point parsing. When any corresponding bit is inconsistent, an invalid check result is formed. This frame is not used as the basis for bridge arm blocking, nor is the load in it interpreted as a new fault phase or bridge arm number. Since the validity judgment is completed before the protection point enters the blocking logic, changes in frame data that occur during line transmission will not directly trigger protection actions.

[0071] When the length of the received frame is insufficient to contain the complete address, type field, service payload, and cyclic redundancy check field, or when the frame end position does not conform to the Ethernet standard frame structure, the receiving end cannot obtain the complete bit sequence to be checked and the verification result of the sending end, and therefore directly treats the frame as an invalid frame. When the frame structure is complete but the CRC32 comparison is inconsistent, it is also treated as an invalid frame. In both cases, the protection short frame bit parsing is not entered, but the data link layer can continue to provide the frame integrity verification result to the original communication disconnection judgment process.

[0072] The specific steps for protecting short frame recognition, field parsing, and point recovery are as follows: After completing the frame validity verification, the receiving end reads the type field. When the type field corresponds to the protection short frame type information, the frame is sent to the protection short frame parsing path. When the type field corresponds to the regular service frame type information, the frame is sent to the original regular service parsing path. The identification here is to compare the received type field with the frame category identifier configured consistently at both ends of the device. If they match, the corresponding parsing path is selected. If they do not match, the frame is not mistakenly identified as a protection short frame. Therefore, protection short frames can bypass the regular service parsing thread without changing the processing of regular operating conditions and heartbeat messages.

[0073] After entering the protection short frame parsing path, the receiving end uses the starting position of the protection load as the positioning reference and reads the data of the fault phase position according to the field position order used by the sending end when writing, to obtain the fault phase information. Then, it reads the data of the bridge arm number position to obtain the fault bridge arm number information. During the reading process, the original bit arrangement of the fields is maintained, and the fault phase information and fault bridge arm number information are not summed, averaged, interpolated or compressed to recover. The two pieces of information are used to determine the phase and the bridge arm identifier under that phase, respectively.

[0074] When restoring the corresponding bridge arm position, the receiving end first selects the phase based on the parsed fault phase information, and then selects the bridge arm within the phase range based on the parsed fault bridge arm number information. The two together determine the protection action target. If only the phase is obtained without the complete bridge arm number, or only the bridge arm number is obtained without the complete phase, the recorded fault bridge arm position cannot be formed, and the corresponding bridge arm blocking basis is not formed for the subsequent stage. Therefore, the receiving end will not expand the blocking range with incomplete positions.

[0075] The valve protection 3-out-of-2 device stores the arrival status of the legitimate protection short frame, the fault phase information, and the fault bridge arm number information in the protection input status of its own communication cycle. The arrival status of the protection short frame indicates that there is legitimate protection information in this cycle, and the two point information items indicate the bridge arm to which the protection information points. When the next device communication cycle arrives, the receiving end updates the protection input status again according to the legitimate protection short frames actually received in this cycle, without unconditionally extending the arrival status of earlier cycles to the current cycle, thereby providing a clear cycle boundary for counting subsequent consecutive unreceived cycles.

[0076] The redundancy voting, interlock command generation, and valve interface execution are detailed as follows: Figure 1 The protection architecture shown includes protection host A, protection host B, and protection host C. The three protection hosts provide protection information to the valve protection three-out-of-two device. The valve protection three-out-of-two device compares the legitimate protection information from each protection host within the same protection processing period. The comparison includes fault status, fault phase information, and fault bridge arm number information. When the fault status given by at least two protection hosts is valid and indicates the same fault phase and the same fault bridge arm, the voting result points to the corresponding bridge arm. When less than two protection hosts form a consistent indication, the voting result for the same bridge arm is not expanded based on a single inconsistent input.

[0077] The specific determination process of the 3-out-of-2 logic is as follows: First, obtain the legal status and position information of the three protection input channels respectively. Then, compare the position information of the first protection input channel with that of the second protection input channel, compare the position information of the first protection input channel with that of the third protection input channel, and compare the position information of the second protection input channel with that of the third protection input channel. When both channels in any pairwise comparison are legal fault inputs and the fault phase and bridge arm number are consistent, a valid voting result for that bridge arm is formed. If all pairwise comparisons are not satisfied, the result for that bridge arm confirmed by the 3-out-of-2 logic is not formed.

[0078] After a valid voting result is obtained, the valve protection three-out-of-two device organizes the fault phase and fault arm numbers obtained from the voting into corresponding arm partition interlocking commands, and then... Figure 1 The communication relationship between the valve interface device and the control is transmitted; the action object of the lockout command directly inherits the fault phase and bridge arm number in the voting result, without reselecting the bridge arm based on other curves, coefficients or scores. Since the voting input and lockout output establish a one-to-one correspondence around the same bridge arm point, the downstream valve interface device can identify the zone that needs to be locked.

[0079] After receiving the zone interlock command, the valve interface device parses the fault phase and bridge arm number carried in the command, determines the phase to which the bridge arm belongs by the fault phase, and determines the specific bridge arm by the bridge arm number. Then, it controls the sub-module drive unit corresponding to the bridge arm to execute the interlock. The valve interface device does not convert a single bridge arm point into other fault categories that are not recorded, nor does it change the voting result of the valve protection three-out-of-two device. Since the interlock output is transmitted directionally along the phase and bridge arm number, the phase-specific temporary interlock can be implemented to the corresponding bridge arm.

[0080] During the fault period, the valve protection three-out-of-two device repeatedly receives legitimate protection short frames and continuously forms the corresponding bridge arm protection input in the continuous device communication cycle, and the valve interface device maintains the bridge arm lockout accordingly. After the fault is recovered, the transmitting end stops inserting protection short frames, the receiving end forms a recovery state according to the continuous non-received cycle criterion, the valve protection three-out-of-two device cancels the lockout command of the corresponding bridge arm, and the valve interface device cancels the fault lockout of the corresponding bridge arm according to the command state. Thus, the action and recovery use the same fault bridge arm point connection.

[0081] The continuous non-received period count and fault reset process are as follows: The receiving end sets a consecutive non-received period count for the protection short frame reception status. The initial state indicates that no non-received periods have been accumulated. At the end of each device communication cycle, the receiving end checks whether a valid message identified as a protection short frame has been received within that cycle. If received, the consecutive non-received period count is restored to the initial state, and the fault point obtained from parsing the protection short frame is maintained as the blocking. If not received, one cycle is added to the consecutive non-received period count of the previous cycle, and the result is used as the updated consecutive non-received period count for the current cycle.

[0082] The fault reset determination cycle number is taken from two consecutive device communication cycles recorded. When comparing, the number of consecutive non-received cycles updated in the current cycle is compared with the fault reset determination cycle number. If the number of consecutive non-received cycles has not reached two cycles, the receiving end does not cancel the bridge arm blocking based on a single non-received cycle. When the number of consecutive non-received cycles reaches two cycles, the receiving end determines that the transmitting end has stopped sending protection short frames, the temporary overcurrent fault of the corresponding bridge arm has disappeared, and then a reset result for canceling the fault blocking of the bridge arm is formed.

[0083] If a valid protection short frame is received again in the next cycle after the first unreceived cycle, the receiver restores the number of consecutive unreceived cycles to the initial state and does not add the previous unreceived cycle to the subsequent non-consecutive unreceived cycles. The count only increases continuously and reaches the fault reset determination cycle number when no valid protection short frame is received in the communication cycles of adjacent devices. Therefore, "consecutive" means that the unreceived state is not interrupted by a valid protection short frame between adjacent cycles.

[0084] When the device communication cycle adopts the recorded 50 microsecond configuration, each count increment represents a 50 microsecond device communication cycle. When the device communication cycle adopts the recorded 100 microsecond configuration, each count increment represents a 100 microsecond device communication cycle. The receiving end does not mix the two cycle configurations for counting, but reads the actual device communication cycle used in this link as the cycle boundary. Therefore, the regression count is always consistent with the actual beat of the sending end's regular service frames and fault continuous retransmission protection short frames.

[0085] Invalid protection frames are not treated as valid protection frames and are not entered into the point parsing process, thus they cannot be used to maintain the corresponding bridge arm protection input. At the same time, the receiver still provides the integrity verification result of the frame to the data link layer for disconnection judgment. The continuous non-received periodic regression criterion is used for bridge arm temporary overcurrent fault status, and the communication disconnection criterion is used for link fault status. The two have different inputs and outputs. The receiver classifies and forms bridge arm power fault results and communication disconnection alarms to avoid merging the two types of fault processing into a single undifferentiated result.

[0086] The physical layer disconnection judgment uses the link hardware detection result provided by the gigabit point-to-point Ethernet port as input. The link hardware detection result indicates whether the physical connection and physical layer communication status are maintained between the sending port and the receiving port. When the physical layer detects that the link hardware status does not meet the normal communication conditions, a physical layer abnormal result is formed. When the link hardware status meets the normal communication conditions, a physical layer normal result is formed. This judgment adopts the original physical layer detection function of the flexible DC control and protection system and is not replaced by the presence or absence of protection short frames.

[0087] The data link layer disconnection detection uses the integrity check result of the received Ethernet frame as input. The receiving end checks the regular service frames and protection short frames according to the aforementioned frame field integrity and CRC32 comparison process. When the received frame structure is complete and the cyclic redundancy check result is consistent, a corresponding valid frame result is formed. When the frame structure is incomplete or the cyclic redundancy check result is inconsistent, a frame abnormal result is formed. The data link layer forms a normal or abnormal result based on the original frame integrity check status to reflect the situation where the physical connection exists but the frame data cannot be legally received.

[0088] Application layer disconnection detection uses the cross-device heartbeat and its reception timing carried in the regular service frame as input. The sending end continuously updates and sends heartbeats according to the device communication cycle. The receiving end checks whether the heartbeats arrive continuously according to the service communication relationship according to the original application layer heartbeat timing verification logic. When the heartbeat reception meets the original timing requirements, a normal application layer result is formed. When the heartbeat reception does not meet the original timing requirements, an abnormal application layer result is formed. The protection short frame is only used for rapid communication during bridge arm faults. Its repeated transmission cannot replace the cross-device heartbeat in the regular service frame.

[0089] When a communication disconnection result is generated, the system obtains the normal or abnormal results from the physical layer, data link layer, and application layer respectively, and makes a judgment according to the recorded two-out-of-three redundant voting relationship. Specifically, the results of the three layers are combined in pairs. As long as at least two layers generate abnormal results, a communication disconnection judgment result is generated. If less than two layers generate abnormal results, a two-out-of-three confirmed communication disconnection result is not generated based on a single layer abnormality. The system continues to retain the detection result of that layer for the original state processing.

[0090] The communication disconnection determination process follows the recorded determination time of no more than 10 milliseconds. The determination time is formed by the original physical layer hardware detection, data link layer frame integrity verification, application layer heartbeat timing verification, and two-out-of-three voting. This implementation does not add a new disconnection threshold. When a communication disconnection result is formed, the system sends a communication disconnection alarm. When the protection short frame is valid and the fault point is voted to form a bridge arm blockade, the system processes the bridge arm power fault and sends the corresponding fault information. Since the formation conditions of the two types of results are different, alarms can be classified.

[0091] After the system is put into communication, protection host A, protection host B, and protection host C form regular service frames according to the device's operation cycle, and send the device status, cross-device heartbeat, and basic interaction data to the two-out-of-three (3-out-of-three) A and 3-out-of-three (3-out-of-three) B via the existing gigabit point-to-point Ethernet link. The two sets of valve protection 3-out-of-three devices verify and parse the regular service frames, and continue to execute the existing status interaction and cross-device heartbeat timing verification. The core boards in the valve control host, bridge arm control, and odd-interface and even-interface chassis follow the established procedures. Figure 1 The communication relationships shown maintain the original business processing.

[0092] When a temporary overcurrent fault occurs in any bridge arm, the overcurrent fault level of the corresponding bridge arm whose protection logic output is already active is obtained from the fault location by the field-programmable gate array (FPGA) at the transmitting end. The fault phase information and fault bridge arm number information are obtained from the fault location, and a protection short frame transmission request is formed. If a regular service frame is being transmitted at this time, the FPGA saves the transmission request and the two location information, waits for the cyclic redundancy check field of the regular service frame to be transmitted, and then forms a complete Ethernet shortest protection frame and sends it from the same communication port.

[0093] After receiving the protection short frame, the 3-out-of-2 A and 3-out-of-2 B devices each obtain the Media Access Control (MAC) data frame body based on the frame start position. They perform the same CRC32 operation on the destination address, source address, type field, and service payload as the sending end, and compare the obtained receiving end verification result with the frame intra-frame verification result bit by bit. After the comparison is consistent, the two valve protection 3-out-of-2 devices select the protection short frame parsing path according to the type field, obtain the fault phase information from the fixed phase position, obtain the fault bridge arm number information from the fixed bridge arm number position, and form the legitimate protection input for the communication cycle of the device.

[0094] The valve protection three-out-of-two device compares the legitimate protection inputs from protection host A, protection host B, and protection host C pairwise according to the fault phase and bridge arm number. A valid voting result is formed when at least two protection hosts indicate the same bridge arm. The valid voting result is organized into a zone interlocking command for the corresponding bridge arm and transmitted to the subsequent stage along the communication relationship between the valve control host and the core board of the interface chassis. The valve interface device reads the fault phase and bridge arm number in the interlocking command and controls the corresponding sub-module drive unit to execute the bridge arm interlocking.

[0095] During the period when the bridge arm overcurrent fault level remains active, each protection host continues to send protection short frames after each subsequent regular service frame. The valve protection three-out-of-two device re-verifies, parses, and obtains the point information every cycle. When a valid protection short frame arrives, the number of consecutive unreceived cycles is restored to the initial state and the blocking is maintained. After the bridge arm overcurrent fault level returns to normal, the protection host stops forming protection short frame sending requests and only sends regular service frames. When the valve protection three-out-of-two device fails to receive a valid protection short frame for two consecutive device communication cycles, a fault return result is formed, and the valve interface device cancels the fault blocking of the corresponding bridge arm accordingly.

[0096] Throughout the entire process of the above actions and recovery, the physical layer continues to monitor the link hardware status, the data link layer continues to record the integrity verification status of regular service frames and protection short frames, and the application layer continues to check the cross-device heartbeats in regular service frames. The results from the three layers are used to determine the communication disconnection based on the original two-out-of-three relationship. Therefore, even if the transmitting end normally stops sending protection short frames due to bridge arm failure recovery, regular service frames and cross-device heartbeats are still used to prove the link communication status. The receiving end can distinguish the bridge arm recovery formed by the normal cessation of sending protection short frames from the communication link disconnection.

[0097] The field-programmable gate array (FPGA) at the transmitting end of the valve protection device is responsible for fault level reading, fault location acquisition, protection short frame generation, full frame boundary waiting, post-frame insertion, continuous fault retransmission, and fault reset and stop transmission. The field-programmable gate array at the receiving end of the valve protection three-out-of-two device is responsible for frame structure identification, CRC32 verification, protection short frame type identification, fault phase and bridge arm number parsing, periodic reception status update, and continuous non-received period counting. The protection logic of the valve protection three-out-of-two device is responsible for three-out-of-two voting and blocking command formation. The valve interface device is responsible for corresponding bridge arm blocking and blocking cancellation.

[0098] Since the transmitting end sends the shortest protection frame along the same gigabit point-to-point Ethernet link after the existing regular service frame, without additional external hardware circuitry, the bridge arm fault location does not need to go through additional contacts, cables, and additional digital interfaces to reach the receiving end, which can reduce the transmission layers added by the conversion of protection information between multiple external interfaces. Since the transmitting end waits for the current regular service frame to finish completely before inserting the protection short frame, the integrity of the regular service frame is maintained, and the protection short frame can occupy a clear line position before the next regular service frame.

[0099] Since the protection short frame payload only transmits fault phase information and fault arm number information at fixed positions, the receiving end recovers the two pieces of information according to the same positional relationship after the CRC32 check is valid. The three-out-of-two device performs a consensus vote around the same arm point. Therefore, the fault discrimination subject, the logic voting subject, and the blocking execution subject use a consistent identifier for the action object, which can direct the blocking command to the corresponding arm. At the same time, the cyclic redundancy check excludes invalid frames before the point enters the blocking logic, reducing the possibility that frame transmission changes are directly interpreted as arm points.

[0100] Because the protection short frame is repeatedly sent after each regular service frame during the fault duration, the periodic arrival status of the legitimate protection short frame can maintain the corresponding bridge arm blocking; because the transmitting end stops frame insertion after the fault is restored, the receiving end only accumulates the unreceived status of the adjacent device communication cycle, and cancels the blocking when no legitimate protection short frame is received for two consecutive cycles, therefore, rapid triggering, continuous maintenance and fault restoration are continuously connected by the same protection short frame status, and it is not necessary to establish a set of restoration messages that are disconnected from the fault location outside the protection short frame.

[0101] Since regular service frames still carry device operating conditions, basic interactive data, and cross-device heartbeats, the original physical layer link hardware detection, data link layer frame integrity verification, and application layer heartbeat timing verification still form the communication disconnection judgment according to the three-out-of-two relationship. Therefore, the existence of protection short frames is used to indicate that the temporary overcurrent fault of the bridge arm continues, and the normal stopping of protection short frames is used to form the bridge arm fault reset. Regular services and three-layer detection are used to judge the communication link status. The system can send bridge arm power fault information and communication disconnection alarms respectively.

[0102] In the absence of logical conflicts, the descriptions in this embodiment regarding gigabit point-to-point Ethernet links, Ethernet II regular service frames, Ethernet shortest protection frames, bridge arm overcurrent fault levels, fault phase information, fault bridge arm number information, post-interpolation frames, continuous fault retransmission, CRC32 checksum, two-out-of-three voting, two consecutive cycle reset, and three-layer disconnection judgment can be combined with each other. The combined processing is still based on the data relationship between the transmitting end, receiving end, valve protection two-out-of-three device, and valve interface device as recorded.

[0103] When the request to send a protection short frame and the signal to start sending a regular service frame occur simultaneously, the sending end first checks whether the regular service frame has already entered the line transmission process. If the regular service frame has started, the sending end retains the request to send the protection short frame and waits for the regular service frame to finish completely. If the regular service frame has not started and the sending end is at the frame boundary after the previous regular service frame has ended, the protection short frame is sent according to the relationship that the protection short frame takes precedence over the next regular service frame. Both processes are constrained by not truncating the complete regular service frame and ensure that the protection short frame is located after a complete frame boundary.

[0104] The protection short frame transmission request remains associated with the original fault location while waiting for the current regular service frame to end. The transmitting end does not recalculate the fault phase or bridge arm number based on the waiting time. When the regular service frame ends, the transmitting end reads the fault level status again. If the fault is still in the active state, the corresponding protection short frame is sent. If the fault has been restored, subsequent duplicate frames are canceled. If the protection short frame has already entered the line for transmission, the frame is kept intact until the end. Thus, changes in the fault status will not cause half-frames with missing check fields or missing payload fields to appear on the line.

[0105] When the receiving end obtains a protection short frame within a device communication cycle, the complete reception of the frame and the completion of the CRC32 comparison are taken as the establishment position of receiving a legitimate protection short frame. Only when the frame start position is received but the complete frame body is not received, the cycle is not pre-recorded as the arrival of a legitimate protection short frame. After the integrity check is completed and the result is consistent, the receiving end sets the protection short frame arrival status of the cycle to valid, so that the reception status used for the reset count is consistent with the range of legitimate frames that can enter the point resolution and blocking logic.

[0106] When the receiving end receives an Ethernet frame with a complete frame structure but whose type field does not indicate a protection short frame, it will either send it to the corresponding regular service processing according to its type or not process it as a protection short frame. The fault phase and bridge arm number will not be read from the fixed position of the service payload of the frame. Therefore, even if a regular service frame arrives at the time and position expected to appear of the protection short frame, it cannot replace the protection short frame to maintain the bridge arm fault state. The reception status of the protection short frame is determined by the frame legality and the frame category.

[0107] When the receiving end receives a frame whose type field indicates a protected short frame but whose CRC32 comparison is inconsistent, it does not read its payload as a fault point, and uses the existence of other legitimate protected short frames in that period as the basis for determining the period's reception status. If there are no other legitimate protected short frames in that period, the number of consecutive unreceived periods is updated according to the unreceived rules. At the same time, the CRC32 inconsistency result is entered into the data link layer frame integrity status for use in the original three-layer communication disconnection judgment. Therefore, the same abnormal frame can affect the legitimate reception status of protected short frames and the data link layer status respectively, but it will not be directly converted into a blocking point.

[0108] After the receiving end parses the fault phase information and the fault arm number information, it first maintains the common source relationship of the two information and then submits a two-out-of-three vote. If the information from different protection input channels points to different arms, the two-out-of-three device searches for at least two identical points according to the pairwise consistency comparison process. It does not cross-concatenate the fault phase of one channel with the arm number of another channel to form a new point. Thus, each candidate point used for voting comes from a complete combination of phase and arm number in a valid protection input channel.

[0109] If, after the transmitting end stops sending protection short frames, regular service frames still arrive normally and the cross-device heartbeat meets the original timing requirements, the receiving end will form a bridge arm fault reset if it does not receive a valid protection short frame for two consecutive device communication cycles. This will not result in a communication disconnection directly derived from the cessation of protection short frame transmission. If regular service frames, heartbeats, physical layer status, or frame integrity are all abnormal at the same time, the communication disconnection will still be determined by the physical layer, data link layer, and application layer results according to a two-out-of-three relationship. The bridge arm reset status and the communication disconnection status will be retained separately.

[0110] After the system returns to a communication state capable of legally receiving regular service frames and protection short frames, the receiving end continues to divide the protection short frame reception interval according to the actual device communication cycle. When a temporary overcurrent fault occurs in a subsequent bridge arm, the transmitting end regenerates the fault phase information, fault bridge arm number information, and protection short frame transmission request based on the fault level, and inserts the protection short frame according to the end position of the current regular service frame. This process still uses the aforementioned frame generation, CRC32 check, field parsing, three-out-of-two voting, and reset counting relationship, without adding any unrecorded recovery parameters to the re-communication state.

[0111] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for fast communication between devices in a flexible DC control and protection system, applied to a transmitting device and a receiving device connected via a gigabit point-to-point Ethernet link, the method comprising: The transmitting device transmits regular service frames carrying device operating conditions and cross-device heartbeats through the gigabit point-to-point Ethernet link according to the device communication cycle that matches the device operation cycle. The field-programmable gate array of the transmitting device acquires the overcurrent fault level of the bridge arm, determines the temporary overcurrent fault based on the overcurrent fault level of the bridge arm, and obtains the fault phase information and the fault bridge arm number information. Based on the fault phase information and the fault arm number information, after the current regular service frame is sent, the shortest Ethernet protection frame carrying the fault phase information and the fault arm number information is immediately sent through the same gigabit point-to-point Ethernet link. While the temporary overcurrent fault persists, the shortest Ethernet protection frame is sent after each regular service frame is sent. The sending of the shortest Ethernet protection frame stops after the temporary overcurrent fault is resolved. The field-programmable gate array of the receiving device verifies the shortest Ethernet protection frame received. When the verification is valid, it parses the fault phase information and the fault arm number information, and drives the corresponding arm to perform a blocking protection action based on the parsing result. Based on the reception status of the shortest Ethernet protection frame within the continuous device communication cycle, when the number of consecutive non-reception cycles reaches the fault reset determination cycle number, it is determined that the temporary overcurrent fault has disappeared and the fault blocking status of the corresponding bridge arm is revoked.

2. The method for fast communication between devices in a flexible DC control and protection system according to claim 1, characterized in that, The sending of regular service frames includes: Configure the communication parameters of the pole control device, valve control device and valve interface device in the flexible DC control and protection system to gigabit fiber optic communication mode. The device communication cycle is determined based on the device's operation cycle; based on the communication parameters and the device communication cycle, standard Ethernet frames are used to transmit regular service data in the gigabit point-to-point Ethernet link, while retaining the sending and receiving logic for cross-device heartbeat messages and status interaction messages.

3. The method for rapid communication between devices in a flexible DC control and protection system according to claim 1 or 2, characterized in that, Both the regular service frame and the Ethernet shortest protection frame include a physical layer synchronization prefix and a media access control data frame body. The physical layer synchronization prefix includes a preamble and a start frame character in sequence. The media access control data frame body includes a destination media access control address, a source media access control address, a type field, a service payload, and a cyclic redundancy check field in sequence. The service payload of the regular service frame carries basic interactive data, and the shortest Ethernet protection frame adopts the minimum valid media access control frame allowed by the Gigabit Ethernet link.

4. The method for fast communication between devices in a flexible DC control and protection system according to claim 3, characterized in that, After the current regular service frame has been sent, the shortest Ethernet protection frame is sent, including: The total number of bits in the regular service frame is determined based on the number of bytes in each field of the regular service frame, and the transmission duration of the regular service frame is determined based on the ratio between the total number of bits in the regular service frame and the transmission rate of the gigabit point-to-point Ethernet link. The total number of bits in the protection frame is determined based on the number of bytes in each field of the shortest Ethernet protection frame, and the transmission duration of the protection frame is determined based on the ratio between the total number of bits in the protection frame and the transmission rate. The transmission duration of the regular service frame is added to the transmission duration of the protection frame to obtain the combined transmission duration of the post-interpolated frame.

5. The method for fast communication between devices in a flexible DC control and protection system according to claim 4, characterized in that, Obtain faulty phase information and faulty bridge arm number information, including: The transmitting device is a valve protection device. The field-programmable gate array of the valve protection device acquires the overcurrent fault level of the bridge arm, determines the temporary overcurrent fault based on the overcurrent fault level of the bridge arm, and determines the phase and bridge arm number of the temporary overcurrent fault as the fault phase information and fault bridge arm number information, respectively.

6. The method for fast communication between devices in a flexible DC control and protection system according to claim 5, characterized in that, After the current regular service frame has been sent, the shortest Ethernet protection frame is immediately sent through the same gigabit point-to-point Ethernet link, including: Configure the transmission priority of the shortest Ethernet protection frame to be higher than that of the regular service frame, and open the protection short frame interpolation channel when the regular service frame has been transmitted. The fault phase information and the fault arm number information are written into the shortest Ethernet protection frame and sent through a protection short frame data channel that is isolated from regular business data.

7. The method for fast communication between devices in a flexible DC control and protection system according to claim 6, characterized in that, Verifying and parsing the shortest Ethernet protection frame includes: Obtain the frame data participating in cyclic redundancy check according to the field arrangement order of the shortest Ethernet protection frame; perform cyclic redundancy check operation on the frame data to obtain the verification result at the receiving end; The receiving end verification result is compared with the sending end verification result carried in the cyclic redundancy check field; When the comparison is consistent, the shortest Ethernet protection frame bypasses the regular service parsing thread, reads the fault phase information and the fault arm number information at a fixed position in the service load, and performs the corresponding arm partitioning blocking according to the fault arm number information.

8. The method for fast communication between devices in a flexible DC control and protection system according to claim 7, characterized in that, Determining that the temporary overcurrent fault has disappeared includes: The receiving device uses the device communication cycle as the timing unit. When it receives a valid Ethernet shortest protection frame in the current device communication cycle, it restores the number of consecutive unreceived cycles to the initial state. If no valid Ethernet shortest protection frame is received during the current communication cycle, the number of consecutive unreceived cycles is updated. The updated number of consecutive non-received cycles is compared with the number of fault return determination cycles. When the number of fault return determination cycles is reached, the temporary overcurrent fault return of the corresponding bridge arm is determined, and the phase-specific temporary blocking of the corresponding bridge arm is released.

9. The method for fast communication between devices in a flexible DC control and protection system according to claim 8, characterized in that, The method further includes: The physical layer disconnection judgment result is formed based on the physical layer link hardware detection result, the data link layer disconnection judgment result is formed based on the data link layer frame integrity verification result, and the application layer disconnection judgment result is formed based on the timing verification result of the application layer cross-device heartbeat. Redundant voting is performed on the disconnection judgment results of each layer to obtain the communication disconnection judgment result; Based on the communication disconnection determination result, the communication disconnection fault and the bridge arm power fault are distinguished, and corresponding alarm information is sent up respectively.

10. A fast communication system between devices in a flexible DC control and protection system, the system implementing the method as described in any one of claims 1 to 9, characterized in that, This includes valve protection devices, valve protection 3-out-of-2 devices, valve interface devices, and gigabit point-to-point Ethernet links connecting the valve protection devices, valve protection 3-out-of-2 devices, and valve interface devices. The valve protection device includes a transmitting field-programmable gate array (FPGA). The transmitting FPGA is used to send regular service frames according to the device communication cycle that matches the device's operation cycle, obtain the overcurrent fault level of the bridge arm, obtain the fault phase information and fault bridge arm number information based on the overcurrent fault level of the bridge arm, and send the shortest Ethernet protection frame carrying the fault phase information and fault bridge arm number information after the current regular service frame is sent. The valve protection three-out-of-two device includes a receiver field-programmable gate array (FPGA), which is used to verify and parse the shortest Ethernet protection frame. The valve protection three-out-of-two device is used to complete logical voting based on the parsing result, issue a corresponding bridge arm locking command to the valve interface device, and determine whether the temporary overcurrent fault has disappeared based on the reception status of the shortest Ethernet protection frame within the continuous device communication cycle. The valve interface device is used to drive the corresponding bridge arm to perform a lockout protection action according to the lockout command, and to cancel the fault lockout state of the corresponding bridge arm when the temporary overcurrent fault disappears.