Relay chain-based cyan-free process circuit board life cycle management system and method

Abstract

The application discloses a cyanide-free process circuit board life cycle management system and method based on a relay chain, and particularly relates to the technical field of circuit board green manufacturing and full life cycle digital thread management, which comprises collecting board identification, link identification, process data, detection data and flow data formed in the processes of raw material entering, cyanide-free copper plating process, resin brushing protection, etching, forming, detection, storage, transportation and recycling disposal of a target board, and writing the data fragments of the corresponding links in the collection order, and outputting the previous fragment set; candidate thread fragment chains are constructed through the data of each link, and admission judgment is performed according to the connection result of the previous fragment of the subsequent adjacent link, the fragments verified by the connection are written into the main thread, and the fragments not passing the connection verification are written into the collateral record, so as to generate a life cycle digital thread containing the main thread and the collateral record.

Description

Technical Field

[0001] This invention relates to the field of green manufacturing and full lifecycle digital thread management technology for circuit boards. More specifically, this invention relates to a lifecycle management system and method for cyanide-free process circuit boards based on relay chains. Background Technology

[0002] In the lifecycle management of cyanide-free circuit boards, existing solutions mainly focus on how to connect the data generated in stages such as raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, warehousing, transportation and recycling, so as to facilitate subsequent status query and responsibility traceability. The usual approach is to have each process equipment, testing station and business system record material identification, process parameters, test results and flow information respectively, and then write them into the traceability platform, business database or chain storage structure in the order of their generation to form the lifecycle record of the corresponding board or batch. For example, in the continuous production scenario of cyanide-free circuit boards with multiple varieties and small batches, the same board may undergo re-plating, re-etching, re-inspection and machine replacement, manual re-judgment and cross-shift continuation during the processing. At the same time, it is also required that the green manufacturing responsibility basis be fully preserved, the upstream and downstream processes be continuously connected, the test results be allowed to be fed back with a lag and be able to verify the previous processing status. Under these application conditions, the existing processing methods are prone to the following verifiable situation: instantaneous parameter fluctuations, short-term equipment offsets, or single re-judgment results collected in the previous process are directly written into the lifecycle master record without being verified by subsequent stages. Although this forms a seemingly continuous traceability link, problems arise in subsequent anomaly attribution, responsibility identification, or compliance verification, such as the inability to correspond between the previous record and the subsequent result, multiple contradictory process interpretations for the same board, and temporary process information being retained as formal lifecycle facts. The reason for this is that the existing method assumes that the previous data can be entered into the master record as soon as it is collected, and lacks a mechanism to make formal access judgments on the previous data based on the results of subsequent stages. The technical problem this application aims to solve is: how to avoid directly writing temporary front-end data that has not yet been verified by subsequent stages into the main thread during the construction of a digital thread for the lifecycle of a cyanide-free process circuit board based on a relay chain, thereby preventing distortion of the lifecycle master record. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a lifecycle management system and method for cyanide-free process circuit boards based on a relay chain. By constructing a candidate thread fragment chain from the data of each stage, and performing an admission determination based on the acceptance results of the previous fragments by subsequent adjacent stages, the verified fragments are written into the main thread, and the fragments that fail the acceptance verification are written into the circumstantial record, thereby generating a lifecycle digital thread containing the main thread and the circumstantial record, thus solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a lifecycle management method for cyanide-free process circuit boards based on a relay chain, comprising: S1. Collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and write them into the data segments of the corresponding links in the order of collection, and output the front segment set. S2. Using relay chain nodes, add the preceding segment identifier and the current segment identifier to each preceding segment in the preceding segment set. Form a candidate thread segment chain according to the connection relationship between the preceding segment identifier and the current segment identifier. Mark the current segment that has not obtained the subsequent link result as the segment to be verified. Output the candidate thread segment chain and the segment to be verified set. S3. Read each segment to be certified, and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the admitted segment, and the admitted segment set and the non-admitted segment set are output. S4. Write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstantial record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstantial record set.

[0005] In a preferred embodiment, it further includes: S5. Based on the lifecycle main thread and the set of supporting evidence records, perform associated expansion, output the full process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieve the corresponding supporting evidence records when reading any main thread segment, and output the traceability results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital threads.

[0006] In a preferred embodiment, S1 includes: S1-1: Read the board identifier, process identifier, process data, inspection data and flow data of the target board in the current stage, perform assembly according to the fixed field order of board identifier, process identifier, data category identifier and acquisition time identifier, and output the original record of the current stage; S1-2. Sort the data items in the original record of the current stage according to the order of collection time, and arrange the data items with the same collection time in a fixed order according to the data category identifier to form the data item sequence of the current stage, and output the current stage segment body; S1-3. Sequentially splice the current segment body with the preceding segment segments that have been collected from the target board in the order of the segments to generate the preceding segment set of the target board.

[0007] In a preferred embodiment, S2 includes: S2-1. Based on each preceding segment in the preceding segment set, extract the segment identifier of the previous preceding segment as the preceding segment identifier according to the adjacent order of each preceding segment in the preceding segment set, and extract the segment identifier of the current preceding segment as the current segment identifier, and output the segment connection pair set. S2-2. Perform sequential concatenation of each fragment connection pair in the fragment connection pair set according to the first-to-last correspondence between the previous fragment identifier and the current fragment identifier, generate a candidate thread fragment chain formed by sequentially connecting multiple previous fragments, and output the candidate thread fragment chain set. S2-3. Based on the chain tail segment before the chain in the candidate thread segment chain set, check whether the chain tail segment before the chain tail has formed a corresponding subsequent segment with the subsequent adjacent links one by one. Mark the chain tail segment before the chain tail that has not formed a corresponding subsequent segment as a segment to be verified, and output the candidate thread segment chain and the segment set to be verified.

[0008] In a preferred embodiment, S3 includes: S3-1. Read each segment to be certified from the segment to be certified set and each segment to be certified from the subsequent segment set. According to the fixed field order of board identifier, link identifier, process data, test data and circulation data, split each segment to be certified and each segment to be certified into a numerical field vector and a code value field vector respectively. Generate an initial segment pair set based on the following retention conditions: the same board identifier, adjacent link identifiers, and the same circulation destination code value in the segment to be certified and circulation source code value in the segment to be certified. Output the field alignment table and the initial segment pair set. S3-2. For each initial fragment pair in the initial fragment pair set, subtract the corresponding numerical field vectors one by one according to the field alignment table to form a process difference matrix. Encode the corresponding detection data one by one into consistent values, reverse values, and missing values ​​to form a detection relation matrix. Perform head-to-tail continuation judgment on the corresponding flow code values ​​one by one to form a flow continuation matrix. Then perform singular value decomposition on the process difference matrix and extract the singular vector group corresponding to the non-zero singular values. Perform projection reconstruction on the singular vector group of the process difference matrix to obtain a stable difference vector. Output the stable difference vector, detection relation matrix, and flow continuation matrix corresponding to each initial fragment pair.

[0009] In a preferred embodiment, S3 further includes: S3-3. For each initial segment pair, count the field position sequence in the stable difference vector, the number of reverse values ​​and missing values ​​in the detection relation matrix, and the number of breakpoints in the transition continuation matrix. Then, arrange the field position sequence, the number of reverse values, the number of missing values, and the number of breakpoints in order to form an edge cost sequence. Construct a bipartite connection graph with the segment to be verified and the subsequent segment as the two sides and each initial segment pair as the connecting edge. Perform lexicographical sorting under the constraint that each segment to be verified is connected to only one subsequent segment and each subsequent segment is connected to only one segment to be verified. Then, select the connecting edges corresponding to the unused segments to be verified and the unused subsequent segments in order according to the sorted edge cost sequence to form a provisional connection edge set. When there are connecting edges with the same edge cost sequence, write the corresponding connecting edges back to the candidate thread segment chain and perform dynamic programming sequence alignment on the previous segment identifier, the current segment identifier, and the subsequent segment identifier. Retain the connecting edges with fewer broken chain bits and output the stable connection edge set. S3-4. Determine the segments to be certified that have connecting edges in the stable receiving edge set as admitted segments and write them into the admitted segment set. Determine the segments to be certified that do not have connecting edges in the stable receiving edge set as unadmitted segments and write them into the unadmitted segment set. Output the admitted segment set and the unadmitted segment set.

[0010] In a preferred embodiment, S4 includes: S4-1. Read each admission fragment in the admission fragment set, along with its predecessor fragment identifier, current fragment identifier, and connection relationships. Construct an admission directed graph with admission fragments as nodes and connection relationships as directed edges. Form an adjacency matrix according to the corresponding column of the predecessor fragment identifier and the corresponding row of the current fragment identifier. Perform transitive closure expansion on the adjacency matrix to obtain the reachability matrix. If there is a return position from the same node to itself in the reachability matrix, delete the corresponding return edge and re-execute the transitive closure expansion until the adjacency matrices obtained in two consecutive rounds are consistent. Output a stable connection graph and a stable adjacency matrix. S4-2. Based on the stable connection graph, read the number of incoming edges and outgoing edges of each admitted segment. For the current segment with more than one incoming edge, perform a triple sorting according to the order of appearance of the preceding segment in the previous segment set, the order of the links, and the order of the segment identifier, and retain the incoming edge corresponding to the first position of the sort. For the preceding segment with more than one outgoing edge, perform a triple sorting according to the order of appearance of the current segment in the subsequent segment set, the order of the links, and the order of the segment identifier, and retain the outgoing edge corresponding to the first position of the sort. Then, write the retained connection edges back to the stable adjacency matrix and perform dynamic programming accumulation of the prefix path length table and the suffix path length table. When the same node corresponds to multiple prefix paths and multiple suffix paths, retain the connection edge whose sum of prefix path length and suffix path length is at the first position of the sort. Output the connection edge set of the main thread.

[0011] In a preferred embodiment, S4 further includes: S4-3. Concatenate the beginning and end of each admission segment according to the main thread connection edge set to form the lifecycle main thread. Calculate the prefix hash value, connection hash value, and suffix hash value for adjacent admission segments in the lifecycle main thread. The prefix hash value is generated by concatenating the segment identifier of the previous admission segment with the previous connection hash value. The connection hash value is generated by concatenating the segment identifier of the current admission segment, the identifier of the preceding segment, and the identifier of the current segment. The suffix hash value is generated by concatenating the segment identifier of the next admission segment with the connection hash value. Then, write the prefix hash value, connection hash value, and suffix hash value into the corresponding admission segment and output the lifecycle main thread. S4-4. Read each unapproved fragment in the unapproved fragment set and each approved fragment in the lifecycle main thread. Generate supporting evidence candidate pairs according to the consistency of board identifier, the adjacent or identical sequence of links, and the correspondence between the source and destination of the flow. Calculate the fragment identifier editing steps, the number of link intervals, and the number of flow breakpoints for each supporting evidence candidate pair and form a supporting evidence cost sequence in chronological order. Then, select the first approved fragment in the sorting order of the supporting evidence cost sequence as the corresponding approved fragment for each unapproved fragment. Write the fragment identifier of the unapproved fragment, the fragment identifier of the corresponding approved fragment, the supporting evidence cost sequence, and the content of the unapproved fragment into the supporting evidence record and output the supporting evidence record set.

[0012] In a preferred embodiment, S5 includes: S5-1. Based on each main thread segment in the lifecycle main thread and each circumstant record in the circumstant record set, expand each main thread segment according to the segment identifier, the preceding segment identifier and the current segment identifier in the main thread segment, and match the corresponding admission segment identifier in each circumstant record with the expanded main thread segment identifier to form a full process record sequence and a segment circumstant record correspondence table. S5-2. For each main thread segment in the whole process recording sequence, retrieve the corresponding circumstantial record synchronously according to the segment circumstantial evidence correspondence table, and perform ternary association expansion in the order of main thread segment, corresponding circumstantial record, and next main thread segment to generate thread expansion unit corresponding to each main thread segment and output thread expansion sequence. S5-3. Perform sequential splicing of each thread expansion unit in the thread expansion sequence according to the fragment order, and generate traceability results, responsibility location results, anomaly attribution results and compliance verification results according to the main thread fragment content and corresponding supporting evidence record content in the thread expansion unit, and output the life cycle digital thread result of the target board.

[0013] A relay-chain-based lifecycle management system for cyanide-free process circuit boards includes a data acquisition module, a queue chain construction module, an acceptance and admission module, a thread generation module, and a thread application module. The data acquisition module is used to collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and writes the data segments of the corresponding process in the order of acquisition, and outputs the front-end segment set; The candidate thread fragment construction module is used to add the preceding fragment identifier and the current fragment identifier to each preceding fragment in the preceding fragment set using the relay chain node, form a candidate thread fragment chain according to the connection relationship between the preceding fragment identifier and the current fragment identifier, and mark the current fragment that has not obtained the subsequent link succession result as the fragment to be verified, and output the candidate thread fragment chain and the fragment set to be verified. The acceptance module is used to read each segment to be certified and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the acceptance segment, and the acceptance segment set and the non-acceptance segment set are output. The thread generation module is used to write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstance record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstance record set. The thread application module performs associated expansion based on the lifecycle main thread and the set of supporting evidence records. It outputs the entire process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieves the corresponding supporting evidence records when reading any segment of the main thread. It outputs the tracing results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital threads.

[0014] The technical effects and advantages of this invention are as follows: 1. This solution sets up a mechanism for determining the segment to be verified, the subsequent segment to be accepted, and the entry segment to be written. This ensures that the data in the front end is verified by the subsequent steps before entering the main thread. This relatively suppresses the distortion of the lifecycle main record caused by instantaneous fluctuations, short-term offsets, and the direct writing of single re-judgment results into the main record. 2. Form a candidate thread segment chain by connecting the preceding segments, and perform item-by-item recognition by combining the processing results, detection results or flow results of subsequent adjacent links, so that the segments in the main thread have a causal correspondence between the preceding and following segments, thereby relatively improving the situation where the preceding records and the following results are difficult to correspond. 3. Divide the segments to be verified into permitted segments and unpermitted segments, and write the unpermitted segments into the corroborating evidence record instead of deleting them directly. This allows the formal life cycle facts and the process information that has not been accepted and verified to be retained in layers, thereby relatively alleviating the problem of multiple contradictory process interpretations for the same board. 4. Construct a stable connection graph, main thread connection edge set, and lifecycle main thread for the input fragments, so that the lifecycle digital threads form a consistent structure in terms of fragment connection, echo cancellation, and unique connection retention, thereby relatively improving the link stability and thread expansion consistency when tracing continuously across links. 5. Establish a set of supporting evidence records outside the main thread of the lifecycle, and synchronously associate the main thread fragments and corresponding supporting evidence records when the results are expanded, so that abnormal fragments, re-judgment fragments and non-approved fragments can still be retrieved and verified, thereby improving the completeness of the basis for responsibility positioning, anomaly attribution and compliance verification to a certain extent. Attached Figure Description

[0015] Figure 1 This is a flowchart outlining the method steps of the present invention; Figure 2 This is a schematic diagram of the system module structure of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figure 1-2 The present invention provides a lifecycle management method for cyanide-free process circuit boards based on a relay chain, comprising: S1. Collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and write them into the data segments of the corresponding links in the order of collection, and output the front segment set. This implementation describes the formation process of the preceding segment set. Its purpose is to organize the scattered data generated by the target board within a single stage into segment segments with fixed field order, clear time sequence, and traceable stage location. These segments are then continuously connected with preceding stage segments, providing a unified input for subsequent segment connection, continuation determination, and main thread generation. The basic principle is: first, data fields are standardized and assembled within the current stage; then, data items are sorted and segment bodies are generated within the current stage; finally, the current stage segment body is connected to the existing segment link of the target board according to the stage sequence. This implementation process includes the following steps: The purpose of S1-1 is to organize the original data collected in the current stage into original records of the current stage with fixed field positions, so as to eliminate the impact of inconsistent output formats from different acquisition terminals on subsequent sorting and splicing. The input quantities are the board identifier, stage identifier, process data, test data and flow data of the target board in the current stage. Among them, the board identifier is used to uniquely identify the target board, the stage identifier is used to indicate the position of the current stage in the predetermined stage sequence table, the process data includes process parameter name, parameter value, parameter unit, acquisition time and station identifier, the test data includes test item name, test value, test conclusion, test time and test equipment identifier, and the flow data includes source stage, destination stage, handover time, handover batch and handover operation identifier. The processing action is to assemble the data according to a fixed field sequence of part identifier, process identifier, data category identifier, and acquisition time identifier. The data category identifier is fixedly divided into process type, inspection type, and circulation type. The acquisition time identifier uses the timestamp written by the acquisition terminal of the current process. During assembly, the part identifier and process identifier are written first, followed by the data category identifier, data item name, data item content, and acquisition time identifier in sequence, so that different types of data have fixed field positions within the same record structure. The output is the original record of the current process, which is written to the current process buffer for S1-2 to read. The exception or missing data handling is as follows: when the part identifier is missing, the assembly of the data is terminated and written to the missing record table; when the process identifier is missing, the process identifier is filled in according to the preset process mapping table corresponding to the workstation of the acquisition terminal; when the data item content is missing, the field position is retained and a null value mark is written to avoid subsequent field position misalignment. The purpose of S1-2 is to organize the discrete data items in the original record of the current stage into a unique segment of the current stage, so as to ensure that the data sequence and category arrangement within the same stage are fixed. The input is the original record of the current stage output by S1-1. The processing action includes two levels of sorting: first, sorting each data item in the original record of the current stage according to the acquisition time identifier, with data items acquired earlier placed first and data items acquired later placed last; then, sorting data items with the same acquisition time according to the data category identifier in a fixed order, with the fixed order being process category first, inspection category in the middle, and circulation category last. After sorting, the sorted data items are concatenated into a data item sequence, and the segment identifier of the current stage is written at the beginning of the data item sequence. The segment identifier is formed by concatenating the board identifier, stage identifier, and current stage segment number in a fixed order. The current stage segment number increases in order of generation of the target board's segments within the stage. The output is the current stage segment body, which is written to the stage segment table for S1-3 to read. The handling of anomalies or missing data is as follows: when two data items are collected at the same time and have the same data category identifier, they are arranged according to the order in which they were written in the original record of the current stage. When the same target board is reworked or re-inspected in the current stage, different current stage segment bodies are formed and different current stage segment numbers are assigned to them. The purpose of S1-3 is to connect the current segment body to the segment link already formed on the target board, generating a front-end segment set for subsequent connection processing. The input is the current segment body output by S1-2 and the preceding segment body that has been collected from the target board. The preceding segment body is read from the front-end segment storage area according to the board identifier of the target board. The processing action is to first determine the position of the current segment body according to the segment sequence table, and then arrange the preceding segment body in the segment order and splice it with the current segment body in sequence. The segment order table is fixed as raw material entry, cyanide-free copper plating process, resin brushing protection, etching, molding, testing, warehousing, transportation and recycling. When there is a conflict between the acquisition time sequence and the segment order, the segment order shall prevail, and the order of data items formed by S1-2 shall be retained within the same segment. After splicing, each segment is written into the same preceding segment set according to its sequential position, and the sequential positional relationship between adjacent segments is recorded; the output is the preceding segment set of the target board, which is written into the preceding segment set storage area for S2 to read; the abnormal or missing handling is as follows: if the current target board does not have a preceding segment, the current segment body is used as the starting segment of the preceding segment set and written directly; if there are repeated segments in the read preceding segment, they are distinguished by the rework round number or the re-inspection round number before participating in splicing to avoid overlapping of the same segment; Through the above processing, the data generated in the current stage is first organized into the original record of the current stage with a fixed field order, then organized into the current stage fragment body with a unique order, and finally written into the previous segment set containing the fragments of the preceding stage. This ensures that the target board has a unified field structure, unified fragment identifier, and unified stage order before entering the subsequent relay chain connection processing. This processing directly limits the input boundary of the subsequent candidate thread fragment chain, avoiding the bias in the succession judgment caused by field misalignment, disordered time order, or unclear stage position. In practical applications: For example, during the cyanide-free copper plating process, a target board sequentially generates records of chemical composition, current parameters, thickness detection, and transfer. The acquisition end first writes these data, along with the board identifier and process identifier, into the original record of the current process. Then, it forms a copper plating process segment body according to the acquisition time and data category. Subsequently, this copper plating process segment body is attached to the raw material entry process segment to form a front-end segment set containing the raw material entry segment and the cyanide-free copper plating process segment, which is then used by subsequent steps to generate candidate thread segment chains.

[0018] S2. Using relay chain nodes, add the preceding segment identifier and the current segment identifier to each preceding segment in the preceding segment set. Form a candidate thread segment chain according to the connection relationship between the preceding segment identifier and the current segment identifier. Mark the current segment that has not obtained the subsequent link result as the segment to be verified. Output the candidate thread segment chain and the segment to be verified set. This implementation describes the generation process of candidate thread segment chains and the set of segments to be verified. Its purpose is to organize the discrete segments in the preceding segment set into candidate chains with clear beginning and end connections, and to identify the tail segments that have not yet obtained subsequent succession results, serving as direct input for subsequent admission determination. The basic principle is: first, extract the connection relationships between adjacent segments from within the preceding segment set; then, form one or more candidate thread segment chains based on these connection relationships; finally, use the tail segment as the verification object to determine whether its corresponding subsequent adjacent segments have formed subsequent segments, thereby outputting the set of segments to be verified. This implementation process includes the following steps: The purpose of S2-1 is to provide a unified fragment connection basis for subsequent thread concatenation. Its working mechanism is to convert two adjacent fragments in the previous fragment set into connection pairs with fixed directions. The input is the previous fragment set of the target board, where each previous fragment in the previous fragment set has a fragment identifier and a link identifier, and the previous fragments have been written in link order. The processing action is to traverse each previous fragment in the adjacent order of the previous fragment set. For the first previous fragment, write an empty identifier into its previous fragment identifier field, and extract the fragment identifier of the first previous fragment as the current fragment identifier. For the remaining previous fragments, extract the fragment identifier of the previous previous fragment as the previous fragment identifier, extract the fragment identifier of the current previous fragment as the current fragment identifier, and then write the previous fragment identifier and the current fragment identifier into a fragment connection pair record in a fixed field order. The output is a set of fragment join pairs, which is written to the join pair buffer for S2-2 to read. The exception or missing information is handled as follows: when the current segment fragment set contains only one previous segment fragment, a single fragment join pair is directly generated with the previous segment identifier as the null identifier and the current segment identifier as the identifier of the previous segment fragment. When the identifier of the previous segment fragment is missing, it does not participate in the generation of fragment join pairs, and the previous segment fragment is written to the missing segment table to avoid the formation of broken links in the future. The purpose of S2-2 is to transform the local connection relationships in the fragment connection pair set into a continuously expandable candidate thread fragment chain. Its working mechanism is to complete the sequential concatenation of fragment connection pairs based on the head-to-tail correspondence between the previous fragment identifier and the current fragment identifier. The input is the fragment connection pair set output by S2-1. The processing action is to first start with the fragment connection pair with the previous fragment identifier being empty as the first connection pair of the chain. Then, in the remaining fragment connection pairs, find the next connection pair whose previous fragment identifier is the same as the current fragment identifier in the current tail connection pair. Connect the next connection pair to the current tail of the chain and continue to search backward in the same way until there is no next connection pair with the head-to-tail correspondence, thus forming a candidate thread fragment chain. When there are multiple connection pairs with empty preceding segment identifiers in the fragment connection pair set, or when there are remaining connection pairs that have not been absorbed by the current candidate thread fragment chain, other candidate thread fragment chains are generated in the same way, thus allowing a target board to form multiple candidate thread fragment chains simultaneously; the output is a set of candidate thread fragment chains, and the set of candidate thread fragment chains is written to the candidate chain storage area for S2-3 to read; the exception or missing handling is as follows: if two connection pairs have the same first and last identifiers but the link order is not continuous during the concatenation process, the concatenation is not performed, and the latter connection pair is retained as a candidate for the starting point of the new chain; if the same current segment identifier corresponds to multiple subsequent connection pairs, all of them are retained into the candidate thread fragment chain set and are not deleted in this step, and the corresponding relationship is processed by the subsequent admission determination step; The purpose of S2-3 is to identify the tail segment of the candidate thread segment chain that has not yet obtained the subsequent succession result. Its working mechanism is to perform a subsequent segment formation status check on the tail segment of each candidate thread segment chain. The input is the candidate thread segment chain set and the subsequent adjacent segment index table output by S2-2. The subsequent adjacent segment index table is established by the subsequent adjacent segment generated according to the board identifier, segment identifier and flow source code value. The processing action is to extract the tail segment of each candidate thread segment chain, read the board identifier, segment identifier and flow destination code value of the tail segment, determine its subsequent adjacent segment identifier according to the preset segment sequence table, and search the subsequent adjacent segment index table for the subsequent segment that simultaneously satisfies the following conditions: consistent board identifier, consistent subsequent adjacent segment identifier, and the flow source code value in the subsequent segment is consistent with the flow destination code value in the tail segment. When a corresponding subsequent segment is found, it is determined that the preceding segment of the chain tail has obtained the subsequent continuation result and is not written into the segment set to be verified; when no corresponding subsequent segment is found, the preceding segment of the chain tail is marked as a segment to be verified and written into the segment set to be verified; the output is a candidate thread segment chain and a segment set to be verified, and is written into the candidate chain result table and the segment set to be verified respectively for S3 to read; the exception or missing handling is as follows: when the flow destination code value of the preceding segment of the chain tail is missing, the preceding segment of the chain tail is directly written into the segment set to be verified and the flow missing is noted in the verification mark; when there is no subsequent adjacent link of the link identifier in the link sequence table, the preceding segment of the chain tail is identified as the chain tail termination segment and is not written into the segment set to be verified. Through the above processing, the front-end fragment set is first organized into a fragment connection pair set, then expanded into one or more candidate thread fragment chains, and finally the chain tail front-end fragments that have not yet formed subsequent succession relationships are identified from each candidate thread fragment chain and written into the fragment set to be verified. This ensures that the subsequent succession determination steps only perform calculations on the chain tail fragments that truly need to be verified, avoid repeated processing of fragments that have formed subsequent succession relationships, and at the same time retain the multi-candidate chain structure in rework, re-inspection, or parallel flow scenarios. In practical applications: For example, if a target board has already formed a raw material entry segment, a cyanide-free copper plating process segment, and an etching segment, the system first extracts two segment connection pairs in adjacent order: "raw material entry segment, cyanide-free copper plating process segment" and "cyanide-free copper plating process segment, etching segment". Then, these pairs are concatenated to form a candidate thread segment chain of "raw material entry, cyanide-free copper plating process, resin protection brushing, etching". Subsequently, the etching segment is used as the front segment of the chain tail to search the forming stage segment index table. If no subsequent segment with the same board identifier, the same forming stage, and the same transfer source code value as the transfer destination code value of the etching segment is found, the etching segment is marked as a segment to be verified and written into the segment set to be verified, so that subsequent steps can continue to determine whether it can enter the main thread.

[0019] S3. Read each segment to be certified, and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the admitted segment, and the admitted segment set and the non-admitted segment set are output. This implementation describes the acceptance determination process for the segment to be certified. Its purpose is to identify which segments from the tail segment of the candidate thread segment chain have been truly accepted by subsequent adjacent stages, and accordingly divide the segment to be certified into admitted segments and disapproved segments. The basic principle is as follows: first, establish a one-to-one correspondence between the segments to be certified and the subsequent segments, and initial segment pairs; then, construct differential results and continuation results from three perspectives: process data, detection data, and flow data; subsequently, uniformly convert the calculation results from the three perspectives into connection edge costs and perform a unique connection solution within the two-part acceptance graph; finally, complete the admitted and disapproved writing based on the solution results. This implementation process includes the following steps: The purpose of S3-1 is to establish a unified input boundary for subsequent acceptance calculations. Its mechanism is to first complete field splitting and field alignment, and then filter out the combination of segments that can participate in the acceptance calculation based on object consistency and link continuity. The input quantities are each segment to be certified in the segment set to be certified and each segment in the subsequent segment set. The segment to be certified and the segment in the subsequent segment set contain board identifier, link identifier, process data, test data and circulation data. The continuous parameter values ​​in the process data and the numerical test values ​​in the test data are used as numerical fields, and the board identifier, link identifier, test conclusion code, circulation source code and circulation destination code are used as code value fields. The processing action is to first establish a field bit mapping table according to a fixed field order. In the field bit mapping table, the same field bits are assigned to process items, test items and circulation items with the same name. Then, each segment to be certified and each segment in the subsequent segment are split into numerical field vectors and code value field vectors respectively. Subsequently, using the segment to be certified as a reference, subsequent segments are traversed, and retention judgments are performed on the board identifier, process identifier, and transfer code. Among them, the consistency of the board identifier is used as the object consistency condition, the judgment of the adjacent process identifier is based on the preset process sequence table, and the same transfer destination code value in the segment to be certified and the same transfer source code value in the subsequent segment are used as the start and end continuation condition. When all three conditions are met, an initial segment pair record is generated. The output is a field alignment table and an initial segment pair set, which are written to the field bit mapping area and the initial segment pair buffer area for S3-2 to read, respectively. The abnormal or missing handling is as follows: when a certain detection item or process item in the segment to be certified or the subsequent segment is missing, only the field bit is retained in the field bit mapping table and a blank mark is written in the corresponding vector position. The segment pair is not directly removed. When there is no subsequent segment that meets the three retention conditions for the segment to be certified, no initial segment pair is generated. The segment to be certified is directly entered into the unadmitted candidate table after the end of this round of calculation. The purpose of S3-2 is to calculate the degree of continuity between the segment to be certified and the subsequent segment from three dimensions: process change, detection correspondence, and flow continuation. Its mechanism is to uniformly convert different types of data into comparable matrix and vector results, providing a unified calculation basis for subsequent edge selection. The inputs are the field alignment table and initial segment pair set output by S3-1. The processing includes three parts: First, subtracting the corresponding numerical field vectors of each initial segment pair item by item to form a process difference matrix, where the rows of the matrix correspond to process items, the columns correspond to the numerical field positions under the corresponding process item, and the matrix elements are determined by the subsequent segment... The result is obtained by subtracting the corresponding value of the segment to be verified from the corresponding value. Missing fields are left empty and do not participate in the subtraction. Second, the corresponding detection data of each initial segment pair is encoded item by item to form a detection relationship matrix. Among them, the same detection conclusion code of the same detection item in the segment to be verified and the subsequent segment is recorded as a consistent value, the opposite detection conclusion code is recorded as a reverse value, and the detection item exists only on one side is recorded as a missing value. Third, the corresponding flow code value of each initial segment pair is subjected to the first and last continuation judgment to form a flow continuation matrix. Among them, the flow destination code value in the segment to be verified and the flow source code value in the subsequent segment are the same as the continuation bit, and they are different as the breakpoint bit. After constructing the three types of matrices, singular value decomposition is performed on the process difference matrix. Singular values ​​corresponding to all-zero vectors in the decomposition result are removed, and the singular vector groups corresponding to the singular values ​​of the non-all-zero vectors are retained. The process difference matrix is ​​then projected onto these singular vector groups and reconstructed to obtain stable difference vectors. Each bit in the stable difference vector corresponds to a difference value retained after reconstruction. The output includes the stable difference vectors, detection relation matrix, and flow continuation matrix for each initial fragment pair, and is written to the calculation result table for S3-3 to read. Abnormal or missing values ​​are handled as follows: when all elements in the process difference matrix are empty or zero, the stable difference vector is written as an all-zero vector; when only missing values ​​exist in the detection relation matrix, the matrix is ​​retained for subsequent statistics and not deleted in this step; when there are breakpoints in the flow continuation matrix, the initial fragment pair is still retained for subsequent sorting and processed uniformly by S3-3. The purpose of S3-3 is to find the unique connecting edge corresponding to each segment to be verified from multiple candidate initial segment pairs. Its working mechanism is to first convert the three types of calculation results of process, detection and flow into a unified edge cost sequence, and then complete the selection of connecting edges under one-to-one constraints, and perform link write-back verification on parallel connecting edges. The input quantities are the stable difference vector, detection relation matrix and flow continuation matrix output by S3-2, and the candidate thread segment chain formed by S2. The processing action is to first count four types of quantities for each initial segment pair: First, sort the field positions in the stable difference vector from largest to smallest according to the absolute value to form a field position sequence, and arrange the field positions with the same absolute value in a fixed order; Second, count the number of reverse values ​​in the detection relation matrix; Third, count the number of missing values ​​in the detection relation matrix; Fourth, count the number of breakpoints in the flow continuation matrix; Then, the field position sequence, the number of reverse values, the number of missing values ​​and the number of breakpoints are arranged in order to form an edge cost sequence. Then, a bipartite connection graph is constructed with the segment to be verified and the subsequent segment as the two sides and each initial segment pair as the connecting edge. Under the constraint that "each segment to be verified is connected to only one subsequent segment and each subsequent segment is connected to only one segment to be verified", all connecting edges are sorted lexicographically according to the edge cost sequence. During sorting, the field position sequence is compared first, then the number of reversed values ​​is compared, then the number of missing values ​​is compared, and finally the number of breakpoints is compared. After sorting, connecting edges are read sequentially from the first to the last position. When the segment to be verified and the subsequent segment corresponding to the connecting edge are not occupied, the connecting edge is written into the provisional connection edge set until the end of the sort is reached. If there are connecting edges with the same cost sequence, the parallel connecting edges are written back to the corresponding candidate thread segment chain, and the previous segment identifier, current segment identifier, and subsequent segment identifier are combined into a ternary identifier sequence. This sequence is then aligned with the ternary identifier sequence at the expected position in the candidate thread segment chain using dynamic programming. The number of inconsistent positions is counted bit by bit as the number of broken links, and connecting edges with fewer broken links are retained. The output is a stable set of receiving edges, which is written to the receiving edge result table for S3-4 to read. The exception or missing edge handling is as follows: if a segment to be verified has no connecting edges entering the provisional receiving edge set, the segment to be verified is marked as having no receiving edges. If the number of broken links is the same after the parallel connecting edges are written back, the connecting edge with the earlier position is retained according to the lexicographical order of the connecting edges. The purpose of S3-4 is to complete the final classification of the segments to be verified based on the stable receiving edge set. Its mechanism is to transfer segments with valid receiving edges to the main thread's admission channel, and segments without valid receiving edges to the non-admission retention channel. The input is the stable receiving edge set and the segment set to be verified output by S3-3. The processing action is to traverse the segment set to be verified one by one, and for each segment, search the stable receiving edge set for a connection edge originating from that segment. When a connection edge is found, the segment to be verified is determined as an admission segment, and the segment identifier, the corresponding subsequent segment identifier, and the connection edge are recorded. The identifier is written to the admission fragment set; if no connecting edge is found, the fragment to be verified is identified as an unapproved fragment, and the identifier of the fragment to be verified and the state of no receiving edge are written to the unapproved fragment set; the output is the admission fragment set and the unapproved fragment set, which are written to the main thread admission area and the circumstantial evidence candidate area for S4 to read respectively; the abnormal or missing handling is as follows: when the same fragment to be verified has multiple connecting edges in the stable receiving edge set, only the unique connecting edge retained by S3-3 is read; after the fragment to be verified is written to the unapproved fragment set in this round, the original fragment content is not directly deleted, but the original fragment content is retained for subsequent circumstantial evidence record generation; Through the above processing, the segment to be certified first establishes field alignment relationships and initial segment pairs with the subsequent segments, and then completes the calculation of the process difference matrix, detection relationship matrix and flow continuation matrix respectively. Stable difference vectors are extracted using singular value decomposition and projection reconstruction. Subsequently, the three types of results are unified into an edge cost sequence, and one-to-one acceptance edge selection is completed in the two-part acceptance graph. Finally, the set of admitted segments and the set of non-admitted segments are output. This not only fixes the field range, calculation order and unique output rules of acceptance judgment, but also incorporates process changes, detection conclusions and flow continuation into the acceptance judgment, avoiding direct admission judgment based solely on the consistency of a single field. In practical applications: For example, if an etched segment of a target board is used as a segment to be certified, the system retrieves two candidate segments with the same board identifier and adjacent stages from the segment set of the later stages of the forming process. First, it constructs a process difference matrix, a detection relationship matrix, and a flow continuation matrix for the two segments respectively. Then, it performs singular value decomposition on the process difference matrix to obtain a stable difference vector. Subsequently, it counts the field position sequence, the number of reversed values, the number of missing values, and the number of breakpoints of the two segments to form the edge cost sequence of the two connecting edges. If the edge cost sequences of the two connecting edges are different, the connecting edges with the higher ranking are directly sorted in lexicographical order. If the edge cost sequences of the two connecting edges are the same, they are written back to the candidate thread segment chain of "raw material entry, cyanide-free copper plating process, resin brushing protection, etching" to perform sequence alignment. It compares which connecting edge produces fewer broken chain bits after being connected to the forming segment and retains the unique connecting edge accordingly. Finally, the etched segments with unique connecting edges are written into the admitted segment set, and the segments to be certified without connecting edges are written into the non-admitted segment set.

[0020] S4. Write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstance record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstance record set. This implementation describes the generation process of the lifecycle main thread and the supporting evidence record set. Its purpose is to organize the valid connection relationships in the admitted fragment set into a unique main thread connection structure, and to write the non-admitted fragments into the supporting evidence record according to their corresponding relationships, so that the target board's lifecycle digital thread simultaneously possesses a main fact chain and a supporting evidence retention chain. Its basic principle is: first, a stable connection graph is formed based on the connection relationships between admitted fragments, and the return relationships are cleared. Then, the connection edges under one-in-one-out constraints are uniquely retained within the stable connection graph. Subsequently, the lifecycle main thread is generated by concatenating the retained connection edges and written into a hash chain. Finally, supporting evidence correspondences are established between the non-admitted fragments and the admitted fragments in the main thread and written into the supporting evidence record. This implementation process includes the following steps: The purpose of S4-1 is to extract stable, non-returning edges from the local connectivity between admitted fragments. Its mechanism involves first graphically representing the connectivity, then using adjacency and reachability matrices to identify and remove returning edges. The input consists of each admitted fragment in the admitted fragment set, its preceding fragment identifier, current fragment identifier, and connectivity relationships. Each admitted fragment corresponds to a unique fragment identifier, and the connectivity relationships are given by the connection between the preceding fragment identifier and the current fragment identifier. The processing involves first constructing a admitted directed graph with each admitted fragment as a node and each connectivity relationship as a directed edge. Then, an adjacency matrix is ​​formed according to the column corresponding to the preceding fragment identifier and the row corresponding to the current fragment identifier. If a connectivity relationship exists between a column and a row in the adjacency matrix, a connectivity value is written; otherwise, an empty value is written. Then, a transitive closure expansion is performed on the adjacency matrix to obtain the reachability matrix. When a node in the reachability matrix has a reachable position from itself, it indicates that the corresponding path forms a return path. At this time, the connecting edges forming the return path are back-checked in reverse order of their appearance in the path, and the incoming edge closest to the current node is deleted. Then, the adjacency matrix is ​​reconstructed and the transitive closure expansion is performed again. Each deletion and recalculation is counted as one round. The process stops when the adjacency matrices obtained in two consecutive rounds are completely identical. The output is a stable connection graph and a stable adjacency matrix, which are written to the connection graph storage area for S4-2 to read. The exception or missing information is handled as follows: the entry fragment with an empty preceding fragment is only written as the starting node and does not participate in the return edge deletion judgment; when there is no connection between the entry fragments, each entry fragment is kept as an independent node and directly written to the stable connection graph. The purpose of S4-2 is to uniquely retain multiple incoming and multiple outgoing edges within a stable connection graph, forming a set of connection edges that can be directly used for concatenation in the main thread. Its mechanism involves first resolving conflicts between incoming and outgoing edges in a fixed order, and then using the accumulated path lengths to resolve remaining conflicts. The inputs are the stable connection graph and stable adjacency matrix output by S4-1, as well as the occurrence order and sequence information of each segment in the preceding and following segment sets. The processing involves two layers: the first layer reads the number of incoming and outgoing edges for each admitted segment. For the current segment with more than one incoming edge, perform a triple sorting based on the order of appearance, segment order, and segment identifier of the corresponding preceding segment in the preceding segment set. The incoming edge at the top of the sorting is retained, and the rest are deleted. For the preceding segment with more than one outgoing edge, perform a triple sorting based on the order of appearance, segment order, and segment identifier of the corresponding current segment in the following segment set. The outgoing edge at the top of the sorting is retained, and the rest are deleted. When the triple sorting fields are completely identical, retain the first-written connecting edges according to the order in which they are written in the stable connection graph. In the second layer, the retained connection edges are written back to the stable adjacency matrix to generate an updated adjacency matrix. Based on the updated adjacency matrix, a prefix path length table and a suffix path length table are established. The entry segment with zero incoming edges is used as the prefix starting point, and its prefix path length is initialized to 1. The entry segment with zero outgoing edges is used as the suffix ending point, and its suffix path length is initialized to 1. The path length of the remaining nodes is accumulated successively along the connection direction. When the same node corresponds to multiple prefix paths and multiple suffix paths, the sum of the prefix path length and suffix path length corresponding to each candidate connection edge is calculated. The nodes are sorted in descending order of the sum. If the sums are the same, they are sorted in descending order of the prefix path length. If they are still the same, they are sorted in the order of the segment identifier. The connection edge corresponding to the first node in the sort is retained. The output is the main thread connection edge set, which is written to the main thread connection edge table for S4-3 to read. The exception or missing edge handling is as follows: if a certain entry segment has no incoming edges and no outgoing edges after deleting conflicting edges, the entry segment is retained as an independent main thread candidate segment and is not removed in this step. The purpose of S4-3 is to form a continuous lifecycle main thread based on the main thread connection edge set, and to write the connection relationship between adjacent admitted segments in the main thread into a hash chain. Its working mechanism is to concatenate admitted segments in the order of the connection edges, and to convert the connection information of each segment with the adjacent segments before and after into a hash input with a fixed concatenation order. The input is the main thread connection edge set and the corresponding admitted segment output by S4-2. The processing action is to start from the admitted segment whose previous segment is marked as empty or has zero incoming edges, and read the subsequent admitted segments one by one according to the connection direction of the main thread connection edge set, and perform the first and last concatenation until there are no subsequent admitted segments, thus forming one or more lifecycle main threads. Subsequently, prefix hash values, connection hash values, and suffix hash values ​​are calculated for adjacent admitted segments in the main lifecycle thread. The prefix hash value is input by concatenating the previous admitted segment identifier and the previous connection hash value in a fixed order. The connection hash value is input by concatenating the current admitted segment identifier, the preceding segment identifier, and the current segment identifier in a fixed order. The suffix hash value is input by concatenating the next admitted segment identifier and the connection hash value in a fixed order. A hash operation is performed on each concatenation result to obtain the corresponding hash value. The prefix hash value, connection hash value, and suffix hash value are then written into the hash field of the current admitted segment. The output is the main lifecycle thread, which is written to the main thread's storage area for S4-4 and S5 to read. Exception or missing data handling is as follows: if the first segment of the main lifecycle thread does not have a previous admitted segment, its prefix hash value is generated by concatenating an empty identifier with an empty previous connection hash value; if the last segment does not have a next admitted segment, its suffix hash value is generated by concatenating an empty identifier with a connection hash value. The purpose of S4-4 is to attach the excluded fragments to the corresponding main thread fragments to form supporting evidence records. Its mechanism is to first generate supporting evidence candidate pairs between the excluded fragments and the main thread fragments, and then select a unique corresponding admitted fragment for each excluded fragment based on the supporting evidence cost sequence. The input consists of each excluded fragment in the excluded fragment set and each admitted fragment in the main thread whose lifecycle is output by S4-3. The processing action is to first traverse each admitted fragment in the main thread based on the excluded fragments, and generate supporting evidence candidate pairs according to the conditions of consistent board identification, the same or adjacent link order, and the beginning and end of the flow source and flow destination corresponding. Specifically, candidate pairs of supporting evidence with the same step order are retained first. If there are no candidate pairs with the same steps, then candidate pairs of supporting evidence with adjacent steps are retained. Then, for each candidate pair of supporting evidence, the number of segment identification editing steps, the number of step intervals, and the number of flow breakpoints are calculated. The number of segment identification editing steps is calculated by comparing the board identification segment, the step identification segment, and the segment sequence number segment one by one. Each change to a segment counts as one editing step. The number of step intervals is calculated based on the position difference in the step order table. The number of flow breakpoints is obtained by comparing the flow destination of the unapproved segment with the flow source of the approved segment, and the flow source of the unapproved segment with the flow destination of the approved segment. Next, the number of editing steps, the number of stage intervals, and the number of flow breakpoints of the segment identifier are arranged in chronological order to form the supporting evidence cost sequence. The supporting evidence candidate pairs corresponding to the same non-admitted segment are sorted according to the supporting evidence cost sequence, and the admitted segment at the top of the sort is selected as the corresponding admitted segment for the non-admitted segment. Finally, the segment identifier of the non-admitted segment, the segment identifier of the corresponding admitted segment, the supporting evidence cost sequence, and the content of the non-admitted segment are written into the supporting evidence record. The output is the supporting evidence record set, which is written into the supporting evidence record storage area for S5 to read. The exception or missing handling is as follows: when there is no supporting evidence candidate pair that meets the generation conditions for a certain non-admitted segment, the corresponding admitted segment identifier of the non-admitted segment is written as an empty identifier and written separately into the isolated supporting evidence table. When the supporting evidence cost sequences of multiple supporting evidence candidate pairs are completely consistent, the admitted segment with the earlier appearance order in the main thread of the life cycle is retained. Through the above processing, the local connection relationships between admitted fragments are first organized into a stable connection graph with no return path. Then, through three layers of processing (inbound edge, outbound edge, and path length), a main thread connection edge set is formed. Subsequently, a lifecycle main thread with a hash chain is generated based on this. Finally, the non-admitted fragments are associated with the corresponding admitted fragments and written into the circumstantial record set. This completes the construction of the main thread and circumstantial evidence of the target board's lifecycle digital thread. This processing not only fixes the unique formation rule of the main thread but also fixes the attachment rule of non-admitted fragments, so that when the subsequent association is expanded, both the formal lifecycle facts and the circumstantial facts that have not entered the main thread can be read at the same time. In practical applications: For example, the set of admission segments for a target board includes raw material entry segments, cyanide-free copper plating process segments, etching segments, and forming segments. The etching segment simultaneously receives two incoming edges from different preceding segments. The system first constructs an admission directed graph and deletes the connecting edges that form a return relationship through transitive closure. Then, it performs a triple sorting on the multiple incoming edges of the etching segment according to the order of appearance of the preceding segments, the order of the process, and the order of the segment identifier, retaining only the first incoming edge in the sorting. Subsequently, based on the retained main thread connecting edge set, the raw material entry segment, the cyanide-free copper plating process segment, the etching segment, and the forming segment are connected to form the lifecycle main thread, and prefix hash values, connection hash values, and suffix hash values ​​are written for adjacent segments. For the re-inspection segments among the non-admitted segments, the system searches for admission segments with the same board identifier and the same process order in the lifecycle main thread, calculates the number of segment identifier editing steps, the number of process intervals, and the number of flow breakpoints, and selects the first-ranked admission segment as the corresponding admission segment. The content of the re-inspection segment is written into the circumstantial record under this admission segment for subsequent steps to expand into a complete lifecycle digital thread.

[0021] S5. Based on the lifecycle main thread and the set of supporting evidence records, perform associated expansion, output the full process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieve the corresponding supporting evidence records when reading any main thread segment, and output the traceability results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital thread; This implementation describes the process of expanding the results of a lifecycle digital thread. Its purpose is to expand the main thread segments within the lifecycle main thread and the corresponding supporting evidence records in the supporting evidence record set into a directly readable full-process record in a unified order. Based on this, traceability results, responsibility location results, anomaly attribution results, and compliance verification results are generated. The basic principle is as follows: First, the main thread segments are expanded into a full-process record sequence according to the main thread connection order, and a correspondence is established between the main thread segments and supporting evidence records. Then, each main thread segment is synchronously linked to the corresponding supporting evidence record to form a thread expansion unit. Finally, four types of results are generated based on the main thread content and supporting evidence content in the thread expansion unit. This implementation process includes the following steps: The purpose of S5-1 is to form the main thread sequence and circumstantial evidence connection relationships required for subsequent thread expansion. Its mechanism is to first expand the main thread of the lifecycle from the start to the end, and then map each circumstantial evidence record to the expanded main thread segment. The input consists of each main thread segment in the lifecycle main thread and each circumstantial evidence record in the circumstantial evidence record set. The main thread segment includes a segment identifier, a preceding segment identifier, and a current segment identifier. The circumstantial evidence record includes the segment identifier of the unadmitted segment, the corresponding admitted segment identifier, the circumstantial evidence cost sequence, and the content of the unadmitted segment. The processing action is to start from the main thread segment with an empty preceding segment identifier, and read the subsequent main thread segments sequentially according to the correspondence between the current segment identifier and the preceding segment identifier of the next main thread segment, until no subsequent main thread segment is read, forming a complete process record sequence. Then, the circumstantial evidence records are read one by one, and the corresponding admitted segment identifier in each circumstantial evidence record is compared with the main thread segment identifier in the complete process record sequence. When they match, a segment circumstantial evidence correspondence relationship is established. When a main thread segment corresponds to multiple supporting evidence records, the supporting evidence cost sequence in the supporting evidence records is sorted and written into the supporting evidence list corresponding to the main thread segment. The output consists of a full process record sequence and a segment supporting evidence correspondence table, which are written into the full process record area and the supporting evidence correspondence area respectively for S5-2 to read. The exception or missing information is handled as follows: when there are multiple independent main threads in the life cycle main thread, they are expanded from the starting point of each main thread and then spliced ​​together according to the segment order of the starting segment to form a full process record sequence. When the corresponding admission segment identifier in the supporting evidence record is an empty identifier, it is not written into the segment supporting evidence correspondence table, but is written into the isolated supporting evidence table separately. The purpose of S5-2 is to expand the main thread segment and its corresponding supporting evidence record according to a unified unit structure. Its working mechanism is to take the main thread segment as the main position, the supporting evidence record as the auxiliary position, and the next main thread segment as the connection position to form a thread expansion unit that can be continuously spliced. The input is the full process record sequence and the segment supporting evidence correspondence table output by S5-1. The processing action is to read each main thread segment one by one according to the segment order in the full process record sequence, and simultaneously retrieve the supporting evidence record group corresponding to the current main thread segment according to the segment supporting evidence correspondence table. Then, it reads the next main thread segment in the full process record sequence of the current main thread segment, and writes "current main thread segment, corresponding supporting evidence record group, next main thread segment" into the same thread expansion unit in a fixed order. Among them, the order of multiple supporting evidence records in the corresponding supporting evidence record group remains unchanged according to the sorting result in S5-1. When the current main thread segment does not have a corresponding supporting record, the corresponding supporting record group is written as an empty group; when the current main thread segment is the end segment of the full process record sequence, the next main thread segment is written as an empty marker; the output is the thread expansion sequence, and the thread expansion sequence is written to the expansion result area for S5-3 to read; the exception or missing handling is as follows: when the supporting record content corresponding to the current main thread segment is missing, the position of the supporting record in the corresponding supporting record group is retained and a content missing marker is written to avoid destroying the corresponding order in the thread expansion unit; when the main thread segment in the full process record sequence is missing, the segment position is not skipped, but a segment missing marker is written and the thread expansion unit continues to be formed; The purpose of S5-3 is to convert the thread expansion sequence into a lifecycle digital thread result that can be directly read by the management side. Its working mechanism is to first splice the thread expansion units in the order of the segments, and then generate four types of results based on the main thread content and the supporting evidence content. The input is the thread expansion sequence output by S5-2. The processing action is to first splice the thread expansion units in the thread expansion sequence in the order of the main thread segments to obtain the complete thread expansion result of the target board; then generate the traceability result based on the main thread segment content in the thread expansion unit, wherein the traceability result outputs the process of each main thread segment in the order of the steps in the whole process record sequence. Responsibility positioning results are generated based on the link identifiers, workstation identifiers, flow relationships, and unapproved segment content in the corresponding circumstantial evidence records of the main thread segment. These results specify the responsible link and associated circumstantial evidence for each main thread segment. Anomaly attribution results are generated based on the comparison between the unapproved segment content in the corresponding circumstantial evidence records and the content of the current and next main thread segment. These results specify the reason for unapproval, the associated main thread position, and the subsequent acceptance status. Compliance verification results are generated based on the process data, testing data, and flow data in the corresponding thread expansion units for raw material entry, cyanide-free copper plating, resin protection, testing, warehousing, transportation, and recycling. These results list verification items and content by link. The output is the target board's lifecycle numerical thread result, written to the lifecycle result area for terminal display and subsequent querying. Anomaly or missing information is handled as follows: when a circumstantial evidence record is missing in a thread expansion unit, the responsibility positioning results and anomaly attribution results are generated solely based on the main thread segment content; when a link does not appear in the entire process record sequence, the compliance verification results include a "no record formed" marker for that link. Through the above processing, the main thread segment in the lifecycle main thread is first expanded into a full-process record sequence, and then a segment-by-segment correspondence is established with the supporting evidence records in the supporting evidence record set. Subsequently, a thread expansion unit organized according to "main thread segment, supporting evidence record group, next main thread segment" is formed. Finally, the tracing results, responsibility location results, anomaly attribution results, and compliance verification results are output respectively. In this way, the main thread and supporting evidence records are uniformly converted into directly readable lifecycle digital thread results. This processing fixes the expansion start point, expansion end point, one-to-many connection relationship of supporting evidence records, and the generation criteria of the four types of results, so that subsequent display, auditing, and responsibility verification are all based on the same expansion structure. In practical applications: For example, the main lifecycle thread of a target board includes the raw material arrival segment, the cyanide-free copper plating process segment, the etching segment, and the forming segment. The system first unfolds the entire process record sequence starting from the raw material arrival segment. Then, it attaches two supporting records corresponding to the etching segment and one supporting record corresponding to the forming segment to the corresponding main thread segment, forming multiple thread unfolding units. Subsequently, these thread unfolding units are spliced ​​together according to the segment order to output the traceability results of the target board from arrival to forming. Based on the etching segment and its supporting records, the system generates responsibility location results and anomaly attribution results. Based on the process data, inspection data, and flow data in the raw material arrival segment, the cyanide-free copper plating process segment, the etching segment, and the forming segment, the system generates compliance verification results, and finally forms the complete lifecycle digital thread result of the target board.

[0022] Furthermore, the present invention also includes a cyanide-free process circuit board lifecycle management system based on a relay chain, the system comprising a data acquisition module, a queue chain construction module, an acceptance and admission module, a thread generation module, and a thread application module: The data acquisition module is used to collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and writes the data segments of the corresponding process in the order of acquisition, and outputs the front-end segment set; The candidate thread fragment construction module is used to add the preceding fragment identifier and the current fragment identifier to each preceding fragment in the preceding fragment set using the relay chain node, form a candidate thread fragment chain according to the connection relationship between the preceding fragment identifier and the current fragment identifier, and mark the current fragment that has not obtained the subsequent link succession result as the fragment to be verified, and output the candidate thread fragment chain and the fragment set to be verified. The acceptance module is used to read each segment to be certified and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the acceptance segment, and the acceptance segment set and the non-acceptance segment set are output. The thread generation module is used to write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstance record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstance record set. The thread application module performs associated expansion based on the lifecycle main thread and the set of supporting evidence records. It outputs the entire process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieves the corresponding supporting evidence records when reading any segment of the main thread. It outputs the tracing results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital threads.

[0023] Working principle: This solution first organizes the data generated by the circuit board at each stage into segments, and then connects these segments according to the process sequence. Next, the system does not directly treat the previously collected data as the final history, but checks whether the subsequent stages are connected and whether they match. Segments that match are entered into the main thread as formal lifecycle records, while segments that do not match are retained as supporting evidence. After this processing, the system obtains a clear main thread and verifiable supporting evidence of the lifecycle digital thread, which can then be used to output the entire process record, the location of responsibility, the source of anomalies, and the compliance status. For example, a cyanide-free circuit board goes through the following steps: raw material arrival, copper plating, etching, forming, and testing. The system first records the data from each step as segments, and then checks whether the etching segments are truly inherited by the subsequent forming and testing results. If they are inherited, the etching segments are put into the main thread; if they are not inherited, they are temporarily placed in the circumstantial evidence. In this way, when you finally look at the history of this board, you see a main process that has been verified by subsequent results. If there are any abnormalities, rework, or conflict records in the middle, you can also see the corresponding circumstantial evidence at the same time. Therefore, you can trace how the entire board went through and also see where the problem occurred.

[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A lifecycle management method for cyanide-free process circuit boards based on relay chains, characterized in that, include: S1. Collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and write them into the data segments of the corresponding links in the order of collection, and output the front segment set. S2. Using relay chain nodes, add the preceding segment identifier and the current segment identifier to each preceding segment in the preceding segment set. Form a candidate thread segment chain according to the connection relationship between the preceding segment identifier and the current segment identifier. Mark the current segment that has not obtained the subsequent link result as the segment to be verified. Output the candidate thread segment chain and the segment to be verified set. S3. Read each segment to be certified, and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the admitted segment, and the admitted segment set and the non-admitted segment set are output. S4. Write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstance record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstance record set. S3 includes: S3-1. Read each segment to be certified from the segment to be certified set and each segment to be certified from the subsequent segment set. According to the fixed field order of board identifier, link identifier, process data, test data and circulation data, split each segment to be certified and each segment to be certified into a numerical field vector and a code value field vector respectively. Generate an initial segment pair set based on the following retention conditions: the same board identifier, adjacent link identifiers, and the same circulation destination code value in the segment to be certified and circulation source code value in the segment to be certified. Output the field alignment table and the initial segment pair set. S3-2. For each initial fragment pair in the initial fragment pair set, subtract the corresponding numerical field vectors item by item according to the field alignment table to form a process difference matrix. Encode the corresponding detection data item by item into consistent values, reverse values, and missing values ​​to form a detection relationship matrix. Perform head-to-tail continuation judgment on the corresponding flow code values ​​item by item to form a flow continuation matrix. Then perform singular value decomposition on the process difference matrix and extract the singular vector group corresponding to the non-zero singular values. Perform projection reconstruction on the singular vector group to obtain a stable difference vector. Output the stable difference vector, detection relationship matrix, and flow continuation matrix corresponding to each initial fragment pair. S3-3. For each initial segment pair, count the field position sequence in the stable difference vector, the number of reverse values ​​and missing values ​​in the detection relation matrix, and the number of breakpoints in the transition continuation matrix. Then, arrange the field position sequence, the number of reverse values, the number of missing values, and the number of breakpoints in order to form an edge cost sequence. Construct a bipartite connection graph with the segment to be verified and the subsequent segment as the two sides and each initial segment pair as the connecting edge. Perform lexicographical sorting under the constraint that each segment to be verified is connected to only one subsequent segment and each subsequent segment is connected to only one segment to be verified. Then, select the connecting edges corresponding to the unused segments to be verified and the unused subsequent segments in order according to the sorted edge cost sequence to form a provisional connection edge set. When there are connecting edges with the same edge cost sequence, write the corresponding connecting edges back to the candidate thread segment chain and perform dynamic programming sequence alignment on the previous segment identifier, the current segment identifier, and the subsequent segment identifier. Retain the connecting edges with fewer broken chain bits and output the stable connection edge set. S3-4. Determine the segments to be certified that have connecting edges in the stable receiving edge set as admitted segments and write them into the admitted segment set. Determine the segments to be certified that do not have connecting edges in the stable receiving edge set as unadmitted segments and write them into the unadmitted segment set. Output the admitted segment set and the unadmitted segment set.

2. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 1, characterized in that: Also includes: S5. Based on the lifecycle main thread and the set of supporting evidence records, perform associated expansion, output the full process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieve the corresponding supporting evidence records when reading any main thread segment, and output the traceability results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital threads.

3. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 2, characterized in that: S1 includes: S1-1: Read the board identifier, process identifier, process data, inspection data and flow data of the target board in the current stage, perform assembly according to the fixed field order of board identifier, process identifier, data category identifier and acquisition time identifier, and output the original record of the current stage; S1-2. Sort the data items in the original record of the current stage according to the order of collection time, and arrange the data items with the same collection time in a fixed order according to the data category identifier to form the data item sequence of the current stage, and output the current stage segment body; S1-3. Sequentially splice the current segment body with the preceding segment segments that have been collected from the target board in the order of the segments to generate the preceding segment set of the target board.

4. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 3, characterized in that: S2 includes: S2-1. Based on each preceding segment in the preceding segment set, extract the segment identifier of the previous preceding segment as the preceding segment identifier according to the adjacent order of each preceding segment in the preceding segment set, and extract the segment identifier of the current preceding segment as the current segment identifier, and output the segment connection pair set. S2-2. Perform sequential concatenation of each fragment connection pair in the fragment connection pair set according to the first-to-last correspondence between the previous fragment identifier and the current fragment identifier, generate a candidate thread fragment chain formed by sequentially connecting multiple previous fragments, and output the candidate thread fragment chain set. S2-3. Based on the chain tail segment before the chain in the candidate thread segment chain set, check whether the chain tail segment before the chain tail has formed a corresponding subsequent segment with the subsequent adjacent links one by one. Mark the chain tail segment before the chain tail that has not formed a corresponding subsequent segment as a segment to be verified, and output the candidate thread segment chain and the segment set to be verified.

5. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 4, characterized in that: S4 includes: S4-1. Read each admission fragment in the admission fragment set, along with its predecessor fragment identifier, current fragment identifier, and connection relationships. Construct an admission directed graph with admission fragments as nodes and connection relationships as directed edges. Form an adjacency matrix according to the corresponding column of the predecessor fragment identifier and the corresponding row of the current fragment identifier. Perform transitive closure expansion on the adjacency matrix to obtain the reachability matrix. If there is a return position from the same node to itself in the reachability matrix, delete the corresponding return edge and re-execute the transitive closure expansion until the adjacency matrices obtained in two consecutive rounds are consistent. Output a stable connection graph and a stable adjacency matrix. S4-2. Based on the stable connection graph, read the number of incoming edges and outgoing edges of each admitted segment. For the current segment with more than one incoming edge, perform a triple sorting according to the order of appearance of the preceding segment in the previous segment set, the order of the links, and the order of the segment identifier, and retain the incoming edge corresponding to the first position of the sort. For the preceding segment with more than one outgoing edge, perform a triple sorting according to the order of appearance of the current segment in the subsequent segment set, the order of the links, and the order of the segment identifier, and retain the outgoing edge corresponding to the first position of the sort. Then, write the retained connection edges back to the stable adjacency matrix and perform dynamic programming accumulation of the prefix path length table and the suffix path length table. When the same node corresponds to multiple prefix paths and multiple suffix paths, retain the connection edge whose sum of prefix path length and suffix path length is at the first position of the sort. Output the connection edge set of the main thread.

6. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 5, characterized in that: S4 also includes: S4-3. Concatenate the beginning and end of each admission segment according to the main thread connection edge set to form the lifecycle main thread. Calculate the prefix hash value, connection hash value, and suffix hash value for adjacent admission segments in the lifecycle main thread. The prefix hash value is generated by concatenating the segment identifier of the previous admission segment with the previous connection hash value. The connection hash value is generated by concatenating the segment identifier of the current admission segment, the identifier of the preceding segment, and the identifier of the current segment. The suffix hash value is generated by concatenating the segment identifier of the next admission segment with the connection hash value. Then, write the prefix hash value, connection hash value, and suffix hash value into the corresponding admission segment and output the lifecycle main thread. S4-4. Read each unapproved fragment in the unapproved fragment set and each approved fragment in the lifecycle main thread. Generate supporting evidence candidate pairs according to the consistency of board identifier, the adjacent or identical sequence of links, and the correspondence between the source and destination of the flow. Calculate the fragment identifier editing steps, the number of link intervals, and the number of flow breakpoints for each supporting evidence candidate pair and form a supporting evidence cost sequence in chronological order. Then, select the first approved fragment in the sorting order of the supporting evidence cost sequence as the corresponding approved fragment for each unapproved fragment. Write the fragment identifier of the unapproved fragment, the fragment identifier of the corresponding approved fragment, the supporting evidence cost sequence, and the content of the unapproved fragment into the supporting evidence record and output the supporting evidence record set.

7. The lifecycle management method for cyanide-free process circuit boards based on relay chains according to claim 6, characterized in that: S5 includes: S5-1. Based on each main thread segment in the lifecycle main thread and each circumstant record in the circumstant record set, expand each main thread segment according to the segment identifier, the preceding segment identifier and the current segment identifier in the main thread segment, and match the corresponding admission segment identifier in each circumstant record with the expanded main thread segment identifier to form a full process record sequence and a segment circumstant record correspondence table. S5-2. For each main thread segment in the whole process recording sequence, retrieve the corresponding circumstantial record synchronously according to the segment circumstantial evidence correspondence table, and perform ternary association expansion in the order of main thread segment, corresponding circumstantial record, and next main thread segment to generate thread expansion unit corresponding to each main thread segment and output thread expansion sequence. S5-3. Perform sequential splicing of each thread expansion unit in the thread expansion sequence according to the fragment order, and generate traceability results, responsibility location results, anomaly attribution results and compliance verification results according to the main thread fragment content and corresponding supporting evidence record content in the thread expansion unit, and output the life cycle digital thread result of the target board.

8. A relay-chain-based lifecycle management system for cyanide-free process circuit boards, used to implement the relay-chain-based lifecycle management method for cyanide-free process circuit boards according to any one of claims 1-7, the system comprising a data acquisition module, a queue chain construction module, an acceptance and admission module, a thread generation module, and a thread application module, characterized in that: The data acquisition module is used to collect the board identification, process identification, process data, test data and circulation data generated during the raw material entry, cyanide-free copper plating process, resin protection, etching, molding, testing, storage, transportation and recycling of the target board, and writes the data segments of the corresponding process in the order of acquisition, and outputs the front-end segment set; The candidate thread fragment construction module is used to add the preceding fragment identifier and the current fragment identifier to each preceding fragment in the preceding fragment set using the relay chain node, form a candidate thread fragment chain according to the connection relationship between the preceding fragment identifier and the current fragment identifier, and mark the current fragment that has not obtained the subsequent link succession result as the fragment to be verified, and output the candidate thread fragment chain and the fragment set to be verified. The acceptance module is used to read each segment to be certified and compare the board identifier, process identifier, process data, test data and flow data in the segment to be certified with the corresponding data in the newly formed subsequent segments of the adjacent process. When the processing result, test result or flow result formed by the segment to be certified is identified in the subsequent segment, the corresponding segment to be certified is determined as the acceptance segment, and the acceptance segment set and the non-acceptance segment set are output. The thread generation module is used to write each admitted fragment in the admitted fragment set into the main thread according to the connection relationship, and write each non-admitted fragment in the non-admitted fragment set into the circumstance record according to the fragment identifier of the corresponding admitted fragment, so as to generate the life cycle digital thread of the target board and output the life cycle main thread and the circumstance record set. The thread application module performs associated expansion based on the lifecycle main thread and the set of supporting evidence records. It outputs the entire process record of the target board in the order of the segments in the lifecycle main thread, and synchronously retrieves the corresponding supporting evidence records when reading any segment of the main thread. It outputs the tracing results, responsibility location results, anomaly attribution results and compliance verification results in the form of lifecycle digital threads.

Images