Edge processing method and system for nuclear fusion power generation and energy storage operation data

Abstract

The application relates to the technical field of nuclear fusion power generation and energy storage data processing, and discloses an edge processing method and system for nuclear fusion power generation and energy storage operation data, which comprises the following steps: forming multiple processing data copies and multiple witness data copies corresponding to the same collection identifier; respectively processing the multiple processing data copies and the multiple witness data copies by different data processing domains, and generating encoding fragments carrying range information according to different encoding tracks; generating witness fragments by a witness node; distributing network paths and sending the fragments; respectively recovering candidate operation data, verifying difference standard data blocks by the witness fragments, and determining abnormal data processing domains; excluding abnormal encoding fragments, recovering target operation data by using effective fragments of other ranges, and determining a receiving state according to the number of the ranges corresponding to the effective fragments. The application can identify and exclude processing abnormalities generated before operation data forms a message, and improves the transmission reliability of nuclear fusion power generation and energy storage operation data.

Description

Technical Field

[0001] This invention relates to the field of nuclear fusion power generation and energy storage data processing technology, and in particular to an edge processing method and system for nuclear fusion power generation and energy storage operation data. Background Technology

[0002] During operation, nuclear fusion power generation devices and their associated energy storage devices continuously generate operational data, including equipment status, power status, diagnostic status, and protection status. This operational data is typically first parsed, formatted, and used to generate communication messages by edge nodes on the device side, before being transmitted to the data center via an industrial communication network.

[0003] Related technologies typically employ link verification, error correction coding, or multipath transmission to handle data corruption and packet loss after message formation. When processors, memory, or internal communication components in an edge node are affected by radiation, electromagnetic interference, or power supply anomalies, the operational data may have changed before message formation. Because edge nodes can regenerate correct message verification information based on the altered data, the receiving end may still identify erroneous data as complete data after performing routine link verification.

[0004] Related technologies can also involve multiple processing modules processing the same running data, selecting one result from multiple processing outcomes, and then uniformly generating error correction fragments based on the selected result. If the selected result is incorrect, all error correction fragments will be generated from that incorrect result, which can only improve the transmission success rate of the incorrect result, but cannot eliminate internal errors at the sending end. Furthermore, recovering data solely based on the number of received fragments cannot determine whether the valid fragments are concentrated in the same anomaly processing range. Summary of the Invention

[0005] The main purpose of this application is to provide an edge processing method and system for nuclear fusion power generation and energy storage operation data, which aims to solve the technical problem that error correction sharding is generated based on a single processing result, resulting in internal errors at the sending end not being identified and eliminated through the sharding recovery process.

[0006] To achieve the above objectives, the present invention provides an edge processing method for nuclear fusion power generation and energy storage operation data, the method comprising the following steps: Acquire operational data to form at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies; Each data processing domain processes a pre-established corresponding copy of the processed data to obtain a processing result. The processing results are then encoded according to different encoding trajectories that are mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information. The generation range information is configured to record the runtime resources on which the encoded fragment group is generated. The witness node processes the witness data copy to obtain witness shard groups, allocates network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and sends them to the receiving node; The receiving node recovers the candidate running data based on each of the coding fragment groups, converts each of the candidate running data into the standard data block sequence, and uses the witness fragment group to verify the difference standard data blocks between each of the candidate running data to determine the abnormal data processing domain. Exclude the encoded fragments corresponding to the abnormal data processing domain, recover the target running data based on the remaining encoded fragments and the witness fragments that can be used for recovery, and determine the reception status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

[0007] To achieve the above objectives, the present invention also provides an edge processing system for nuclear fusion power generation and energy storage operation data, comprising: A forming module is used to acquire runtime data and form at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies; An encoding module is used to process a pre-established corresponding copy of the processed data by each data processing domain to obtain a processing result, and to encode each processing result according to encoding trajectories that are different from each other but respectively mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information; wherein, the generation range information is configured to record the runtime resources on which the encoded fragment group is generated; The sending module is used to process the witness data copy by the witness node to obtain witness shard groups, allocate network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and send them to the receiving node; The identification module is used to recover candidate running data by the receiving node based on each of the coded fragment groups, convert each candidate running data into the standard data block sequence, and use the witness fragment group to verify the difference standard data blocks between each candidate running data to determine the abnormal data processing domain. The recovery module is used to exclude the encoded fragments corresponding to the abnormal data processing domain, recover the target running data based on the remaining encoded fragments and the witness fragments that can be recovered, and determine the reception status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

[0008] One or more technical solutions proposed in this application have at least the following technical effects: This application enables different data processing domains to generate coded fragments based on their respective processed data copies, eliminating the need to pre-select one result from multiple processing results and uniformly generate all error-correcting fragments. Even if an error occurs in one data processing domain before message formation, fragments generated by other data processing domains and witness nodes can still serve as independent sources of recovery.

[0009] This application employs encoding trajectories that can map to the same standard data block sequence but have different data arrangement relationships and fragmentation boundaries, enabling the same standard data block to correspond to different fragmentation positions in different encoding fragmentation groups. The fragmentation position is used to locate the relevant encoding fragments and recalculation range, and, in conjunction with the standard data block summary before encoding, the recalculation results with the same parameters, and the retransmission results, distinguishes between abnormal running data processing, fragmentation encoding, and network transmission.

[0010] This application uses the recoverable conditions of each target standard data block or preset recovery group and the coverage of the separated generation range as the formal recovery conditions, avoiding the use of fragments that cannot be fully recovered or whose sources are concentrated simply because the total number of fragments meets the requirements, and enabling the acquisition of supplementary fragments from the uncovered generation range when the source is insufficient.

[0011] This application can also locate the standard data blocks that have conflicted, reprocess and repair only the corresponding data range, and accurately correct subsequent historical data anomalies based on the generated range information. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments are briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0014] Figure 1 This is a flowchart illustrating an embodiment of the edge processing method for nuclear fusion power generation and energy storage operation data in this application. Figure 2 This is a schematic diagram showing the correspondence between the first and second coding trajectories of this application; Figure 3 This is a schematic diagram illustrating the process of restoring candidate runtime data in this application; Figure 4 This is a schematic diagram of the abnormal data processing domain identification process in this application; Figure 5 This is a schematic diagram illustrating the scope of conflict in this application; Figure 6 This is a schematic diagram of the edge processing system for nuclear fusion power generation and energy storage operation data in this application.

[0015] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] It should be understood that the specific embodiments described herein are only for explaining the technical solutions of this application and are not intended to limit this application. To better understand the technical solutions of this application, a detailed description is provided below in conjunction with the accompanying drawings and specific embodiments.

[0017] The main solution of this application embodiment is as follows: multiple distinct data copies are formed for the running data corresponding to the same acquisition identifier; different data processing domains process and encode their respective data copies; witness nodes generate witness fragments based on independently acquired data copies; receiving nodes restore candidate running data respectively, use witness fragments to identify abnormal data processing domains, and exclude encoded fragments generated by abnormal data processing domains; and restore target running data when each target standard data block or preset recovery group meets the recoverable conditions and separation range coverage requirements.

[0018] Error-correcting fragments in related technologies are generally generated based on the same result to be sent, which can handle fragment loss or content corruption that occurs during network transmission, but cannot correct content errors that already exist in the result to be sent before encoding. This application enables each data processing domain to directly generate encoded fragments based on its own processing results, maintaining a correspondence between the source of the processing results and the source of the fragments. The receiving node can identify and exclude abnormal data processing domains by generating range relationships.

[0019] It should be noted that the execution entity in this embodiment can be an edge processing system located on the side of a nuclear fusion power generation device, an energy storage device, or a plant-level communication network. The edge processing system has data acquisition, data processing, encoding, network communication, and program execution functions, and can communicate with at least one witness node and at least one receiving node.

[0020] The operating data referred to in this application may be digital data output by a nuclear fusion power generation device, an energy storage device, or its associated measuring, diagnostic, and protection equipment. This application performs edge processing and network transmission on the operating data, without directly altering the state of the nuclear fusion reaction or the energy storage state of the energy storage device.

[0021] This application provides an edge processing method for nuclear fusion power generation and energy storage operation data, referring to... Figure 1 The method includes steps S100 to S500.

[0022] Step S100: Obtain running data to form at least two processing data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each processing data copy.

[0023] It should be noted that this step is used to provide mutually distinguishable data inputs for different data processing domains and witness nodes. Unlike establishing multiple access references to the same storage address, in this embodiment, each data copy is formed through different read, write, or acquisition processes, thereby reducing the possibility that an abnormal storage area or abnormal read process will simultaneously affect all data copies. Step S100 includes the following sub-steps: Step S110: Identify running resources and generate range information.

[0024] In this embodiment, after the edge node starts, it reads the hardware configuration, storage configuration, software task configuration, and encoding configuration to identify the runtime resources involved in data processing and encoding. These runtime resources can be divided into resource categories, and each resource within a category has a unique resource identifier. Based on the actual usage relationships of the resources, the edge node writes the resource identifiers invoked by the same data processing domain into the corresponding generation range information.

[0025] For example, if the first data processing domain calls processor core A, storage area A, and the first encoding unit, and the second data processing domain calls processor core B, storage area B, and the second encoding unit, then two generation ranges are formed respectively. If the two data processing domains share the same power supply component or internal communication component, this sharing relationship is also written into the generation range information. In this embodiment, the two are not considered completely separate simply because they have different processors.

[0026] In one executable implementation, processing resource identifiers, storage resource identifiers, power supply resource identifiers, and coding resource identifiers are set as critical resource identifiers. The critical resource identifiers in two generation ranges are compared. If two or more of the same critical resource identifiers are present, the two generation ranges are determined to be overlapping generation ranges; if fewer than two of the same critical resource identifiers are present, the two generation ranges are determined to be separate generation ranges.

[0027] In this embodiment, the generation range information is used to describe the runtime resources on which the encoded fragments depend. Different network paths can only separate the transmission process after fragmentation, but cannot eliminate data errors that already exist before fragmentation. By recording the generation range, the receiving node can exclude fragments that depend on a certain runtime resource when it detects an anomaly in that runtime resource.

[0028] Step S120: Receive running data and assign a collection identifier.

[0029] In this embodiment, the data acquisition module receives the operating data output by the nuclear fusion power generation device or energy storage device through the industrial communication interface, and determines whether a set of operating data is complete based on the data length, end field or group status in the communication protocol.

[0030] After confirming the completeness of a set of operational data, the data acquisition module assigns an acquisition identifier to that set of data. The acquisition identifier can consist of a data source identifier and a sequential number. For example, the data source identifier for a specific power conversion device remains unchanged, and the sequential number increments by 1 for each complete set of operational data generated by that device.

[0031] It is easy to understand that the acquisition identifier is used to associate the processing data copy, witness data copy, encoded fragment, candidate running data, and target running data. As long as it can prevent data from different acquisition events from being combined into the same recovery result at the receiving node, it can be used as the acquisition identifier described in this application.

[0032] Step S130: Read the running data corresponding to the collection identifier, and write the data read each time to different storage areas to obtain at least two copies of the processed data.

[0033] In this embodiment, an independent copy management module can be set up to perform at least two distinct read and write operations on the running data corresponding to the same acquisition identifier. The first read operation data in the acquisition buffer is written to a first storage area to form a first processed data copy. The second read operation data is written to a second storage area to form a second processed data copy. Then, the first processed data copy is provided to a first data processing domain, and the second processed data copy is provided to a second data processing domain.

[0034] It's important to note that the first and second processed data copies cannot simply refer to the same storage content through two different address references. If two data processing domains read from the same storage address, and the data at that address changes, both domains may retrieve the same erroneous content. Writing to different storage areas makes local storage anomalies more easily manifested as replica differences.

[0035] After each processed data copy is written, a copy summary can be generated to check for significant differences in the writing results. If the copy summaries of the first and second processed data copies are different, both processed data copies are retained and a copy anomaly is recorded, rather than using one processed data copy to overwrite the other.

[0036] In one executable implementation, the length of a set of runtime data is 4096 bytes, and the first and second storage areas are located in different memories. If the replica digests of the two copies are different, two reads can be performed again from the acquisition buffer; if they are still inconsistent after rewriting, the corresponding runtime data is marked as data to be verified, and the required preset number of sources is increased during subsequent recovery.

[0037] Step S140: Obtain the running data through a collection process that is different from that of each processing data copy to obtain the witness data copy.

[0038] In this embodiment, the witness data copy should be distinct from the processing results formed by the first data processing domain and the second data processing domain, and then the witness data copy is provided to the witness node. The witness node can directly obtain the data source output through an independent acquisition interface, obtain the original operating data through an independent communication channel, or synchronously receive the same operating data from another physical edge node.

[0039] When equipment conditions are limited, the witness data copy can be read a third time independently by the data acquisition module and written to the storage area corresponding to the witness node. This storage area does not share the same storage control path as the storage areas used by each processed data copy.

[0040] It should be noted that the witness data copy is used to provide a data source independent of the processing results. In this embodiment, subsequent witness shards are formed independently based on the witness data copy and cannot be copied from any coded shard group to avoid the same processing error from simultaneously entering the candidate running data and the witness shard.

[0041] Step S200: Each data processing domain processes a pre-established corresponding copy of the processed data to obtain a processing result. The processing results are then encoded according to different encoding trajectories that are mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information. The generation range information is configured to record the runtime resources on which the encoded fragment group is generated.

[0042] It should be noted that this step is used to establish a direct correspondence between the processing results, encoded fragments, and generated ranges. Each data processing domain does not copy the final processing results from each other; instead, they complete the processing and encoding separately. Different encoded trajectories can all be recovered to the same standard data block sequence, but the same standard data block corresponds to different encoded fragments in different encoded trajectories. Step S200 includes the following sub-steps: Step S210: Each data processing domain processes the corresponding processed data copy to obtain a processing result, and the two processing results are respectively determined as the first processing result and the second processing result.

[0043] In this embodiment, the first data processing domain reads a copy of the first processed data and executes data processing according to the first processing path. The first processing path may sequentially check the data length, identify the field directory, extract the target field, convert the data type, check the field range, and arrange the fields according to a unified structure to obtain the first processing result.

[0044] The second data processing domain reads the second processed data copy and executes data processing according to the second processing path. The second processing path can sequentially extract the target field, check the data length, convert the data type, identify the field directory, check the field range, and arrange the fields according to the unified structure to obtain the second processing result.

[0045] The first and second processing paths produce results in the same format, but their internal actions are performed in different orders. Normally, the first and second processing results should have the same content; however, when an anomaly occurs in a software task, field parsing process, or storage / retrieval process, different processing paths can increase the likelihood that the anomaly will manifest as different results.

[0046] Step S220: Divide the data into a first standard data block sequence and a second standard data block sequence with the same block number according to a unified data structure.

[0047] In this embodiment, a standard data block is a data unit obtained by dividing the processing results according to a unified data structure. The first data processing domain and the second data processing domain divide the first processing result and the second processing result into a first standard data block sequence and a second standard data block sequence, respectively, according to the same field order and block number.

[0048] Standard data blocks can be divided according to device function, field category, or preset data length. For example, a set of 4096 bytes of processing results can be divided into 8 standard data blocks, each recording the block number, valid data length, and corresponding field range. Before encoding, each data processing field is serialized sequentially according to the block number, valid data length, and valid data. A preset digest algorithm is used to generate a digest of the standard data blocks before encoding, and the digest algorithm identifier and digest value are recorded; one executable implementation uses the SHA-256 algorithm.

[0049] A standard data block sequence is used to establish common data locations between different coded trajectories. Although the first and second coded trajectories use different data arrangements and fragmentation boundaries, they are both converted into a standard data block sequence with the same block numbering order after the receiving node completes reverse processing. Therefore, the receiving node can locate content differences based on block numbers, unaffected by changes in the coded fragmentation positions.

[0050] Step S230: Generate coded fragment groups according to the first and second coded trajectories.

[0051] like Figure 2 As shown in the embodiment of this application, the first encoding module reads the first standard data block sequence and generates a first data fragment according to the first arrangement relationship and the first fragment boundary. The first arrangement relationship can adopt the natural numbering order of the standard data blocks, and the first fragment boundary can be set between adjacent standard data blocks.

[0052] For example, the first standard data block sequence includes blocks 1 to 8. The first encoding trajectory combines blocks 1 and 2 into first data fragment 1, blocks 3 and 4 into first data fragment 2, blocks 5 and 6 into first data fragment 3, and blocks 7 and 8 into first data fragment 4. The first encoding module also generates first check fragments based on the first data fragments to recover the corresponding content when some of the first data fragments are missing.

[0053] The second encoding module reads the second standard data block sequence and generates second data fragments according to a second arrangement relationship that is different from the first encoding trajectory and the second fragment boundary. For example, the second encoding trajectory combines block 1 and block 5 into second data fragment 1, blocks 3 and 7 into second data fragment 2, blocks 2 and 6 into second data fragment 3, and blocks 4 and 8 into second data fragment 4.

[0054] The second encoding module generates a second check fragment according to a second check combination relationship that is different from the first check combination relationship, so that the same standard data block corresponds to different combination positions in the two encoding fragment groups. The specific error correction encoding method can adopt an encoding method that can recover missing content by utilizing partial data fragments and check fragments; this application does not limit the specific algorithm.

[0055] Each encoded fragment is written with an acquisition identifier, data processing domain identifier, generation range identifier, encoding trajectory identifier, encoding parameter version, fragment number, and standard data block mapping information, and carries the corresponding pre-encoding standard data block digest algorithm identifier and digest value. The standard data block mapping information is used to describe the standard data blocks or combinations of standard data blocks contained in the current encoded fragment.

[0056] It is easy to understand that both types of encoded trajectories preserve complete mapping information, enabling receiving nodes to reconstruct a unified standard data block sequence according to the correspondence.

[0057] When an error occurs in standard data block 5 during the processing stage, the difference corresponds to block 5 in both candidate running data, but it is located in different fragments in the first and second coding fragment groups. The fragment position is only used to locate the relevant coding fragment and the recalculation range, and cannot be used alone to determine the abnormal stage; the abnormal stage is determined by combining the standard data block summary before coding, the recalculation result obtained using the same coding parameters, and the retransmission result.

[0058] Through step S200, each data processing domain forms a group of encoded fragments with different generation ranges, and each fragment has a clear mapping relationship with the corresponding standard data block.

[0059] In step S300, the witness node processes the witness data copy to obtain witness shard groups, allocates network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and sends them to the receiving node.

[0060] It should be noted that this step is used to form independent witness fragment groups, ensuring a identifiable separation between the scope of coded fragment formation and the network transmission process. Path allocation not only selects different paths based on network interface names but also checks the actual switching devices, forwarding nodes, and network areas traversed by each path. Step S300 includes the following sub-steps: Step S310: Generate witness shard groups.

[0061] In this embodiment of the application, the witness node processes the witness data copy according to the same unified data structure as the first processing result and the second processing result, and obtains a sequence of witness standard data blocks with the same block number.

[0062] Witness nodes generate verification witness fragments and recoverable witness fragments. For verification witness fragments, the witness node prioritizes selecting standard data blocks belonging to different encoding fragments in the first and second encoding trajectories, and writes the block number, valid data length, and valid data sequentially in ascending order of block number. The resulting byte sequence is then used to generate a digest using a preset digest algorithm. The verification witness fragment carries a collection identifier, a cross-block combination relationship identifier, a set of block numbers, a digest algorithm identifier, and a digest value. One executable implementation uses the SHA-256 algorithm; for example, witness fragment 1 corresponds to blocks 1, 3, and 7; witness fragment 2 corresponds to blocks 2, 4, and 8; and witness fragment 3 corresponds to blocks 1, 5, and 6.

[0063] Specifically, when a discrepancy standard data block needs to be verified, the receiving node selects the verification witness fragment containing that discrepancy standard data block and checks whether the other standard data blocks corresponding to that verification witness fragment are consistent in each candidate running data. If the other standard data blocks are consistent, the receiving node serializes each candidate discrepancy standard data block with the other standard data blocks according to the block number order, effective data length, and digest algorithm recorded in the verification witness fragment, and generates a verification digest; when the block number set, combination relationship identifier, and verification digest are all consistent, it is determined to be a match with the verification witness fragment.

[0064] When a verification witness shard contains multiple differential standard data blocks, the receiving node can select other verification witness shards to continue verification; if the existing verification witness shards cannot distinguish between the candidate running data, a supplementary witness request containing the collection identifier, the specified block number, other combined block selection rules, and the digest algorithm identifier is sent to the witness node.

[0065] Recoverable witness fragments can be generated during the initial generation stage according to a preset target coding trajectory, or they can be generated after receiving a patch request according to the target coding trajectory identifier and coding parameter version specified in the request. The witness node generates recoverable witness fragments for the witness standard data block using the same system code data rows or check matrix rows as the target coding trajectory, and writes the fragment number and standard data block mapping information. The receiving node only decodes the recoverable witness fragments together with the coding fragments whose coding trajectory identifiers and coding parameter versions are consistent. For example, when the target coding trajectory uses Reed-Solomon (4,2) erasure coding, the witness node generates supplementary fragments according to the specified data rows or check rows of the coding matrix.

[0066] Step S320: Obtain candidate network paths and transmission range information.

[0067] In this embodiment, candidate network paths can be obtained from the network control device or the local path status table by setting the path allocation module. Each candidate network path records the starting interface, intermediate network nodes, forwarding gateway, target interface, and recent transmission status.

[0068] The path allocation module generates transmission range information based on the network components traversed by the candidate network paths. If two network paths use different starting interfaces but both pass through the same main switching device, the two network paths have overlapping transmission ranges; if two network paths pass through different interfaces, different main switching devices, and different forwarding gateways, they can be determined to have separate transmission ranges.

[0069] In one feasible implementation, network interfaces, main switching devices, and forwarding gateways are designated as critical network components. When two candidate network paths have more than two identical critical network components, the two candidate network paths are grouped into the same transmission range group. The specific number mentioned above can be set to other preset values ​​based on the network topology.

[0070] It should be noted that the generation range information and the transmission range information represent different processing stages. The system pre-establishes a resource topology table to uniformly register at least one common fault range identifier among the power supply area, cabinet, physical area, radiation impact area, and network aggregation area for processing resources, coding resources, and network components; the generation range and transmission range are respectively converted into a set of common fault range identifiers and then compared.

[0071] Step S330: Allocate network paths according to the generation range and transmission range.

[0072] In this embodiment, the path allocation module preferentially allocates the first coded fragment group, the second coded fragment group, and the witness fragment group to different transmission ranges. For example, the first coded fragment group is transmitted through a first industrial communication network, the second coded fragment group is transmitted through a second optical fiber network, and the witness fragment group is transmitted through a third network where the witness node is located.

[0073] For multiple fragments used to recover the same standard data block, the path allocation module calculates the overlap degree by dividing the number of intersection identifiers of the generated range fault identifier set and the candidate path fault identifier set by the number of identifiers in their union, and prioritizes candidate network paths with an overlap degree lower than a preset path threshold; one executable implementation sets the preset path threshold to 0.5. If two data processing domains share the same power supply area identifier, at least some witness fragments are allocated to physical gateways that do not have that power supply area identifier for transmission.

[0074] In one executable implementation, the first and second coded fragment groups each include four data fragments and two check fragments. The path allocation module allocates the first coded fragment to the first path group, the second coded fragment to the second path group, and the witness fragment to the third path group, ensuring that the three path groups do not pass through the same primary switching device.

[0075] Step S340: Send each fragment and record the actual process information.

[0076] In this embodiment, a multi-path sending module can be configured to submit each encoded fragment and witness fragment to the allocated network interface, and record the fragment number, sending interface, next network node, and sending time. Intermediate forwarding devices can attach their own node identifier, or the network control device can provide the actual path record to the receiving node.

[0077] When the network changes its forwarding path during transmission, the transmission range is determined by the actual network range traversed by the fragment, not just by the path name assigned before transmission. For example, if both the first and second paths pass through the same backup switching equipment after a fault switchover, the receiving node will recognize that the two paths have overlapping transmission ranges.

[0078] The receiving node periodically sends fragment reception acknowledgments, recording received fragments and fragments that failed link verification. If a fragment is corrupted during transmission, the sending node can retransmit the same fragment via an alternative network path. If the same network path fails a preset number of times consecutively, its use is suspended and an alternative transmission range is selected.

[0079] It should be noted that the above retransmission is used to resolve network transmission anomalies. If the receiving node lacks a fragmentation generation range, the node in the uncovered generation range should reread the corresponding data copy and generate supplementary fragments.

[0080] Through step S300, the receiving node can simultaneously know the generation range and actual transmission range of each fragment, and can distinguish between anomalies that occur during the fragmentation stage and the network transmission stage.

[0081] In step S400, the receiving node recovers the candidate running data based on each of the coded fragment groups, converts each of the candidate running data into the standard data block sequence, and uses the witness fragment group to verify the difference standard data blocks between each of the candidate running data to determine the abnormal data processing domain.

[0082] like Figures 3-4 As shown, it should be noted that this step is used to evaluate the processing results formed by different data processing domains. The receiving node first groups the fragments according to the encoding trajectory and generation range, recovers the candidate running data respectively, and then performs standard data block comparison and witness verification. Step S400 includes the following sub-steps: Step S410: Perform link verification and grouping on the received fragments.

[0083] In this embodiment, after receiving the encoded fragment or witness fragment, the receiving node first performs the link verification specified by the communication protocol. Fragments that fail the link verification are set as invalid fragments, and their corresponding transmission range is recorded.

[0084] For fragments that pass the link verification, the receiving node reads the acquisition identifier, the generation range identifier, and the encoding trajectory identifier, and groups the encoding fragments that belong to the same acquisition identifier and the same encoding trajectory into the same encoding fragment group.

[0085] If multiple fragments with the same fragment number have the same content, one fragment's content is retained, along with the corresponding generation range and actual path record for each fragment; if multiple fragments with the same fragment number have different content, they are designated as conflicting fragments.

[0086] The receiving node establishes the first coded fragment group, the second coded fragment group, and the witness fragment group respectively, and checks whether the standard data block mapping information carried by each fragment is complete.

[0087] Step S420: Restore the first candidate running data and the second candidate running data respectively.

[0088] In this embodiment, when the number of valid fragments in the first coded fragment group reaches the corresponding recovery quantity, the receiving node recovers the first standard data block sequence according to the first coded trajectory. When some first data fragments are missing, the missing content is recovered using the first check fragments.

[0089] The receiving node processes the second coded fragment group in the same way and recovers the second standard data block sequence according to the second coded trajectory. Since the data arrangement order of the two coded trajectories is different, the receiving node performs reverse arrangement according to its respective mapping information, converting both sets of results into standard data block sequences with consistent block numbering order, forming the first candidate running data and the second candidate running data respectively.

[0090] When recovering candidate running data, the receiving node does not mix encoded fragments from different data processing domains. If fragments are mixed before comparing candidate running data, content generated by abnormal data processing domains may enter the same recovery result, making it impossible to determine the source of the difference.

[0091] When the number of valid fragments in a coded fragment group has not yet reached the corresponding recovery quantity, the receiving node can either recover another candidate running data and set the current data to a temporary state, or request additional coded fragments to be sent to the corresponding generation range.

[0092] Step S430: Locate the abnormal stage based on the coded trajectory.

[0093] In this embodiment of the application, the receiving node compares the first candidate running data and the second candidate running data block by block, locates the difference standard data blocks with different contents, and determines the fragment position of the difference standard data blocks in the first coding trajectory and the second coding trajectory according to the mapping relationship.

[0094] When the differences correspond to the same block number but are located in different fragment positions in different coding trajectories, the receiving node determines the relevant coding fragment and recalculation range solely based on this. The receiving node obtains the corresponding pre-coding standard data block digest and the recalculation result with the same parameters: if the content corresponding to the initial digest fails the witness verification, but the content obtained by rereading the processed data copy passes the witness verification, it is determined to be a transient anomaly in the processing stage; if the content corresponding to the pre-coding digest passes the witness verification, but the initial coding fragment is different from the recalculated coding fragment generated using the same coding parameters, it is determined to be an anomaly in the coding stage.

[0095] If the sending node retransmits the same encoded fragment as the sending record, and the receiving node's link verification returns to normal, then the previous anomaly was determined to have occurred during network transmission. The aforementioned encoded trajectory location is used to locate the recalculation object. The anomaly stage is determined based on the pre-encoding digest, the same-parameter recalculation, and the retransmission results. The anomaly data processing domain also needs to be determined in conjunction with the witness verification results.

[0096] Step S440: Verify the discrepancy standard data block using witness fragmentation.

[0097] In this embodiment of the application, the receiving node selects the witness shard containing the current difference standard data block from the witness shard group, and checks whether the other standard data blocks corresponding to the witness shard are consistent in the first candidate running data and the second candidate running data.

[0098] When other standard data blocks are consistent, the receiving node uses the difference standard data blocks in the first candidate running data and the second candidate running data respectively to serialize with the other standard data blocks according to the block number order and effective data length of the verification witness sharding record, and uses the digest algorithm to generate the first verification digest and the second verification digest.

[0099] If the first verification digest matches the digest value in the verification witness fragment, but the second verification digest does not match the digest value, then the difference standard data block in the first candidate running data is determined as the target standard data block, and the second data processing domain is determined as the abnormal data processing domain; if the second verification digest matches but the first verification digest does not match, then the first data processing domain is determined as the abnormal data processing domain.

[0100] When a witness shard involves multiple difference standard data blocks, the receiving node sequentially selects other witness shards to narrow down the difference range. If the existing witness shards cannot distinguish between two candidate running data, a supplementary witness request containing the specified block number is sent to the witness node.

[0101] In one executable implementation, the two candidate runtime data differ only in block 5, and the verification witness shard 3 corresponds to blocks 1, 5, and 6. Since blocks 1 and 6 are identical in both candidate runtime data, the receiving node serializes the two candidate blocks 5 and the identical blocks 1 and 6 in the order of blocks 1, 5, and 6, respectively, and generates a verification digest using the SHA-256 algorithm. The first verification digest matches the digest value in verification witness shard 3, while the second verification digest does not; therefore, the second data processing domain is designated as the anomalous data processing domain.

[0102] Through step S400, the receiving node can evaluate the candidate running data formed by different data processing domains respectively, and use the independently formed witness fragments to determine the abnormal data processing domains.

[0103] Step S500: Exclude the encoded fragments corresponding to the abnormal data processing domain, restore the target running data based on the remaining encoded fragments and the witness fragments that can be used for recovery, and determine the receiving status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

[0104] It should be noted that this step is used to apply the anomaly identification results to the runtime data recovery. The receiving node excludes the relevant coded fragments formed by the anomaly data processing domain and checks whether the remaining fragments meet the recoverability conditions and separation range coverage requirements of each target standard data block or preset recovery group. Step S500 includes the following sub-steps: Step S510: Determine the recoverability of the standard data block and the range coverage generated by the separation.

[0105] In this embodiment, the receiving node groups valid fragments that have passed link verification and have not been excluded according to their acquisition identifier, target encoding trajectory identifier, and encoding parameter version. It then establishes decoding relationships for each target standard data block or preset recovery group based on the fragment number and standard data block mapping information. Only encoding fragments with compatible encoding trajectory identifiers and encoding parameter versions, as well as recoverable witness fragments, enter the same decoding relationship; fragments from different encoding trajectories are not directly mixed.

[0106] When a valid fragment in a decoding relationship can uniquely identify the corresponding target standard data block or preset recovery group, the recovery condition is met. The receiving node then counts the number of separation generation ranges that can provide the corresponding data or compatible decoding relationship for each target standard data block or preset recovery group; only when all target standard data blocks or preset recovery groups meet the recovery condition and the corresponding number of separation generation ranges reaches the preset source number, is the formal recovery condition met.

[0107] In one executable implementation, the first encoded trajectory includes four data fragments and two check fragments, dividing the eight standard data blocks into four preset recovery groups. Each recovery group should have valid fragments that satisfy the encoding parameters and originate from at least two separate generation ranges. Although the receiving node can recover all the standard data blocks using the four data fragments formed by the first data processing domain, it cannot directly set the recovery result to the formal reception state if each recovery group only covers one generation range.

[0108] The preset recovery quantity represents the minimum number of fragments required to achieve unique recovery within the same compatible decoding relationship, rather than a simple count of all fragments across different encoded trajectories. Verification witness fragments carrying only digest values ​​are not included in the decoding relationship and are not counted as recovery sources; recoverable witness fragments generated according to the target encoded trajectory parameters can be counted as recovery sources for the corresponding standard data block or preset recovery group. The preset source quantity can be set to 2 or 3 depending on the importance of the data.

[0109] Step S520: Eliminate abnormal encoded fragments and restore the target running data.

[0110] In this embodiment, after the receiving node determines the abnormal data processing domain, it sets all encoded fragments generated by that data processing domain as excluded fragments. When the abnormality only involves a specific encoding task or a specific standard data block, only the encoded fragments corresponding to the corresponding task or standard data block can be excluded.

[0111] The receiving node reconstructs the target standard data block sequence using coded fragments and recoverable witness fragments that have not been excluded and belong to a compatible decoding relationship. For standard data blocks that have passed witness verification and have no discrepancies, the corresponding standard data block in the non-abnormal candidate running data can be used; for missing standard data blocks, they are recovered using check fragments or recoverable witness fragments under the same coded trajectory.

[0112] After recovery is complete, the receiving node reassembles the target runtime data according to the standard data block number and generates a target runtime data summary. The receiving node then verifies the recovery results again using witness fragmentation to prevent new data errors from occurring during fragmentation exclusion or reassembly.

[0113] In essence, this step enables error-correcting fragments to participate in error correction within the sender's internal system, rather than simply restoring pre-selected processing results. Even if an error occurs in a data processing domain before a communication message is formed, the receiving node can still utilize other fragments that generate the desired range to recover the target running data.

[0114] Step S530: Perform directional patching for the uncovered area.

[0115] In this embodiment of the application, when any target standard data block or preset recovery group does not meet the recoverable conditions or the number of separated generation ranges is insufficient, the receiving node determines the generation range that has been effectively covered by fragments, and selects the uncovered generation range from other data processing domains or witness nodes registered in the system.

[0116] The receiving node sends a patch request to nodes in the uncovered generation range. The patch request includes the acquisition identifier, target encoding trajectory identifier, encoding parameter version, standard data block to be recovered or preset recovery group, required fragment number, generation range corresponding to existing fragments, and abnormal generation ranges that are not allowed to be used.

[0117] Upon receiving a patch request, the node rereads the data copy stored on its own node, re-executes the corresponding processing and encoding according to the target encoding trajectory and encoding parameter version, and generates a supplementary patch carrying the required patch number and standard data block mapping information.

[0118] For example, the receiving node has recovered blocks 1 to 8 using the four data fragments formed by the first data processing domain, but each of the four preset recovery groups only covers one generation range of the first data processing domain. The receiving node sends a patch request to the witness node, which, based on the witness data copy and according to the encoding parameters of the first encoding trajectory, generates recoverable supplementary fragments covering the four preset recovery groups. After incorporating each supplementary fragment into the corresponding compatible decoding relationship, the receiving node re-determines the recoverability conditions and the number of separated generation ranges for each recovery group.

[0119] As is easily understood, a patch request is used to supplement the recovery source coverage of a specific target standard data block or a preset recovery group; repeatedly sending more fragments from an existing generated range can only increase the number of fragments, not the number of separate generated ranges for the corresponding data range.

[0120] Step S540: Perform conflict range recalculation and local repair.

[0121] like Figure 5 As shown in this embodiment, when the existing witness shard cannot directly determine the abnormal data processing domain, or when it is necessary to further determine the abnormal type, the receiving node determines the difference standard data block and the preset number of standard data blocks adjacent to it as the conflict range.

[0122] Adjacent standard data blocks are included in the conflict range because incorrect field lengths or parsing positions may alter the starting position of subsequent fields. For example, an incorrect field length interpretation in block 4 might change the data boundary of block 5, and simply reprocessing block 5 cannot determine the cause of the anomaly. In this case, blocks 3 through 6 can be collectively identified as the conflict range.

[0123] The receiving node sends a conflict range recalculation request to the first data processing domain, the second data processing domain, and the witness node. Each data processing domain rereads its own saved copy of the processed data, does not use the initial processing result cache as the recalculation input, and re-executes the processing and encoding actions related to the conflict range.

[0124] The witness node rereads the witness data copy and generates a recalculated witness fragment for the corresponding conflict range. The receiving node compares the pre-encoding standard data block digest, initial processing result, recalculation result, initial encoding fragment, recalculated encoding fragment, and recalculated witness fragment for each data processing domain.

[0125] When the digest obtained by rereading the processed data copy in a certain data processing domain is different from the initial digest, and the recalculation result passes the witness verification, it can be determined that a transient anomaly occurred during the initial processing. When the recalculated digest is consistent with the initial digest, but the corresponding content fails the witness verification, it can be determined that there may be a persistent anomaly in the corresponding processed data copy, storage area, or processing configuration.

[0126] If the content corresponding to the standard data block digest before encoding can be verified by witness, but the initial encoded fragment of a certain encoding trajectory is different from the recalculated encoded fragment generated using the same encoding parameters, it can be determined that the anomaly occurred in the encoding process; when the received content returns to normal after resending the same encoded fragment consistent with the sending record, it can be determined that the previous anomaly occurred in the network transmission process.

[0127] The non-abnormal data processing domain generates local repair fragments based on the recalculation results. Each local repair fragment carries a collection identifier, conflict range, and corresponding block number. The receiving node only replaces the standard data blocks corresponding to the conflict range in the target running data; after replacement, it re-executes witness verification, and then proceeds to the receiving status determination step after verification.

[0128] In one executable implementation, a set of runtime data comprises eight standard data blocks, with only block 5 differing. The receiving node can identify blocks 4 through 6 as the conflict range, and each data processing domain only reprocesses these three standard data blocks. Compared to retransmitting all 4096 bytes of runtime data, partial repair reduces the amount of supplementary data to be transmitted and processing time.

[0129] Step S550: Determine the reception status and generate reception feedback.

[0130] In this embodiment of the application, when the target running data is fully recovered, the difference standard data block and the local repair results are verified by witness, and each target standard data block or preset recovery group meets the recoverable conditions and the corresponding number of separation generation ranges reaches the preset source number, the receiving node sets the target running data to the formal receiving state.

[0131] When the target running data can be recovered, but there are still target standard data blocks or preset recovery groups that have not reached the preset source quantity, witness fragments that have not arrived completely, or conflict ranges that are still being recalculated, the receiving node will set the target running data to a temporary state. Data in the temporary state can be used for auxiliary display, but it will not be used as the final running record until further verification is completed.

[0132] If none of the candidate runtime data passes witness verification, or if the target standard data block still cannot be determined after obtaining supplementary fragments, the receiving node will set the corresponding data to an isolated state. Data in the isolated state will not be written to the formal runtime data area.

[0133] The receiving node generates a receiving feedback based on the receiving status, the abnormal data processing domain, the difference standard data block, the effective fragments actually used and their range, and sends the receiving feedback to each data processing domain and witness node.

[0134] Each data processing domain updates its operation record based on received feedback. If a second data processing domain is identified as an abnormal data processing domain after a preset number of consecutive occurrences, its formal encoding task is suspended, and the corresponding processor, storage area, and encoding configuration are checked. If encoding anomalies occur consecutively in a certain encoding trajectory, the corresponding data arrangement relationship or fragmentation boundary is adjusted.

[0135] When all data processing domains and network paths remain normal, the number of verification shards generated under normal circumstances can be reduced; when abnormal operation resources are detected, the number of witness shards can be increased or the number of preset sources can be increased.

[0136] Step S560: Correct historical running data.

[0137] In this embodiment, some runtime resource anomalies may only be detected after the runtime data has been received. For example, a memory self-test indicates that the second storage area has experienced consecutive errors within the previous 10 minutes, while the second data processing domain used the second storage area to generate multiple coded fragments during that time period.

[0138] Edge nodes query range information based on the time range corresponding to the abnormally running resources to determine which collection identifiers correspond to which coding fragments were formed by the abnormally running resources. Based on the standard data block mapping information recorded for each coding fragment, the affected standard data blocks corresponding to the coding fragments can be further determined.

[0139] The receiving node excludes the affected coded fragments and checks whether the historical target runtime data still has other valid fragments of the separated generated range. If the remaining historical fragments meet the recovery requirements, the affected standard data blocks are restored; if the number of generated ranges is insufficient, supplementary historical fragments are requested from the data processing domain or witness node that were not affected by the abnormal runtime resources.

[0140] When an anomaly only affects a portion of the standard data blocks in a single coded fragment, the receiving node only re-verifies and restores the corresponding standard data blocks, without canceling other unaffected standard data blocks under the same acquisition identifier.

[0141] After the correction is completed, the receiving node saves the original receiving status, the reason for the correction, the affected range, and the target running data after the correction, thereby keeping the historical processing traceable and avoiding invalidating all unaffected data.

[0142] The operation process of this application will be described below with reference to a specific embodiment.

[0143] In this embodiment, a first edge node and a second edge node are set on the nuclear fusion power generation device side, a witness node is set on the energy storage device side, and a receiving node is set on the plant-level data center side. The first edge node includes a first data processing domain and a first encoding module, and the second edge node includes a second data processing domain and a second encoding module.

[0144] The first data processing domain uses processor core A, storage area A, and the first encoding module, while the second data processing domain uses processor core B, storage area B, and the second encoding module. Witness nodes obtain copies of the witness data through independent acquisition interfaces. The three nodes utilize different primary power supply areas and communication exits.

[0145] The data acquisition module receives a set of power operation data with a length of 4096 bytes, assigns an acquisition identifier to it, and forms a first processed data copy, a second processed data copy, and a witness data copy.

[0146] The first data processing domain obtains the first processing result according to the first processing path, and the second data processing domain obtains the second processing result according to the second processing path. Both processing results are divided into 8 standard data blocks.

[0147] The first encoding module forms the first encoding fragment group according to the natural order of the block numbers, wherein block 1 and block 2 are located in the first data fragment 1, block 3 and block 4 are located in the first data fragment 2, block 5 and block 6 are located in the first data fragment 3, and block 7 and block 8 are located in the first data fragment 4.

[0148] The second encoding module forms a second encoding fragment group according to a data arrangement relationship distinct from the first encoding trajectory. Blocks 1 and 5 are located in second data fragment 1, blocks 3 and 7 in second data fragment 2, blocks 2 and 6 in second data fragment 3, and blocks 4 and 8 in second data fragment 4. The witness node serializes the corresponding witness standard data blocks in the order of blocks 1, 5, and 6, and generates verification witness fragments using the SHA-256 algorithm.

[0149] During processing, a content error occurs in storage area B of the second data processing domain, causing an error in block 5 of the second processing result. The second encoding module generates a second encoded fragment based on the erroneous content, but the communication verification information of each second encoded fragment is consistent with its current content.

[0150] The receiving node recovers the first candidate running data and the second candidate running data respectively, and converts them into a unified standard data block sequence. The two candidate running data differ only in block 5. Block 5 is located in the first data fragment 3 in the first coding trajectory and in the second data fragment 1 in the second coding trajectory. The receiving node uses this to locate the recalculation range of the relevant coding fragments and block 5, but does not use this to determine the abnormal stage separately.

[0151] The receiving node performs verification using the verification witness fragments corresponding to blocks 1, 5, and 6. Blocks 1 and 6 are identical in the two candidate running data sets. The receiving node substitutes the two candidate blocks 5 into each and generates an SHA-256 verification digest. The first verification digest matches the witness digest, but the second verification digest does not; therefore, the second data processing domain is determined as the abnormal data processing domain.

[0152] The receiving node excludes the relevant encoded fragments generated by the second data processing domain and recovers blocks 1 to 8 using the first encoded fragment group. Each preset recovery group meets the recoverable condition, but the fragments currently used for recovery only cover one generation range of the first data processing domain, which has not yet reached the preset source quantity of 2.

[0153] The receiving node sends a patch request to the witness node, containing the first encoded trajectory identifier, the encoding parameter version, and four preset recovery groups. The witness node rereads the witness data copy and generates recoverable supplementary fragments covering each preset recovery group according to the encoding parameters of the first encoded trajectory. After incorporating the supplementary fragments into the corresponding decoding relationships, the receiving node confirms that each recovery group is covered by the range generated separately by the first data processing domain and the witness node.

[0154] The receiving node performs witness verification on the recovery results again. After all recovery groups meet the recoverable conditions and the preset number of sources, the recovery results are set to the formal receiving state and a receiving feedback is sent. After the second data processing domain rereads the second processed data copy, block 5 still fails the witness verification, and the recalculated digest is consistent with the initial digest. Therefore, it is determined that there is a persistent anomaly in storage area B, and the system is switched to the standby storage area.

[0155] Through the above process, this application uses different processed data copies, different encoding trajectories, and independently formed witness fragments to identify the range of anomalies, and uses other ranges of anomalies to restore the correct operating data.

[0156] This application also provides an edge processing system for nuclear fusion power generation and energy storage operation data. Please refer to [link / reference]. Figure 6The edge processing system for the nuclear fusion power generation and energy storage operation data includes: A forming module is used to acquire runtime data and form at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies; An encoding module is used to process a pre-established corresponding copy of the processed data by each data processing domain to obtain a processing result, and to encode each processing result according to encoding trajectories that are different from each other but respectively mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information; wherein, the generation range information is configured to record the runtime resources on which the encoded fragment group is generated; The sending module is used to process the witness data copy by the witness node to obtain witness shard groups, allocate network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and send them to the receiving node; The identification module is used to recover candidate running data by the receiving node based on each of the coded fragment groups, convert each candidate running data into the standard data block sequence, and use the witness fragment group to verify the difference standard data blocks between each candidate running data to determine the abnormal data processing domain. The recovery module is used to exclude the encoded fragments corresponding to the abnormal data processing domain, recover the target running data based on the remaining encoded fragments and the witness fragments that can be recovered, and determine the reception status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

[0157] Compared with the prior art, the beneficial effects of the edge processing system for nuclear fusion power generation and energy storage operation data provided in this application are the same as those of the edge processing method for nuclear fusion power generation and energy storage operation data provided in the above embodiments. Furthermore, the other technical features of the edge processing system for nuclear fusion power generation and energy storage operation data are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0158] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0159] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct or indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. An edge processing method for nuclear fusion power generation and energy storage operation data, characterized in that, The method includes the following steps: Acquire operational data to form at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies; Each data processing domain processes a pre-established corresponding copy of the processed data to obtain a processing result. The processing results are then encoded according to different encoding trajectories that are mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information. The generation range information is configured to record the runtime resources on which the encoded fragment group is generated. The witness node processes the witness data copy to obtain witness shard groups, allocates network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and sends them to the receiving node; The receiving node recovers the candidate running data based on each of the coding fragment groups, converts each of the candidate running data into the standard data block sequence, and uses the witness fragment group to verify the difference standard data blocks between each of the candidate running data to determine the abnormal data processing domain. Exclude the encoded fragments corresponding to the abnormal data processing domain, recover the target running data based on the remaining encoded fragments and the witness fragments that can be used for recovery, and determine the reception status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

2. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 1, characterized in that, The steps for generating range information include: Identify the runtime resources involved in runtime data processing and encoding, and assign resource identifiers to each of the runtime resources; Write the resource identifiers and resource sharing relationships assigned to the same data processing domain into the corresponding generation scope information, and compare the key resource identifiers in different generation scope information. If the number of identical key resource identifiers reaches a preset overlap threshold, the corresponding generation range is determined as the overlap generation range; if the number does not reach the preset overlap threshold, the corresponding generation range is determined as the separation generation range.

3. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 1, characterized in that, The steps of forming at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies include: Read the running data corresponding to the collection identifier respectively, and write the data read each time to different storage areas to obtain at least two copies of the processed data; The runtime data is obtained by acquiring data through a process distinct from that of each processing data copy, resulting in a witness data copy. Each processing data copy is then provided to its corresponding data processing domain, and the witness data copy is provided to the witness node.

4. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 1, characterized in that, The step of encoding each processing result according to a different encoding trajectory that is mapped to the same standard data block sequence to obtain an encoded fragment group carrying the generation range information includes: The two processing results are respectively designated as the first processing result and the second processing result, and are divided into a first standard data block sequence and a second standard data block sequence with the same block number according to a unified data structure. The first standard data block sequence is combined according to the first arrangement relationship and the first fragmentation boundary to obtain the first coded fragment group; The second standard data block sequence is combined according to a second arrangement relationship and a second fragmentation boundary that are different from the first arrangement relationship and the first fragmentation boundary to obtain the second coded fragmentation group; The mapping relationship between each coded fragment in the first coded fragment group and the second coded fragment group and the block number is recorded respectively.

5. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 4, characterized in that, The steps of having the witness node process the witness data copy to obtain witness shard groups, allocating network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and sending them to the receiving node include: The witness data copy is processed according to the standard data block sequence to obtain the witness standard data block sequence; The witness standard data blocks belonging to different coding segments in each of the coding trajectories are selected, and the verification witness segments are generated according to the preset cross-block combination relationship. The recoverable witness segments are generated according to the coding parameters of the target coding trajectory to obtain the witness segment group. Obtain the transmission range information of the candidate network path, compare the overlap between the generated range information and the transmission range information, and allocate the encoded fragments and witness fragments used to recover the same standard data block to the candidate network path with an overlap degree lower than a preset path threshold. Each coded fragment group and the witness fragment group are sent according to the assigned candidate network path, and the network range actually traversed by each fragment is recorded.

6. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 5, characterized in that, The steps for verifying the difference standard data blocks between each candidate running data using the witness shard group to determine the abnormal data processing domain include: Compare the standard data block sequences corresponding to each candidate running data to locate the difference standard data block; Based on the mapping relationship, determine the corresponding coded fragments of the differential standard data block in different coded trajectories; The difference standard data blocks in each candidate running data are combined with other standard data blocks corresponding to the same verification witness shard according to the preset cross-block combination relationship, and the combination result is compared with the verification witness shard. The difference standard data block that can be matched with the verification witness shard is identified as the target standard data block, and the data processing domain that has not generated the target standard data block is identified as the abnormal data processing domain.

7. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 6, characterized in that, The step of determining the reception status of the target running data based on the number of generation ranges corresponding to the valid fragments participating in the recovery includes: Based on the coding trajectory identifier, coding parameters and standard data block mapping information carried by each effective fragment, determine whether each target standard data block or preset recovery group meets the recoverable conditions, and statistically separate the range number of effective fragments that meet the recoverable conditions. If each of the target standard data blocks or preset recovery groups meets the recoverable conditions, and the number of separate generation ranges corresponding to each of the target standard data blocks or preset recovery groups reaches the preset source number, then the recovered target running data is set to the formal receiving state. If there are target standard data blocks or preset recovery groups that do not meet the recoverable conditions or the preset source quantity, the uncovered production range is determined according to the production range corresponding to the existing effective fragments, and a patch request containing the target coding trajectory identifier, coding parameters and the number of the block to be recovered is sent to the nodes in the uncovered production range. The nodes within the uncovered range reread the corresponding data copy according to the encoding parameters of the target encoding trajectory and generate supplementary fragments, and redetermine the reception status based on the supplementary fragments.

8. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 7, characterized in that, The process includes the following steps before setting the target running data to the formal receiving state: The difference standard data block and a preset number of standard data blocks adjacent to the difference standard data block are defined as the conflict range; A recalculation request containing the conflict range is sent to each of the data processing domains and the witness node. Each of the data processing domains rereads the corresponding processing data copy and processes the conflict range. The witness node rereads the witness data copy and generates a recalculated witness fragment. Compare the initial processing results, recalculation results, and the recalculation witness fragments of each data processing domain, and select the standard data block that matches the recalculation witness fragment to generate a local repair fragment; The local repair fragment replaces the standard data block corresponding to the conflict range in the target running data, the witness verification is re-executed on the replaced target running data, and the receiving status is determined based on the verification result.

9. The edge processing method for nuclear fusion power generation and energy storage operation data as described in claim 8, characterized in that, After determining the reception status of the target running data based on the number of generation ranges corresponding to the valid fragments participating in the recovery, the process further includes: Based on the receiving status, the abnormal data processing domain, and the difference standard data block, a receiving feedback is generated, and the running records corresponding to each data processing domain and each encoding trajectory are updated based on the receiving feedback; When an abnormality in historical operating resources is detected, the affected coded fragments generated by the historical operating resources are determined according to the generation range information, and the affected standard data blocks corresponding to the affected coded fragments are determined according to the mapping relationship. Exclude the affected encoded fragments and use historical fragments corresponding to other production ranges to restore the affected standard data blocks, or request historical supplementary fragments from production ranges that are not affected by the historical operating resources; Update the reception status of the corresponding historical running data based on the restored standard data blocks.

10. An edge processing system for nuclear fusion power generation and energy storage operation data, characterized in that, The system includes: A forming module is used to acquire runtime data and form at least two processed data copies corresponding to the same acquisition identifier and a witness data copy corresponding to each of the processed data copies; An encoding module is used to process a pre-established corresponding copy of the processed data by each data processing domain to obtain a processing result, and to encode each processing result according to encoding trajectories that are different from each other but respectively mapped to the same standard data block sequence to obtain an encoded fragment group carrying generation range information; wherein, the generation range information is configured to record the runtime resources on which the encoded fragment group is generated; The sending module is used to process the witness data copy by the witness node to obtain witness shard groups, allocate network paths to each of the encoded shard groups and the witness shard groups according to the generated range information, and send them to the receiving node; The identification module is used to recover candidate running data by the receiving node based on each of the coded fragment groups, convert each candidate running data into the standard data block sequence, and use the witness fragment group to verify the difference standard data blocks between each candidate running data to determine the abnormal data processing domain. The recovery module is used to exclude the encoded fragments corresponding to the abnormal data processing domain, recover the target running data based on the remaining encoded fragments and the witness fragments that can be recovered, and determine the reception status of the target running data according to the number of generation ranges corresponding to the valid fragments participating in the recovery.

Images