Production process-based material batch whole-process tracing system

Abstract

The application discloses a material batch whole-process tracing system based on production processes, and particularly relates to the field of production and manufacturing tracing, and is used for solving the continuity and accuracy of material batch whole-process tracking. The batch tracking coordinates are generated by constructing a process mapping matrix and an assembly topology index, real-time detection of flow transfer cavities and insertion of placeholder markers are performed, the chain breaking credibility is determined based on path integrity and structural complexity, the traceable chain segments are compressed, the batch flow transfer patch is generated by performing field dynamic alignment, the global batch flow transfer graph is refreshed by writing into a distributed ledger, the historical abnormal batches are located and the assembly risk prompt list is output, and the material batch whole-process continuous tracing view is established. Furthermore, the abnormal positioning time is significantly shortened, the real-time performance is improved, the affected batches are accurately locked, the recall range is reduced, the production line decision efficiency is optimized, and efficient and reliable technical support is provided for quality management and batch tracking in the production process.

Description

Technical Field

[0001] This invention relates to the field of production and manufacturing traceability, and more specifically, to a material batch full-process traceability system based on production processes. Background Technology

[0002] The production workshop operates on a batch-by-batch basis, continuously processing cutting, forming, testing, and assembly. After entering the assembly stage, materials are broken down into multiple sub-components, which then sequentially move through different processes. Each transfer relies on tags, logs, and interfaces to describe its destination. A traceability solution needs to connect all nodes to quickly pinpoint the source and limit the scope of impact in case of later quality anomalies. In practice, assembly and testing stations are often alternated, resulting in high information writing frequency and a geometrically increasing topology depth with the product structure. The mapping relationship between the data persistence layer and the real-time interface becomes complex and dynamic, thus lengthening the traceability chain.

[0003] The continuity of batch tracking in a multi-branch assembly-inspection-reassembly chain is susceptible to three types of factors: First, transfers between workstations rely on manual labeling and barcode scanning; label detachment, damage, or misapplication can leave breakpoints deep within the topology. Second, different workstations use independent log templates, resulting in a lack of consistency in field meaning and time granularity, and misalignment between the writing order and the actual flow order. Third, when rework or modification operations are performed after assembly, the original parent-child relationships must be rewritten; if the real-time interface is not synchronized in time, the traceability algorithm cannot reconstruct the complete path. These factors combined create gaps in the batch flow trajectory from raw materials to sub-components and then to the complete machine, making it difficult to quickly identify affected batches when quality issues arise, expanding the recall scope, and delaying production line decisions.

[0004] To address the aforementioned problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a material batch full-process traceability system based on production processes. This system generates batch tracking coordinates by constructing a process mapping matrix and assembly topology index, detects flow gaps in real time and inserts placeholder markers, determines the credibility of broken chains based on path integrity and structural complexity, compresses traceable chain segments, dynamically aligns execution fields to generate batch flow patches, writes to a distributed ledger to refresh the global batch flow graph, locates historical abnormal batches and outputs an assembly risk warning list, thus establishing a continuous traceability view of the entire material batch process. This significantly shortens the anomaly location time, improves real-time performance, accurately identifies affected batches, narrows the recall scope, and optimizes production line decision-making efficiency. It provides efficient and reliable technical support for quality management and batch tracking in the production process, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A material batch traceability system based on production processes includes:

[0008] Mapping Index Module: Constructs a process mapping matrix at the first transfer station and derives an assembly topology index in real time as batch tracking coordinates;

[0009] Chain break detection module: The chain break catcher monitors the tag event stream in the process mapping matrix in real time, inserts a placeholder mark after detecting a hole and records the trigger time;

[0010] Chain segment repair module: Placement markers trigger the construction of a local batch flow graph, and the credibility of the broken chain is formed by combining the path integrity measure and the structural complexity measure. Based on this, it is determined whether to compress the parent-child relationship at both ends of the gap into a traceable chain segment.

[0011] Semantic alignment module: The semantic fusion engine performs dynamic alignment of traceable chain segment fields based on the workstation context, generates batch transfer patches, and uniformly describes log differences;

[0012] Path update module: After the batch flow patch is written to the distributed ledger, the path refactorer is triggered to refresh the global batch flow graph;

[0013] Anomaly Location Module: The global batch flow diagram uses the playback module to locate historical anomaly batches and outputs an assembly risk warning list for quality decision-making.

[0014] In a preferred embodiment, the mapping index module includes the following:

[0015] At the first transfer station, a two-dimensional table-style process mapping matrix is ​​constructed for the material batch to record the flow relationship between the workstations. Initially, the elements of the process mapping matrix are zero, and updating the elements to one indicates that a flow status has occurred. A tree-structured assembly topology index is generated to record the hierarchical relationship of the material batches. In the assembly topology index, nodes represent material batches or sub-components, and connecting lines represent parent-child relationships. Initially, the assembly topology index only contains the root node. Batch tracking coordinates are generated by integrating flow path information and hierarchical structure information and using a hash function to generate a fixed-length unique identifier to ensure the uniqueness of the material batches.

[0016] In a preferred embodiment, the chain break detection module includes the following:

[0017] The technical features of the chain break catcher, which monitors the tag event stream in the process mapping matrix in real time, inserts a placeholder mark in the process mapping matrix and records the trigger time after detecting a hole, can be summarized as follows: continuously monitor the dynamic changes of the tag event stream in the process mapping matrix, and when a hole is detected due to an interruption in the tag event stream, immediately insert a placeholder mark at the corresponding position in the process mapping matrix and record the trigger time of the corresponding hole detection.

[0018] In a preferred embodiment, the chain segment repair module includes the following:

[0019] The construction of a local batch flow graph is triggered by placeholder markers. Flow events related to the workstations before and after the void are extracted from the process mapping matrix and converted into directed edges. At the same time, the missing edges at the void locations are marked.

[0020] In a preferred embodiment, the chain segment repair module further includes the following:

[0021] The chain segment coherence index is calculated based on identifying all paths from the pre-hole station to the post-hole station in the local batch flow diagram. The coherence of each path is calculated by multiplying the inverse of the time interval, and the geometric mean of the coherence of all paths is taken to quantify the integrity of the flow path in the local batch flow diagram.

[0022] The topological chaos index is calculated based on the ratio of the inverse sum of the shortest path lengths between all pairs of nodes in the local batch flow graph to the sum of the node degrees, and is used to quantify the structural complexity of the local batch flow graph.

[0023] In a preferred embodiment, the chain segment repair module further includes the following:

[0024] The chain break credibility is generated by combining the chain segment coherence index and the topological chaos index through a logical function. When the chain break credibility is higher than the preset threshold, topological compression is performed to merge the workstation before the hole, the workstation after the hole, and the intermediate workstation into virtual nodes to generate a traceable chain segment.

[0025] In a preferred embodiment, the semantic alignment module includes the following:

[0026] The semantic fusion engine loads workstation log templates, field definitions, and time granularity standards from a pre-configured workstation information database and stores them as an internal mapping table. For flow events in a traceable chain, it extracts the workstation set and uses the predefined field semantic mapping table to convert the workstation fields into a standard field set, generating aligned flow events. It performs time granularity unification operations on the aligned flow events, truncating or interpolating them according to the standard time granularity. It organizes the aligned flow events and the unified timestamps into a batch flow patch in JSON format. It records the details of the field alignment and time adjustment operations, generates a difference report, and adds it to the batch flow patch metadata field.

[0027] In a preferred embodiment, the path update module includes the following:

[0028] Batch flow patches are serialized into data packets in JSON format and sent to the distributed ledger write nodes. After verifying the integrity and digital signature of the data packets, the write nodes record the data packets in the distributed ledger. The distributed ledger uses blockchain technology to store the hash value and timestamp of the data packets. The path refactorer subscribes to the write events of the distributed ledger, receives and parses the batch identifier, workstation pair and alignment flow event set in the newly written batch flow patches. The path refactorer updates the subgraph part of the corresponding batch in the global batch flow graph according to the parsed alignment flow event set, and refreshes the global batch flow graph by adding new edges or updating edge attributes. The path refactorer records the update time and the hash value of the batch flow patches to generate a version log, and provides a real-time query interface to access the latest global batch flow graph data.

[0029] In a preferred embodiment, the anomaly localization module includes the following:

[0030] The replay module loads the latest snapshot of the global batch flow graph from the distributed ledger and obtains the predefined quality anomaly criteria. It then uses a depth-first search algorithm to traverse the flow path of the global batch flow graph in reverse, records the batches that trigger the quality anomaly criteria to the abnormal batch list, and then uses Dijkstra's algorithm to calculate the shortest path length from the abnormal batch to the downstream batch. Finally, it uses a path dependency decay model to calculate the risk propagation factor and generates an assembly risk warning list.

[0031] In a preferred embodiment, the anomaly localization module further includes the following:

[0032] The assembly risk warning list is in tabular form and includes batch identifier, risk value, anomaly type and recommended measures. Finally, the playback module transmits the assembly risk warning list to the quality decision module, supporting the technical features of quality control operations.

[0033] The technical effects and advantages of this invention, which is a material batch full-process traceability system based on production processes, are as follows:

[0034] By constructing a process mapping matrix and assembly topology index to generate batch tracking coordinates, real-time detection of flow gaps and insertion of placeholder markers, determination of chain break credibility based on path integrity and structural complexity, compression of traceable chain segments, dynamic alignment of execution fields to generate batch flow patches, writing to a distributed ledger to refresh the global batch flow graph, locating historical abnormal batches and outputting an assembly risk warning list, a continuous traceability view of the entire material batch process is established. This significantly shortens the anomaly location time, improves real-time performance, accurately identifies affected batches, narrows the recall scope, optimizes production line decision-making efficiency, and provides efficient and reliable technical support for quality management and batch tracking in the production process. Attached Figure Description

[0035] Figure 1This is a schematic diagram of the material batch full-process traceability system based on the production process of the present invention.

[0036] Figure 2 This is a schematic diagram of the steps in the anomaly location module of the material batch full-process traceability system based on the production process of the present invention. Detailed Implementation

[0037] 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.

[0038] Example 1: Figure 1 This invention presents a material batch full-process traceability system based on production processes, comprising:

[0039] Mapping Index Module: Constructs a process mapping matrix at the first turn station and immediately derives an assembly topology index as batch tracking coordinates.

[0040] Chain break detection module: The chain break catcher monitors the tag event stream in the process mapping matrix in real time, inserts a placeholder mark after detecting a hole and records the trigger time.

[0041] Chain segment repair module: Placement markers trigger the construction of a local batch flow graph, which combines path integrity and structural complexity metrics to form the chain break credibility, and determines whether to compress the parent-child relationship at both ends of the gap into a traceable chain segment.

[0042] Semantic alignment module: The semantic fusion engine performs dynamic alignment of traceable chain segment fields based on the workstation context, generates batch transfer patches, and uniformly describes log differences.

[0043] Path update module: After the batch flow patch is written to the distributed ledger, the path refactorer is triggered to refresh the global batch flow graph, reflecting the assembly relationship in real time.

[0044] Anomaly Location Module: The global batch flow diagram uses the playback module to locate historical anomaly batches and outputs an assembly risk warning list for quality decision-making.

[0045] In modern manufacturing, full-process traceability of material batches is of great significance for ensuring product quality and optimizing production management. Production workshops advance continuous processes such as cutting, forming, testing, and assembly in batches. After entering the assembly stage, materials are broken down into multiple sub-components, which flow through different processes in sequence, forming a complex flow path.

[0046] Traditional traceability methods rely on manual labeling, barcode scanning, and log recording, but these methods are prone to disruptions due to factors such as label detachment, inconsistent log formats, or delayed synchronization of rework operations. When quality anomalies occur, traditional traceability methods struggle to quickly locate affected batches, leading to expanded recalls and production line decision-making delays.

[0047] Therefore, there is an urgent need for a material batch full-process traceability method based on production processes, which can improve the efficiency and accuracy of locating abnormal batches by building a continuous and reliable traceability system. The material batch full-process traceability system based on production processes proposed in this invention is specifically designed for the above scenario. By constructing a process mapping matrix and assembly topology index at the first transfer station, it lays the foundation for subsequent traceability and ensures the traceability of batch flow.

[0048] The goal of the mapping index module is to establish an initial data structure for material batch flow tracking at the first transfer station (the first station where a material batch enters the production process). Specifically, this includes building a process mapping matrix, deriving an assembly topology index, and generating batch tracking coordinates to record the flow relationship and hierarchical structure of material batches in the production process, providing a data foundation for subsequent steps.

[0049] The mapping index module includes the following:

[0050] S1.1, Construction of the process mapping matrix:

[0051] At the first transfer station in the production process, a process mapping matrix is ​​constructed for each material batch to record the flow relationship between the material batches at each station.

[0052] The process mapping matrix is ​​a two-dimensional table where rows and columns correspond to all workstations in the production process. For example, if the production process includes workstations A, B, and C, then the rows and columns of the matrix are named workstations A, B, and C, respectively. Each element in the matrix represents the flow status from the workstation corresponding to the row to the workstation corresponding to the column. Initially, all elements in the matrix are set to zero, indicating that the material batch has not yet flowed between any workstations. When a material batch flows from one workstation to another, the system updates the element at the corresponding position in the matrix to 1, indicating that the flow has occurred. For example, when a material batch flows from workstation A to workstation B, the element at the intersection of row A and column B in the matrix changes from zero to 1. By continuously updating the matrix elements, the system can completely record the flow path of material batches in the production process.

[0053] The process mapping matrix records the flow relationships in a two-dimensional table format. Its structure is intuitive, easy to implement, and easy to understand. Through the correspondence between rows and columns, the system can quickly locate the flow status between any workstations, facilitating real-time querying and analysis. Updating matrix elements only involves changing zeros to 1s, resulting in low computational complexity and making it suitable for efficient execution during production.

[0054] S1.2, Derivation of the topological index:

[0055] While constructing the process mapping matrix, an assembly topology index is derived for each material batch to represent the hierarchical relationship and structural changes of the material batch during the assembly process.

[0056] The assembly topology index uses a tree structure for recording, where each node represents a material batch or its sub-components, and the connecting lines between nodes indicate the parent-child relationship between them. At the first transfer station, the assembly topology index tree structure contains only a root node, which corresponds to the initial material batch. As the production process progresses, if a material batch is split into multiple sub-components in subsequent processes, the system adds a new node to the tree structure for each sub-component and connects it to the parent node representing the original material batch via connecting lines.

[0057] For example, if a material batch is split into sub-component X and sub-component Y in a certain process, two new child nodes are added under the root node in the tree structure, corresponding to sub-component X and sub-component Y respectively, and connected to the root node through connecting lines. By continuously expanding the tree structure, the system can accurately record the hierarchical changes of the material batch during the assembly process.

[0058] The assembly topology index uses a tree structure to record hierarchical relationships, clearly reflecting the structural characteristics of material batches during the splitting or assembly process. The intuitive design of the tree structure's nodes and connections makes it suitable for representing complex parent-child relationships and offers good scalability to adapt to the complexity requirements of different assembly processes. By maintaining the tree structure, the system can track the association between material batches and their sub-components, providing structured data support for analyzing the integrity of the assembly process or locating anomalies. Furthermore, the tree structure facilitates the extraction of hierarchical information through traversal operations, enhancing the flexibility of data processing.

[0059] S1.3, Generation of batch tracking coordinates:

[0060] To uniquely identify and locate each material batch in the production process, the system generates a batch tracking coordinate system. The generation process of batch tracking coordinates consists of three stages.

[0061] First, extract the flow path information of the material batch from the process mapping matrix, that is, record all the station sequences that the material batch passes through, such as the order from station A to station B and then to station C.

[0062] Secondly, extract the hierarchical structure information of material batches from the assembly topology index, that is, the path from the root node to the current node in the tree structure, such as the connection relationship from the root node to the child component X.

[0063] The third step is to concatenate the flow path information and hierarchical structure information into a string in sequence, and then process the string using a hash function to generate a fixed-length unique identifier.

[0064] Hash functions ensure the uniqueness of output identifiers even when input information is similar by mapping input strings to fixed-length outputs. For example, even if two material batches have highly similar flow paths and hierarchical structures, the hashing property of a hash function can generate different identifiers. The generated batch tracking coordinates serve as global identifiers for the material batches throughout the entire production process.

[0065] Batch tracking coordinates integrate flow path information and hierarchical structure information, and utilize hash functions to generate unique identifiers, ensuring that each material batch has a unique identifier within the system. This method avoids identifier collisions and maintains identifier distinguishability even in large-scale production. The fixed-length identifiers generated by hash functions are easy to store and transmit, making them suitable for data association and retrieval in distributed systems. Furthermore, batch tracking coordinates compress multi-dimensional information into a single identifier, simplifying the data management process and improving the efficiency and consistency of the system in tracking and locating material batches.

[0066] The mapping index module has built a process mapping matrix and derived an assembly topology index at the first transfer station, serving as the basic coordinates for batch tracking. However, transfers between stations rely on manual labeling and barcode scanning. Label detachment, damage, or misapplication can lead to missing transfer information, creating breaks in the traceability chain. If these breaks are not detected and addressed in a timely manner, the continuity of batch tracking will be compromised, affecting the efficiency and accuracy of subsequent anomaly localization. Therefore, the chain break detection module focuses on real-time monitoring of the label event stream in the process mapping matrix, detecting breaks, and taking measures to ensure the integrity of the traceability chain.

[0067] The goal of the broken chain detection module is to detect gaps in the flow path by monitoring the tag event stream in the process mapping matrix in real time, and to insert a placeholder mark and record the trigger time after detecting a gap, so as to ensure the continuity of the material batch traceability link.

[0068] The chain break detection module includes the following:

[0069] S2.1, Real-time monitoring mechanism:

[0070] The broken chain catcher is a real-time monitoring module responsible for subscribing to and parsing the tagged event stream in the process mapping matrix. The tagged event stream refers to the sequence of records generated by scanning equipment as material batches flow between production stations. Each record contains the material batch number, starting station, target station, and the time the record was generated. The broken chain catcher subscribes to update events in the process mapping matrix to acquire and parse these records in real time to check the integrity of the material batch flow path. The parsing process involves reading the record content one by one to confirm the logical continuity of each record with the previous record in terms of time and station.

[0071] S2.2, Void Detection:

[0072] A void refers to a missing link in the material batch flow path, resulting in a discontinuity in the traceability chain.

[0073] The broken-chain catcher identifies gaps by analyzing the continuity of the tagged event flow. Specifically, it constructs a flow path graph for each material batch, using workstations as nodes and flow records as edges connecting them. Starting from the first workstation the material batch enters, the broken-chain catcher checks each record sequentially, verifying the existence of a direct connection from the current workstation to the next. If a record's starting workstation doesn't match the target workstation of the previous record, or if a record is missing a connection from one workstation to another, a gap is identified in the flow path, and the specific location of the gap—the missing workstation pair—is recorded.

[0074] By constructing a workflow path diagram and verifying the connectivity of each record, the system can clearly identify breakpoints in the workflow path. This method utilizes the chronological order of records and workstation relationships to intuitively and accurately locate gaps. Using a workflow path diagram facilitates automated analysis of workflow records by the system.

[0075] S2.3, Insert a placeholder marker and record the trigger time:

[0076] Upon detecting a void, the chain break catcher inserts a placeholder marker into the process mapping matrix to identify the missing flow record. The placeholder marker contains the following information: the material batch number, the starting and target workstations of the missing flow record, and the specific time the void was detected and the marker was inserted. The insertion process involves writing the placeholder marker into the corresponding position of the missing flow record in the process mapping matrix, while simultaneously recording the time the void was detected as the trigger moment and storing it in the system for subsequent analysis and repair.

[0077] The purpose of inserting placeholder markers is to maintain the continuity of the flow path by temporarily identifying missing flow records. This method ensures that even if data is missing, the system can still track the location of material batches, preventing a complete interruption of the traceability chain.

[0078] In the production process, the mapping index module has completed the construction of the process mapping matrix, which is used to record the flow relationship of material batches between workstations. The chain break detection module directly uses the data in the process mapping matrix to monitor the tag event stream in real time using a chain break catcher, detects gaps in the flow path, inserts a placeholder mark after detecting a gap, and records the trigger time. The chain segment repair module will use these placeholder marks and trigger time information to construct a local batch flow diagram and perform chain break repair, thereby ensuring the continuity of material batch tracking and the reliability of data.

[0079] The chain break detection module achieves continuous monitoring of the material batch flow path through a real-time monitoring mechanism, void detection method, and recording of placeholder marker insertion and trigger times. The real-time monitoring of the chain break catcher ensures the system can quickly respond to changes in the flow records; the construction and verification of the flow path map provides accurate void location; and the recording of placeholder marker insertion and trigger times maintains the integrity of the traceability link and provides necessary information for subsequent analysis. This systematic processing logic significantly improves the reliability and efficiency of material batch tracking during the production process.

[0080] The chain break detection module uses a chain break catcher to monitor the tag event stream in real time. Upon detecting an information gap, it inserts a placeholder marker and records the trigger time, providing a clear identifier for the chain break location in subsequent processing. However, the existence of information gaps threatens the continuity of the batch flow path, especially in multi-branch assembly-inspection-reassembly links. Factors such as tag detachment, inconsistent log formats, or unsynchronized rework operations can lead to interruptions in the traceability link. The chain segment repair module aims to construct a local batch flow graph using the placeholder markers inserted in the chain break detection module, analyze the integrity and structural complexity of the flow path, generate chain break confidence, and perform topology compression based on the judgment results to optimize the flow path structure.

[0081] The chain segment repair module includes the following:

[0082] S3.1, Construct a local batch flow diagram:

[0083] The construction of a local batch flow diagram is triggered by placeholder markers inserted in the chain break detection module. The local batch flow diagram uses workstations as nodes and material batch flow events between workstations as edges connecting the nodes, focusing on the empty locations indicated by the placeholder markers and their adjacent workstations. The construction process includes the following steps:

[0084] First, extract the flow events related to the workstations before and after the void from the process mapping matrix; second, convert the extracted flow events into directed edges to generate a local batch flow graph; finally, mark the missing edges at the void locations to clearly identify the broken chain locations.

[0085] S3.2, Calculate the chain coherence index:

[0086] The chain segment coherence index is used to quantify the completeness of the flow path in a local batch flow diagram, reflecting the path continuity from the station before the hole to the station after the hole. The calculation process includes the following steps:

[0087] First, identify all possible paths from the pre-hole station to the post-hole station in the local batch flow diagram; second, for each path, calculate its coherence based on the time interval of the flow events, where coherence is defined as the product of the reciprocals of the time intervals between adjacent stations on the path; finally, the chain segment coherence index is the geometric mean of the coherence of all paths, and if no path exists, the chain segment coherence index is zero.

[0088] The chain segment coherence index, calculated by multiplying the inverses of time intervals, emphasizes the temporal compactness of each link in the path; a higher value indicates a more continuous path. The application of the geometric mean balances the coherence of multiple paths, avoiding the influence of deviations from a single path and ensuring the comprehensiveness of the assessment. This calculation method intuitively reflects the integrity of the flow path, providing crucial quantitative evidence for generating chain break credibility.

[0089] S3.3, Calculate the topological chaos index:

[0090] The topological chaos index is used to quantify the structural complexity of a local batch flow graph, reflecting the degree of disorder in the connections between nodes. The calculation process includes the following steps:

[0091] First, calculate the reciprocal sum of the shortest path lengths between all pairs of nodes in the local batch flow graph; second, calculate the sum of the degrees of all nodes, where the degree of a node is defined as the number of edges directly connected to that node; finally, the topological chaos index is the ratio of the reciprocal sum of the shortest path lengths to the sum of the node degrees.

[0092] The topological chaos index measures the compactness of a graph by the inverse of the shortest path length and the connectivity density by the sum of node degrees; the ratio reflects the structural complexity of the graph. A higher value indicates a more complex structure and a greater likelihood of broken links. This quantification method integrates global and local characteristics of the graph, providing a structural-level assessment basis for generating the confidence level of broken links.

[0093] S3.4, Generate broken link credibility:

[0094] The chain break confidence assessment uses a chain break confidence model that integrates the chain segment coherence index and the topological chaos index to evaluate the authenticity of a chain break at a hole location. The chain break confidence model employs a logistic function, taking the chain segment coherence index and the topological chaos index as inputs, and calculates the chain break confidence using pre-trained parameters. The output range is limited to zero to one, with the value closer to one indicating a higher probability of a chain break.

[0095] For example, the following methods can be used for calculation and processing:

[0096] Broken chain confidence model:

[0097] Using logical functions:

[0098] Where β0, β1, and β2 are pre-training parameters, representing the baseline value, the influence coefficient of SCI, and the influence coefficient of TCI, respectively.

[0099] Parameter explanation:

[0100] SCI: Segment Coherence Index, the lower the value, the less complete the path;

[0101] TCI: Topological Chaos Index, a higher value indicates a more complex structure;

[0102] β0: Model baseline offset;

[0103] β1, β2: Coefficients obtained by training with historical broken link data, which adjust the influence of SCI and TCI.

[0104] The broken chain confidence model transforms the chain coherence index and topological chaos index into broken chain confidence through a logic function, thereby enabling intelligent determination of the authenticity of broken chains.

[0105] S3.5, Determine whether to perform topology compression:

[0106] The system determines whether to perform topology compression on the flow path at the void location based on a comparison between the chain break reliability and a preset threshold. The specific process is as follows:

[0107] If the chain break confidence level is higher than a preset threshold, the hole location is determined to be a real chain break, and topology compression is performed; if the chain break confidence level is lower than or equal to the preset threshold, the current path structure is retained. The topology compression method merges the workstations before and after the hole, as well as the intermediate workstations, into a virtual node, generating a traceable chain segment, and simultaneously recording the parent-child relationship before and after compression.

[0108] By comparing preset thresholds, the system can intelligently choose whether to perform topology compression, avoiding unnecessary computational operations and improving resource utilization efficiency. Topology compression simplifies the flow path structure by merging intermediate workstations.

[0109] The chain segment repair module achieves intelligent analysis and path optimization of chain break locations by constructing a local batch flow graph, calculating the chain segment coherence index and topological chaos index, generating chain break credibility, and performing topology compression. This improves the reliability and efficiency of material batch tracking, optimizes the flow path structure, and provides strong support for the overall system performance.

[0110] The chain segment repair module constructs a local batch flow graph and performs a chain break credibility determination to generate a traceable chain segment. However, different workstations use independent log templates with inconsistent field meanings and time granularities, resulting in semantic differences in the traceable chain segments, directly affecting the accuracy of batch flow information tracing. The semantic alignment module addresses this issue by using a semantic fusion engine to dynamically align the fields of the traceable chain segments, generating batch flow patches and unifying log differences, providing a consistent semantic view for the path update module to reconstruct the global batch flow graph.

[0111] The semantic alignment module includes the following:

[0112] S4.1, Initialization of the semantic fusion engine:

[0113] The semantic fusion engine, as the core processing module of the semantic alignment module, is responsible for semantically aligning fields in the traceable chain. During the initialization phase, the semantic fusion engine loads the context information of each workstation from a pre-configured workstation information database. This context information includes each workstation's log template, field definitions, and time granularity standard. The log template consists of a set of fields with clear semantic meanings, such as material number and processing status, and is associated with specific data types, such as strings or integers. The time granularity standard defines the precision of time events recorded by each workstation, for example, in seconds or milliseconds. After loading, the semantic fusion engine stores this context information as an internal mapping table. The internal mapping table organizes data in key-value pairs, where the key is the workstation identifier, and the value is the corresponding log template, field definition, and time granularity standard, enabling fast querying and comparison in subsequent processing.

[0114] S4.2, Dynamic field alignment:

[0115] The semantic fusion engine performs dynamic field alignment on the traceable segments generated by the segment repair module to ensure semantic consistency of fields across different workstations. A traceable segment contains batch flow information between production workstations, consisting of a set of flow events, each recording several field values, such as material number or processing status. The dynamic field alignment process involves three steps. First, all workstations involved in the traceable segment are identified, forming a workstation set. Second, for each workstation, the corresponding log template and field set are extracted from the internal mapping table, clarifying the semantic meaning and data type of each field. Finally, based on a predefined field semantic mapping table, the workstation-specific fields are converted into a unified set of standard fields. The field semantic mapping table is a pre-built dictionary that records the correspondence between workstation fields and standard fields; for example, converting the "Material ID" field of a workstation into the standard field "Material_ID". Through this conversion, the semantic fusion engine adjusts each flow event in the traceable segment to the aligned flow event, generating an aligned traceable segment.

[0116] Dynamic field alignment achieves field standardization through a predefined semantic mapping table, eliminating differences in field meanings between different workstations and ensuring semantic consistency of workflow events.

[0117] S4.3, Unified time granularity:

[0118] The semantic fusion engine performs a time granularity unification operation on the timestamps in the aligned backtrackable chain segment to ensure that the time records of all flow events use consistent precision. The time granularity unification process includes the following steps: First, determine the standard time granularity used by the system. Second, for each flow event in the aligned backtrackable chain segment, check whether its timestamp granularity is consistent with the standard time granularity. If an inconsistency is found, adjustments are made according to the granularity difference.

[0119] Specifically, if the timestamp granularity of a flow event is finer than the standard granularity, such as in milliseconds, then the millisecond portion is removed through a truncation operation, converting the timestamp to seconds. If the timestamp granularity of a flow event is coarser than the standard granularity, such as in minutes, then an intermediate timestamp conforming to the standard granularity is generated using linear interpolation. The adjusted timestamp is then updated in the corresponding flow event, generating a traceable chain segment with unified time granularity.

[0120] Uniform time granularity eliminates the accuracy differences in time records between different workstations through truncation and linear interpolation methods, ensuring the consistency of timestamps for all flow events.

[0121] S4.4, Generate batch transfer patch:

[0122] The semantic fusion engine generates batch flow patches based on the traceable chain segment with unified time granularity to correct and unify log differences in the flow path. Each batch flow patch contains an aligned set of flow events and timestamps with unified time granularity, stored in JSON format. This includes batch number, workstation pair, aligned list of flow events, and unified timestamps. The generation process involves the semantic fusion engine organizing the flow events and timestamps from the traceable chain segment with unified time granularity into batch flow patches, ensuring that the recorded fields and timestamps conform to standardized requirements. The generated batch flow patches serve as input for subsequent steps to correct global flow paths.

[0123] Batch workflow patching provides standardized workflow path correction data by integrating aligned workflow events and unified timestamps. This approach facilitates rapid application of patches to correct log differences in the workflow path, ensuring the accuracy and consistency of the global batch workflow graph.

[0124] S4.5, unifying the description of log differences:

[0125] The semantic fusion engine uses batch flow patches to uniformly describe log differences in traceable segments, ensuring a consistent semantic view for subsequent processing. The uniform description process includes the following steps: First, comparing the original flow events with the aligned flow events analyzes the specific operations of field mapping and time adjustment. Second, recording the mapping rules used during field alignment and the truncation or interpolation methods employed during time granularity unification, generating a difference report. This difference report contains detailed descriptions of the field alignment rules and time adjustment methods and is added to the metadata fields of the batch flow patch. Finally, the batch flow patch includes not only the aligned data but also a complete description of the alignment and adjustment process.

[0126] The semantic alignment module, through the semantic fusion engine, performs dynamic field alignment and time granularity unification operations on the traceable chain segments, successfully generating batch flow patches and eliminating differences in field meaning and time granularity between log templates of different workstations. The initialization phase ensures accurate loading of workstation context information; dynamic field alignment achieves data standardization through semantic mapping; time granularity unification adjusts timestamps through truncation and interpolation; and the generation of batch flow patches and difference descriptions provides standardized data support for the reconstruction of the global batch flow graph, guaranteeing semantic consistency and traceability accuracy of the flow path.

[0127] The semantic alignment module generates batch flow patches through the semantic fusion engine, which unifies the data fields of different workstations and aligns them with the time granularity. The path update module uses these batch flow patches to write them into the distributed ledger and triggers the path refactorer to refresh the global batch flow graph, realizing real-time updates of flow relationships. The anomaly localization module performs anomaly localization and quality decision analysis based on the updated global batch flow graph.

[0128] The path update module includes the following:

[0129] S5.1, batch transfer of patches to the distributed ledger:

[0130] In a material batch end-to-end traceability system based on production processes, the batch flow patch generated by the semantic alignment module is first written to the distributed ledger. The batch flow patch is standardized data containing aligned flow events and a unified timestamp. The distributed ledger is a decentralized data storage system used to ensure the immutability and traceability of the data. The writing process includes the following:

[0131] First, the batch transfer patches are converted into JSON format data packets. Then, this data packet is transmitted to the write nodes of the distributed ledger. Upon receiving the data packet, the write nodes verify its integrity and the validity of the digital signature. After successful verification, the write nodes record the data packet in the distributed ledger. The distributed ledger uses blockchain technology. Each block stores the hash values ​​and timestamps of multiple batch transfer patches, and the blocks are interconnected through a hash chain to ensure data continuity and security.

[0132] The process of writing batch-based patches into the distributed ledger ensures data security and traceability. The decentralized nature of the distributed ledger avoids the risk of data being tampered with by a single node, while the application of hash chains and digital signatures guarantees data integrity and authenticity.

[0133] S5.2, Trigger the path refactorer:

[0134] The path refactorer is the module in the system responsible for updating the global batch flow graph in real time. Its triggering process depends on write operations in the distributed ledger. When a batch flow patch is successfully written to the distributed ledger, the distributed ledger generates an event notification. The path refactorer receives the newly written batch flow patch data by subscribing to the distributed ledger's write event and uses this as the basis to initiate subsequent update operations.

[0135] By subscribing to write events of the distributed ledger, the path refactorer can obtain the latest batch flow patches in a timely manner, ensuring that the updates to material batch flow relationships are highly timely. Compared to the polling method of periodically checking the ledger, this event-driven mechanism reduces unnecessary resource consumption and processing latency, improves system response speed and operating efficiency, and provides technical support for real-time management.

[0136] S5.3, refresh the global batch flow graph:

[0137] The global batch flow diagram is a directed graph structure where nodes represent workstations in the production process and edges represent flow events between batches, reflecting the flow relationships of all material batches. The process of refreshing the global batch flow diagram includes the following:

[0138] First, based on the information in the batch flow patch, locate the subgraph portion in the global batch flow graph related to that batch. Then, for each alignment event in the batch flow patch, check if a corresponding edge already exists in the global batch flow graph. If no corresponding edge exists, add a new edge to the global batch flow graph. If a corresponding edge already exists, update the edge's attributes, such as modifying the timestamp or supplementing the event description. After processing all alignment events, ensure that the updated subgraph accurately reflects the latest flow relationship recorded in the batch flow patch.

[0139] The method of refreshing the global batch flow diagram enables the system to update the flow relationships of material batches in a timely manner, ensuring the integrity and accuracy of the traceability chain. By incrementally updating the sub-graph, the system avoids the computational overhead of full graph reconstruction, while supporting rapid query and analysis needs.

[0140] S5.4, reflects assembly relationships in real time:

[0141] After refreshing the global batch flow chart, the assembly relationships of material batches are updated in real time. After completing the global batch flow chart refresh, the path refactorer records the update time and the hash value of the batch flow patch, generating a version log to support historical data backtracking. Simultaneously, the system provides a real-time query interface, allowing users and the quality control module to access the latest global batch flow chart data to meet real-time monitoring needs during production.

[0142] The real-time processing of assembly relationships ensures the timeliness and accuracy of material batch flow during production. Version log generation facilitates tracking of historical updates, supports data auditing and anomaly backtracking, and enhances the system's traceability. The real-time query interface provides users and the quality control module with convenient data access, improving the system's interactivity and application value.

[0143] The path update module writes batch flow patches to the distributed ledger and refreshes the global batch flow graph, forming a complete view reflecting real-time assembly relationships. However, during production, quality anomalies may occur in historical batches (such as exceeding processing time limits or inspection results deviating from the normal range). A backtracking mechanism is needed to accurately locate these abnormal batches and assess their impact on downstream assembly to support quality decisions. The anomaly location module, based on the global batch flow graph generated by the path update module, uses a backtracking module to locate historical abnormal batches and output an assembly risk warning list, providing precise data for production line quality control.

[0144] The process of the anomaly location module is as follows: Figure 2 As shown, it includes the following:

[0145] S6.1 Initialization of the replay module:

[0146] First, the replay module loads the latest version of the global batch flow chart during initialization and retrieves predefined quality anomaly criteria. The global batch flow chart records the flow path and related attributes of material batches in the production process, and the latest version is updated by the path update module. Quality anomaly criteria are represented by a set of anomaly patterns, each containing anomaly type and a corresponding threshold, such as processing time exceeding the normal range or detection results deviating from preset standards. The replay module reads the latest snapshot of the global batch flow chart from the distributed ledger, ensuring that the analysis is based on real-time updated data. After initialization, the replay module has the ability to traverse and analyze the global batch flow chart.

[0147] The replay module loads the latest version of the global batch flow chart and predefined quality anomaly standards, ensuring that anomaly localization is based on the latest production data, thereby guaranteeing the accuracy and timeliness of the analysis. Reading snapshot data from the distributed ledger maintains data integrity and consistency, avoiding data omissions or errors, and providing reliable support for subsequent anomaly detection and risk assessment.

[0148] S6.2, Locating Historical Abnormal Batches:

[0149] The replay module identifies batches that trigger quality anomaly criteria by reverse-traversing the flow paths in the global batch flow graph. The specific process is as follows:

[0150] Starting from the end station of the global batch flow chart, the flow events of each batch at each station are checked in reverse to determine whether their attributes meet any criterion in the abnormal pattern set. For example, for processing time abnormalities, if the processing time of a certain flow event exceeds the preset maximum allowable time, the corresponding batch is marked as an abnormal batch; for detection result abnormalities, if the detection value of a certain flow event exceeds the preset normal range, the corresponding batch is also marked as an abnormal batch. All batches that trigger the abnormal criteria are recorded in the abnormal batch list.

[0151] By traversing the global batch flow diagram in reverse, from the end of the production process back to the source, the system can comprehensively and accurately identify batches that cause quality anomalies. Through item-by-item comparison with predefined anomaly criteria, the system achieves automatic anomaly detection, avoiding the inefficiency and subjective bias of manual judgment.

[0152] S6.3, Generate an assembly risk warning list:

[0153] Based on the list of abnormal batches, the replay module analyzes the impact range of abnormal batches in the global batch flow diagram and generates an assembly risk warning list. The specific process is as follows:

[0154] For each batch in the abnormal batch list, its downstream batches are identified—that is, subsequent batches affected by the abnormal batch in the production process—and a risk propagation factor is calculated to quantify the degree of impact. The risk propagation factor is calculated using a path-dependent decay model, comprehensively considering the shortest path length from the abnormal batch to downstream batches in the global batch flow graph, as well as the severity of the abnormality. As the shortest path length increases, the risk value gradually decreases exponentially. Finally, all downstream batches are sorted from highest to lowest risk value, generating an assembly risk warning list. This list includes the abnormal batch identifier, the identifiers of the affected downstream batches, the risk value, and a description of the abnormality.

[0155] For example, the processing method can be as follows:

[0156] For each abnormal batch B k ∈L abnormal Identify its downstream batch B dourn (i.e., subject to B) k (affecting subsequent batches), and calculate the risk propagation factor R. k,down To quantify the degree of impact.

[0157] A path-dependent decay model is adopted:

[0158] in:

[0159] d k,down B k To B down The shortest path length in the global batch flow graph;

[0160] α is the attenuation coefficient (with a value range of 0 < α < 1, for example 0.8), which represents the attenuation of risk as the path length decreases;

[0161] I k For B k The severity of the anomaly is predefined based on the anomaly type and degree of deviation (e.g., the percentage of processing time exceeding the limit);

[0162] Labnormal This is a list of abnormal batches.

[0163] By calculating the risk propagation factor using a path-dependent attenuation model, the system can objectively quantify the impact of abnormal batches on downstream batches, reflecting the propagation and attenuation patterns of risk in the production process. This facilitates prioritizing the handling of downstream batches with higher risks and optimizing resource utilization efficiency. The generation of an assembly risk alert list supports the rapid identification of high-risk areas, reducing the further impact of quality anomalies on the production process.

[0164] S6.4, Output assembly risk warning list:

[0165] The playback module outputs the generated assembly risk warning list to the quality decision module to support anomaly handling and production line decisions. The assembly risk warning list is presented in tabular form, with each row including batch identifier, risk value, anomaly type, and recommended action, such as "isolate batch" or "re-inspect." Based on the risk values ​​and recommended actions in the list, the quality decision module performs corresponding quality control operations, such as suspending production of the relevant batch or arranging further inspection.

[0166] The anomaly localization module achieves accurate identification and risk assessment of quality anomalies through the initialization of the playback module, the localization of historical anomaly batches, and the generation and output of an assembly risk warning list. The playback module reverse-traverses the flow path based on the latest version of the global batch flow graph, accurately marking anomaly batches and quantifying the impact of anomalies on downstream batches using a path dependency decay model, ultimately generating structured risk warning information. This step follows the real-time data updates from the path update module, ensuring the accuracy of anomaly localization and the timeliness of risk assessment, providing technical support for the production workshop to quickly pinpoint the scope of impact and implement effective control when quality anomalies occur.

[0167] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0168] It should be noted that the system of the present invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting a variety of hardware environments and usage requirements.

[0169] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0170] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely to distinguish one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0171] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A material batch full-process traceability system based on production processes, characterized in that, include: Mapping Index Module: Constructs a process mapping matrix at the first transfer station and derives an assembly topology index in real time as batch tracking coordinates; Chain break detection module: The chain break catcher monitors the tag event stream in the process mapping matrix in real time, inserts a placeholder mark after detecting a hole and records the trigger time; Chain segment repair module: Placement markers trigger the construction of a local batch flow graph, and the credibility of the broken chain is formed by combining the path integrity measure and the structural complexity measure. Based on this, it is determined whether to compress the parent-child relationship at both ends of the gap into a traceable chain segment. Semantic alignment module: The semantic fusion engine performs dynamic alignment of traceable chain segment fields based on the workstation context, generates batch transfer patches, and uniformly describes log differences; Path update module: After the batch flow patch is written to the distributed ledger, the path refactorer is triggered to refresh the global batch flow graph; Anomaly Location Module: The global batch flow chart locates historical anomaly batches via the playback module and outputs an assembly risk warning list for quality decision-making; The mapping index module includes the following: At the first transfer station, a two-dimensional table-style process mapping matrix is ​​constructed for the material batch to record the flow relationship between the workstations. Initially, the process mapping matrix has 0 elements, and updating the elements to 1 indicates that a flow status has occurred. A tree-structured assembly topology index is generated to record the hierarchical relationship of the material batches. In the assembly topology index, nodes represent material batches or sub-components, and connecting lines represent parent-child relationships. Initially, the assembly topology index only contains the root node. Batch tracking coordinates are generated by integrating flow path information and hierarchical structure information and using a hash function to generate a fixed-length unique identifier to ensure the uniqueness of the material batch.

2. The material batch full-process traceability system based on production process as described in claim 1, characterized in that, The chain break detection module includes the following: The technical features of the chain break catcher, which monitors the tag event stream in the process mapping matrix in real time, inserts a placeholder mark in the process mapping matrix and records the trigger time after detecting a hole, can be summarized as follows: continuously monitor the dynamic changes of the tag event stream in the process mapping matrix, and when a hole is detected due to an interruption in the tag event stream, immediately insert a placeholder mark at the corresponding position in the process mapping matrix and record the trigger time of the corresponding hole detection.

3. The material batch full-process traceability system based on production process as described in claim 2, characterized in that, The chain segment repair module includes the following: The construction of a local batch flow graph is triggered by placeholder markers. Flow events related to the workstations before and after the void are extracted from the process mapping matrix and converted into directed edges. At the same time, the missing edges at the void locations are marked.

4. The material batch full-process traceability system based on production process as described in claim 3, characterized in that, The chain segment repair module also includes the following: The chain segment coherence index is calculated based on identifying all paths from the pre-hole station to the post-hole station in the local batch flow diagram. The coherence of each path is calculated by multiplying the inverse of the time interval, and the geometric mean of the coherence of all paths is taken to quantify the integrity of the flow path in the local batch flow diagram. The topological chaos index is calculated based on the ratio of the inverse sum of the shortest path lengths between all pairs of nodes in the local batch flow graph to the sum of the node degrees, and is used to quantify the structural complexity of the local batch flow graph.

5. The material batch full-process traceability system based on production process as described in claim 4, characterized in that, The chain segment repair module also includes the following: The chain break credibility is generated by combining the chain segment coherence index and the topological chaos index through a logical function. When the chain break credibility is higher than the preset threshold, topological compression is performed to merge the workstation before the hole, the workstation after the hole, and the intermediate workstation into virtual nodes to generate a traceable chain segment.

6. The material batch full-process traceability system based on production process as described in claim 5, characterized in that, The semantic alignment module includes the following: The semantic fusion engine loads workstation log templates, field definitions, and time granularity standards from a pre-configured workstation information database and stores them as an internal mapping table. For the flow events in the traceable chain segment, it extracts the workstation set and uses the pre-defined field semantic mapping table to convert the workstation fields into a standard field set, generating aligned flow events. Perform time-granularity uniform operations on the aligned flow events, and truncate or interpolate them according to the standard time granularity. Organize the aligned events and unified timestamps into batch flow patches in JSON format; Record details of field alignment and time adjustment operations, generate a difference report, and add it to the batch transfer patch metadata field.

7. The material batch full-process traceability system based on production process as described in claim 6, characterized in that, The path update module includes the following: The batch transfer patch is serialized into a data packet in JSON format and sent to the distributed ledger write node. After verifying the integrity of the data packet and the digital signature, the write node records the data packet in the distributed ledger. The distributed ledger uses blockchain technology to store the hash value and timestamp of the data packet. The path refactoring subscribes to write events from the distributed ledger, receives and parses the batch identifier, workstation pair, and alignment flow event set in the newly written batch flow patch. The path refactoring updates the subgraph portion of the corresponding batch in the global batch flow graph based on the parsed alignment flow event set. It refreshes the global batch flow graph by adding new edges or updating edge attributes. The path refactoring records the update time and the hash value of the batch flow patch to generate a version log and provides a real-time query interface to access the latest global batch flow graph data.

8. The material batch full-process traceability system based on production process as described in claim 7, characterized in that, The anomaly localization module includes the following: The replay module loads the latest snapshot of the global batch flow graph from the distributed ledger and obtains the predefined quality anomaly criteria. It then uses a depth-first search algorithm to traverse the flow path of the global batch flow graph in reverse, records the batches that trigger the quality anomaly criteria to the abnormal batch list, and then uses Dijkstra's algorithm to calculate the shortest path length from the abnormal batch to the downstream batch. Finally, it uses a path dependency decay model to calculate the risk propagation factor and generates an assembly risk warning list.

9. The material batch full-process traceability system based on production process as described in claim 8, characterized in that, The anomaly localization module also includes the following: The assembly risk warning list is in tabular form and includes batch identifier, risk value, anomaly type and recommended measures. Finally, the playback module transmits the assembly risk warning list to the quality decision module to support quality control operations.

Images