Multi-stage quality and safety dynamic traceability method and system based on big data linkage

Abstract

This invention discloses a multi-stage dynamic traceability method and system for quality and safety based on big data linkage, belonging to the field of big data analysis technology. The method includes: introducing a big data linkage engine to construct a multi-stage product traceability link library and matching the traceability link of the target product; collecting multi-stage quality-related data streams according to the link, assigning codes and storing them to generate dynamic traceability code quality data; constructing and activating a two-way traceability trigger mechanism to trace the dynamic traceability code quality data and locate quality and safety nodes; and handling and dynamically optimizing the traceability link based on node records. This solves the technical problems in existing product quality and safety traceability processes, such as insufficient multi-stage quality data correlation and fragmented traceability links, which lead to difficulties in quickly locating quality and safety anomalies and untimely handling feedback. It achieves the technical effect of improving the accuracy of product quality and safety traceability positioning and the efficiency of closed-loop management by dynamically assigning codes to multi-stage quality-related data and implementing two-way safety traceability.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, specifically to a multi-stage dynamic traceability method and system for quality and safety based on big data linkage. Background Technology

[0002] With the continuous expansion of product production, distribution, and consumption scenarios, product quality and safety management has gradually shifted from single-stage testing to comprehensive supervision covering multiple stages, including raw material procurement, production and processing, warehousing and transportation, sales and distribution, and end-user feedback. Particularly in the fields of food, pharmaceuticals, agricultural products, industrial products, and supply chain collaborative management, products generate a large amount of quality testing data, distribution record data, environmental monitoring data, batch management data, and user feedback data at different stages. This data provides a crucial foundation for product quality and safety traceability. Existing quality and safety traceability methods typically rely on single-node records, batch codes, or manual ledgers to track the product distribution process. While these methods can achieve product origin inquiry and flow tracking to a certain extent, they still suffer from problems such as scattered data sources, insufficient link connections, fragmented stage data, and inconsistent coding systems in multi-industry, multi-entity, and multi-stage collaborative scenarios. When product quality and safety anomalies occur, it is difficult to quickly determine the stage and responsible node of the anomaly based on the quality data of the entire chain. This results in low traceability efficiency, insufficient location accuracy, and difficulty in timely feedback of subsequent handling results to the traceability chain. As a result, a continuously optimized closed-loop management mechanism for quality and safety cannot be formed, leading to the recurrence of similar quality and safety problems. The intelligence level and adaptability of the traceability system are clearly insufficient. Summary of the Invention

[0003] This application provides a multi-stage dynamic traceability method and system for quality and safety based on big data linkage. Its key feature is that it solves the technical problems in the existing product quality and safety traceability process, which are difficult to quickly locate quality and safety anomalies and are not timely in handling and feedback due to insufficient correlation of multi-stage quality data and fragmented traceability links. It achieves the technical effect of improving the accuracy of product quality and safety traceability and the efficiency of closed-loop management by dynamically assigning and storing multi-stage quality correlation data and performing two-way safety traceability.

[0004] The first aspect of this application provides a multi-stage dynamic traceability method for quality and safety based on big data linkage, the method comprising: A big data linkage engine is introduced to construct a multi-stage product traceability link library. Based on this library, a target multi-stage traceability link for the target product is matched. Multi-stage quality-related data streams of the target product are collected according to the target multi-stage traceability link, and traceability codes are assigned and stored to generate dynamic product traceability code quality data. A two-way traceability triggering mechanism is constructed to activate the dynamic product traceability code quality data for bidirectional security tracing, locating product quality and safety traceability nodes. Based on these traceability nodes, traceability processing records and dynamic closed-loop optimizations are performed on the target multi-stage traceability link.

[0005] A second aspect of this application provides a multi-stage dynamic traceability system for quality and safety based on big data linkage, the system comprising: Link Matching Module: Introduces a big data linkage engine to construct a multi-stage product traceability link library. Based on this library, it matches the target multi-stage traceability link of the target product. Coding and Storage Module: Collects multi-stage quality-related data streams of the target product according to the target multi-stage traceability link, assigns traceability codes to the data streams, and stores them to generate dynamic product traceability code quality data. Security Traceability Module: Constructs a two-way traceability triggering mechanism to activate the mechanism for two-way security traceability of the dynamic product traceability code quality data, locating product quality and safety traceability nodes. Traceability Handling Module: Based on the product quality and safety traceability nodes, it records and dynamically optimizes the traceability handling of the target multi-stage traceability link.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, a big data linkage engine integrates and analyzes product quality data from different sources to establish a traceability chain model covering multiple stages of product circulation. Then, a corresponding full-process traceability path is matched based on the characteristics of the target product. Next, quality-related information at each stage of the product is collected, and data is associated and stored using a hierarchical identification method to form dynamically queryable quality traceability data. When a quality anomaly occurs, a two-way traceability mechanism is used to track and analyze from both the product flow direction and the problem feedback direction, quickly identifying key links and anomaly nodes affecting product quality. Finally, based on the location results, the traceability chain is recorded and optimized to achieve a dynamic closed loop in the product quality and safety management process, improving the ability to track quality problems and control risks. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of the multi-stage dynamic traceability method for quality and safety based on big data linkage provided in the embodiments of this application.

[0009] Figure 2 A heat map illustrating the weight distribution of node anomalies provided in an embodiment of this application.

[0010] Figure 3 This is a schematic diagram of the structure of a multi-stage dynamic traceability system for quality and safety based on big data linkage, provided in an embodiment of this application.

[0011] Explanation of reference numerals in the attached diagram: Link matching module 11, code storage module 12, security traceability module 13, source tracing and handling module 14. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown, this application provides a multi-stage dynamic traceability method for quality and safety based on big data linkage, wherein the method includes: A big data linkage engine is introduced to construct a multi-stage product traceability link library. Based on the multi-stage product traceability link library, the target multi-stage traceability link of the target product is matched.

[0014] In this embodiment, a big data linkage engine is first deployed on the regulatory cloud server, and industry product classification directories and multi-industry product association data are accessed within the engine. Subsequently, the multi-industry product association data is categorized and labeled according to the industry product classification directories, and the categorized data undergoes data cleaning and traceability stage splitting to obtain multi-stage traceability sets for different industry products. Then, multi-industry product traceability link templates are generated based on the obtained multi-stage traceability sets and integrated into a multi-stage product traceability link library. When traceability of a target product is required, the product batch number, product name, industry code, product category code, and product sub-category code of the target product are read. These fields are precisely matched with the product category codes in the link template, and the traceability link corresponding to the target product is selected from the multi-stage product traceability link library to obtain the target multi-stage traceability link for the target product. This ensures that the target product follows this target multi-stage traceability link during subsequent data collection, coding storage, and two-way traceability processes.

[0015] Furthermore, a big data linkage engine will be introduced to build a multi-stage product traceability chain library, including: A big data linkage engine is introduced to access industry product classification directories and multi-industry product association data. This multi-industry product association data includes product circulation history data, product quality traceability data, and product quality testing reference data. The multi-industry product association data is classified and labeled according to the industry product classification directories to obtain product association data for different industries. Data cleaning and traceability stage decomposition are performed on the different industry product association data to obtain multi-stage traceability sets for different industries. Based on these multi-stage traceability sets, multi-industry product traceability link templates are generated and integrated to construct a multi-stage product traceability link library.

[0016] Preferably, a big data linkage engine is deployed on the regulatory cloud server, and the following tables are pre-established within the big data linkage engine: an industry product classification directory table, a product association raw data table, a product classification identifier table, a cleaned association data table, a traceability stage rule table, a product stage data table, a link template table, and a link library index table. First, the industry product classification directory is received through a unified data access interface, and the industry code, product category code, product sub-category code, product standard name, and product stage rule number are written into the industry product classification directory table. Simultaneously, multi-industry product association data is received through the same data access interface, and the product batch number, industry code, product category code, product sub-category code, product circulation order number, circulation time, circulation entity code, upstream entity code, downstream entity code, quality traceability record number, traceability event type, quality testing items, quality testing results, testing time, and testing institution code are written into the product association raw data table. Subsequently, the big data linkage engine reads the industry code, product category code, and product sub-category code from the original product association data table and performs precise matching with the corresponding fields in the industry product classification directory table. Successfully matched data is written to the product classification identifier table, and an industry product classification identifier is generated for each data entry. This identifier is formed by sequentially combining the industry code, product category code, and product sub-category code, thus obtaining multi-industry product association data categorized by industry and product type. Next, duplicate records are deleted using the product batch number, product circulation order number, quality traceability record number, and testing time as joint deduplication fields. Circulation time and testing time are uniformly converted to standard time fields according to a preset time format. Finally, the product batch number, industry code, product sub-category code, traceability event type, and quality testing result are checked for emptiness according to field integrity rules, and the data that passes the check is written to the cleaned association data table. After data cleaning, the traceability stage rule table is read. This table pre-defines a one-to-one correspondence between traceability event types and traceability stages. Specifically, raw material receiving events correspond to the raw material receiving stage; production and processing events to the production and processing stage; quality inspection events to the quality inspection stage; warehousing and storage events to the warehousing and storage stage; logistics and transportation events to the logistics and transportation stage; sales and distribution events to the sales and distribution stage; and terminal feedback events to the terminal feedback stage. The big data linkage engine assigns each piece of data in the cleaned and associated data table to a unique traceability stage according to these correspondences. It also generates a stage node number, stage sequence number, stage responsible entity code, and a set of stage data fields for each traceability stage, writing these to the product stage data table to obtain a multi-stage set of product traceability for different industries.

[0017] Next, using product stage data under the same industry product classification identifier as the processing object, the nodes of each stage are arranged in ascending order of stage sequence number. Data connection relationships between adjacent stage nodes are established based on upstream entity code, downstream entity code, product batch number, and product circulation order number, forming a continuous node relationship from the raw material receiving stage to the terminal feedback stage. Then, the stage node number, stage sequence number, stage data field set, and adjacent node relationship corresponding to the same industry product classification identifier are combined to generate a multi-industry product traceability link template for that industry product category. The link template number, industry product classification identifier, stage node set, node sequence relationship, and node data connection field are written into the link template table. Finally, the multi-industry product traceability link templates corresponding to each industry product category are indexed according to industry code, product category code, and product subcategory code, and the index results are written into the link library index table. The link template table and the link library index table together constitute the multi-stage product traceability link library. Through the above process, scattered product circulation history data, product quality traceability data, and product quality testing reference data can be transformed into traceability link templates with fixed stage nodes and continuous node relationships according to a unified classification identifier, thereby improving the standardization of the construction of multi-stage product traceability link library and the accuracy of subsequent target product link matching.

[0018] For example, taking cold chain dairy products in the food industry as an example, an industry product traceability link template can be generated. The industry code for this template is SP, the product category code is SP-RZ, the product sub-category code is SP-RZ-LL, the industry product classification identifier is SP-SPRZ-SPRZLL, and the link template number is TPL-SP-RZ-LL-001. This template includes the following stages in sequence: raw material receiving stage, production and processing stage, quality inspection stage, warehousing and storage stage, logistics and transportation stage, sales and distribution stage, and terminal feedback stage. Specifically, the stage node number for the raw material receiving stage is N001, and the stage data fields include the raw milk batch number, supplier farm code, receiving time, receiving temperature, and initial raw material inspection results; the stage node number for the production and processing stage is N002, and the stage data fields include the production batch number, production line number, sterilization temperature, sterilization duration, filling time, and operator code; the stage node number for the quality inspection stage is N003, and the stage data fields include the test report number, test items, total bacterial count, and protein content. The data includes: test results and testing agency code; stage node number N004 for the warehousing and receiving stage, with data fields including receiving order number, warehouse code, receiving time, warehouse temperature, and storage location number; stage node number N005 for the logistics and transportation stage, with data fields including transport order number, carrier vehicle code, departure time, arrival time, transport temperature record, and carrier entity code; stage node number N006 for the sales and distribution stage, with data fields including sales order number, distributor code, store code, shelf placement time, and sales area code; and stage node number N007 for the terminal feedback stage, with data fields including feedback number, consumer feedback time, feedback type, feedback content, and problematic product batch number. In this template, N001 establishes a connection with N002 through the raw milk batch number and production batch number; N002 establishes a connection with N003 through the production batch number and test report number; N003 establishes a connection with N004 through the production batch number and warehouse entry number; N004 establishes a connection with N005 through the warehouse entry number and transport order number; N005 establishes a connection with N006 through the transport order number and sales order number; and N006 establishes a connection with N007 through the sales order number and problematic product batch number. This forms a multi-stage traceability link template for cold chain dairy products from raw material reception to end-user feedback.

[0019] According to the target multi-stage traceability link, the multi-stage quality-related data stream of the target product is collected, the multi-stage quality-related data stream is traceable and coded and stored, and dynamic traceability code quality data of the product is generated.

[0020] In one embodiment, after obtaining the target product's multi-stage traceability chain, the system reads the stage node number, stage sequence number, stage collection field, and node data receiving field from the target multi-stage traceability chain, and establishes a quality data collection task table, a stage quality data table, a traceability coding table, a coding mapping table, and a dynamic traceability code quality data table in the regulatory cloud server. The system uses the target product batch number as the collection primary key and generates quality data collection tasks sequentially according to the preset stage order in the target multi-stage traceability chain: raw material receiving stage, production and processing stage, quality inspection stage, warehousing and storage stage, logistics and transportation stage, sales and distribution stage, and terminal feedback stage. The system then writes the stage node number, collection field name, collection time requirement, and responsible entity code corresponding to each collection task into the quality data collection task table. Subsequently, according to the field requirements in the quality data collection task table, the system collects quality-related data generated by the target product at each stage. Specifically, the raw material receiving stage collects raw material batch number, supplier code, receiving time, and initial raw material inspection results; the production and processing stage collects production batch number, production equipment code, processing parameters, and production completion time; the quality inspection stage collects inspection report number, inspection items, inspection results, and inspection agency code; the warehousing and storage stage collects warehousing slip number, warehouse code, warehousing time, and warehousing environment parameters; the logistics and transportation stage collects transport slip number, vehicle code, transportation time, and transportation environment parameters; the sales and distribution stage collects sales slip number, sales entity code, sales area, and shelf-listing time; and the terminal feedback stage collects feedback number, feedback time, feedback type, and feedback content. The system writes the data collected at each stage into the stage quality data table according to the product batch number and stage node number, forming a multi-stage quality-related data stream for the target product. Subsequently, based on a pre-built hierarchical coding mechanism, the collected multi-stage quality-related data streams are traced and coded. A mapping relationship is then established between the coding results and the multi-stage quality-related data streams. Through this mapping relationship, the multi-stage quality-related data streams can be encrypted and stored, generating dynamic traceability code quality data for the product, which is then written to the dynamic traceability code quality data table. Through this process, the quality data generated by the target product at each traceability stage has a corresponding traceability identifier and link mapping relationship, improving the storage standardization, stage identifiability, and accuracy of subsequent two-way secure traceability of the multi-stage quality-related data.

[0021] Furthermore, the multi-stage quality-related data stream is traced, coded, and stored to generate dynamic product traceability code quality data, including: A hierarchical coding mechanism is constructed, and the multi-stage quality-related data stream is traced and coded using the hierarchical coding mechanism to obtain the target product traceability identifier code; a multi-stage data mapping relationship is established between the multi-stage quality-related data stream and the sub-codes of each stage in the target product traceability identifier code; the multi-stage quality-related data stream is encrypted and stored based on the multi-stage data mapping relationship to generate product dynamic traceability code quality data.

[0022] Preferably, the system establishes product root code assignment rules and product stage sub-code assignment rules based on the basic attribute information of the target product and the nodes of each stage in the target multi-stage traceability link. By merging these two rules, a hierarchical coding mechanism is constructed. This hierarchical coding rule adopts the coding structure of "product root code + stage sub-code + data sequence number + check code". Among them, the product root code is used to uniquely identify the target product; the stage sub-code is used to identify the specific traceability stage of the target product; the data sequence number is used to identify the order of quality data records within the same stage; and the check code is used to verify the integrity of the code. Subsequently, the industry code, product sub-category code, manufacturer code, and product batch number of the target product are read, and the product root code is generated in the order of "industry code - product sub-category code - manufacturer code - product batch number". Then, the stage node number and stage sequence number in the target multi-stage traceability link are read, and the stage sub-codes are generated in the order of "stage sequence number - stage node number". For example, the first stage sub-code is generated in the raw material receiving stage, the second stage sub-code is generated in the production and processing stage, the third stage sub-code is generated in the quality inspection stage, and so on until the final stage sub-code is generated in the terminal feedback stage. The system combines the product root code with the sub-codes for each stage, and generates data sequence numbers according to the order in which quality data for each stage is written into the stage quality data table. Then, it performs hash digest calculations on the product root code, stage sub-code, and data sequence number, extracting the last four digits of the hash digest as a checksum. This generates the target product traceability identifier code corresponding to each stage quality data record. After coding, using the product batch number as the primary key and the stage node number as the grouping field, each quality data record in the multi-stage quality-related data stream is bound to its corresponding stage sub-code. The system then writes the product batch number, product root code, stage sub-code, traceability identifier code, quality data record number, upstream stage sub-code, downstream stage sub-code, collection time, and responsible entity code into the quality data mapping table. This establishes a multi-stage data mapping relationship between the multi-stage quality-related data stream and the stage sub-codes in the target product traceability identifier code.

[0023] Next, the stage quality data records corresponding to the same traceability identifier are encapsulated into standard data packets, including the stage quality data body, product root code, stage sub-code, collection time, responsible entity code, and node acceptance field. Then, the data encryption key corresponding to the product root code in the key index table is called to encrypt the standard data packets. For example, AES symmetric encryption can be used. That is, the standard data packets are converted into byte sequences, and the byte sequences are grouped according to the 128-bit block length of the AES algorithm. Data that is less than one block is padded with PKCS#7. Then, the data encryption key corresponding to the product root code is read from the key index table, and the initialization vector corresponding to this encryption is generated. Then, the data encryption key is used to perform round key addition, byte substitution, row shift, column mixing, and round key iteration operations on each data block in sequence, so that the plaintext quality data in the standard data packets is converted into an unrecognizable ciphertext byte stream, thus obtaining the stage encrypted quality data. Finally, the traceability identifier, product root code, stage sub-code, upstream stage sub-code, downstream stage sub-code, encrypted quality data, encryption time, and key index number are written into the encrypted storage table to generate dynamic product traceability code quality data. Through this process, the quality data of each stage of the target product can be located via the stage sub-code, and a continuous mapping relationship is established between the traceability identifier and upstream and downstream stage data, improving the storage security, link correlation, and accuracy of subsequent traceability retrieval of multi-stage quality data.

[0024] Furthermore, constructing a layered coding mechanism includes: Obtain the basic attribute information of the target product, design coding rules based on the basic attribute information, and generate product root code coding rules; perform feature extraction and coding parsing on each stage node in the target multi-stage traceability link to obtain product stage sub-code coding rules; merge the product root code coding rules and the product stage sub-code coding rules to construct a hierarchical coding mechanism.

[0025] Optionally, using the product batch number of the target product as the primary key, the system reads the basic attribute information of the target product from the product basic attribute table, including the industry code, product category code, product sub-category code, manufacturer code, production date, and product batch number. Then, the system designs coding rules for the basic attribute information according to a preset field order, setting the industry code as the first root code field, the product sub-category code as the second root code field, the manufacturer code as the third root code field, and the product batch number as the fourth root code field. The system specifies that each field is separated by a connector, forming a product root code coding format of "industry code - product sub-category code - manufacturer code - product batch number". The system writes this product root code coding format, field order, field length, and field source into the root code rule table, generating product root code assignment rules so that the target product can be uniquely identified through its product root code. Subsequently, the system reads each stage node in the target multi-stage traceability chain, extracts features from the stage node number, stage sequence number, stage name, stage responsible entity code, and stage collection fields for each stage node, and writes the extraction results into the stage node feature table. The system encodes and parses each stage node according to the stage sequence number, setting the stage sequence number as the first sub-code field, the stage node number as the second sub-code field, and, if necessary, setting the stage responsible entity code as the third sub-code field, and the data type identifier corresponding to the stage collection field as the fourth sub-code field, forming a product stage sub-code encoding format consisting of at most "stage sequence number - stage node number - stage responsible entity code - data type identifier". The system writes the product stage sub-code encoding format, field source, field order, and stage node number corresponding to each stage node into the stage sub-code rule table to obtain the product stage sub-code assignment rules, enabling the quality data generated by the target product in different traceability stages to be distinguished through the stage sub-code. Finally, the system reads the product root code assignment rules from the root code rule table and the product stage sub-code assignment rules from the stage sub-code rule table, using the product root code as the upper-level unique identifier of the target product and the stage sub-code as the lower-level data identifier of each traceability stage of the target product, and stipulates that the complete traceability identifier code adopts a hierarchical encoding structure of "product root code - stage sub-code - data sequence number - check code". The system incorporates the aforementioned hierarchical coding structure, product root code assignment rules, product stage sub-code assignment rules, data sequence number generation rules, and check code generation rules into a hierarchical coding rule table, thereby constructing a hierarchical coding mechanism. Through this process, the basic attribute information of the target product and the stage node characteristics in the target multi-stage traceability chain can be uniformly incorporated into the coding system, enabling the generated traceability identifier to possess both unique product identification capabilities and stage data location capabilities, thus improving the accuracy of subsequent quality data coding, mapping, and traceability retrieval.

[0026] A two-way traceability triggering mechanism is constructed, which is activated to perform two-way security traceability of the product's dynamic traceability code quality data and locate the product quality and safety traceability node.

[0027] In one embodiment, firstly, the system reads the product root code, stage sub-code, stage quality data, upstream stage sub-code, downstream stage sub-code, quality inspection results, terminal feedback content, and collection time from the dynamic traceability code quality data table, and configures fixed product quality traceability trigger conditions in the traceability trigger rule table. The system scans the dynamic traceability code quality data table according to a preset cycle. When any stage quality data of the target product meets the product quality traceability trigger conditions, the corresponding product root code, stage sub-code, abnormal field, abnormal value, trigger time, and trigger reason are written into the quality abnormality record table, and a bidirectional traceability task number is generated, thereby activating the bidirectional traceability trigger mechanism. After activation, the system executes the forward quality and safety tracing set in the bidirectional traceability trigger mechanism to obtain the product's distribution nodes, and then executes the reverse quality and safety tracing to obtain the product quality and safety traceability nodes. Through the above process, forward flow tracing and reverse cause backtracking can be automatically triggered after a quality abnormality occurs, allowing product quality and safety issues to be located to specific stage nodes, improving the timeliness and accuracy of product quality and safety traceability node location.

[0028] Furthermore, activating the bidirectional traceability triggering mechanism to perform bidirectional security traceability of the product's dynamic traceability code quality data and locate the product quality and safety traceability node includes: According to the bidirectional traceability triggering mechanism, the product quality traceability triggering conditions are obtained; when the product quality traceability triggering conditions are triggered, the bidirectional traceability triggering mechanism is activated, which includes forward quality traceability and reverse quality traceability; the bidirectional traceability triggering mechanism is used to perform forward and reverse security traceability on the product dynamic traceability code quality data to locate the product quality security traceability node.

[0029] Preferably, the system first reads the product root code, stage sub-code, stage quality data, upstream stage sub-code, downstream stage sub-code, quality inspection results, collection time, and terminal feedback content from the product dynamic traceability code quality data. Based on the target product's product category, it then retrieves the corresponding product quality traceability trigger conditions from the traceability trigger rule table. These conditions include: inspection results exceeding preset quality standard thresholds; more than two abnormal feedbacks within a preset time range for the same product batch; storage or transportation environment parameters exceeding allowable ranges; missing stage quality data; and inconsistent data continuity fields between adjacent stages. The system reads the product dynamic traceability code quality data according to a preset scanning cycle and compares the quality data of each stage with the product quality traceability trigger conditions. When any stage quality data meets the trigger condition, the corresponding product root code, stage sub-code, abnormal field, abnormal value, trigger condition number, and trigger time are written to the traceability trigger record table. Simultaneously, a bidirectional traceability task number is generated, thereby activating the bidirectional traceability trigger mechanism. After the bidirectional traceability trigger mechanism is activated, the system uses the stage sub-code that triggered the abnormality as the task starting node and establishes forward and reverse quality traceability tasks in the bidirectional traceability task table. For forward quality traceability tasks, based on the downstream stage sub-codes in the product's dynamic traceability code quality data, the dynamic traceability code quality data of subsequent stages is read sequentially from the task start node to obtain the set of product-related distributed nodes corresponding to forward safety traceability. For reverse quality traceability tasks, the product terminal feedback node obtained from forward quality traceability is used as the backtracking starting point. Based on the upstream stage sub-codes in the product's dynamic traceability code quality data, the dynamic traceability code quality data of the preceding stages is read level by level to obtain the product quality and safety traceability node corresponding to reverse safety traceability. Finally, the product root code, stage sub-code, stage node number, anomaly field, anomaly cause, responsible entity code, upstream associated node, downstream affected node, and bidirectional traceability task number corresponding to the product quality and safety traceability node are written into the quality and safety node positioning table. Through the above process, forward flow tracing and reverse cause backtracking can be automatically initiated when a product quality anomaly meets the triggering conditions, and the quality and safety problem can be located to the specific traceability stage and responsible node, improving the response speed and node positioning accuracy of product quality and safety traceability.

[0030] Furthermore, the bidirectional traceability triggering mechanism is used to perform forward and reverse security traceability of the product dynamic traceability code quality data, locating the product quality and safety traceability node, including: The bidirectional traceability triggering mechanism is used to perform forward security traceability of the product dynamic traceability code quality data to obtain a set of distributed nodes involved in the product; based on the set of distributed nodes involved in the product, the product terminal feedback node is determined; and the product dynamic traceability code quality data is performed in reverse security traceability along the product terminal feedback node to locate the product quality and safety traceability node.

[0031] Optionally, when the two-way traceability trigger mechanism is activated, the system first reads the product root code, stage sub-code, and traceability identifier code corresponding to the triggered anomaly. Using the stage sub-code as the starting node for forward safety traceability, it queries the downstream stage sub-code corresponding to that stage sub-code in the product dynamic traceability code quality data. Following the pointing relationship of the downstream stage sub-codes, it reads the dynamic traceability code quality data of subsequent stages level by level, sequentially obtaining the stage node number, traceability identifier code, responsible entity code, circulation time, circulation area, node acceptance field, and quality status information corresponding to the target product in the warehousing, logistics, sales, and terminal feedback stages. The system treats each subsequent stage node as a product-related distribution node and writes the product root code, stage node number, stage sub-code, responsible entity code, node type, node location, circulation time, upstream stage sub-code, and downstream stage sub-code into the product distribution node table, thus obtaining the set of product-related distribution nodes. Subsequently, the distribution nodes with the node type of terminal feedback stage are selected from the distribution node set involved in the product, and the feedback number, feedback time, feedback subject identifier, feedback content, problematic product batch number and feedback anomaly type corresponding to the node are read. When multiple terminal feedback stage nodes exist under the same product root code, they are sorted according to the consistency between the feedback exception type and the current trigger exception field, the proximity of the feedback time and the trigger time, and the matching degree of quality exception keywords in the feedback content. That is, a feedback matching score is calculated for each terminal feedback stage node. Nodes whose feedback exception type and the current trigger exception field belong to the same exception category are assigned a first matching value, and nodes that do not belong to the same exception category are assigned a second matching value. The time difference between the feedback time and the trigger time is calculated, and a time proximity value is generated according to the rule that the smaller the time difference, the higher the time matching value. At the same time, the feedback content is matched with a preset quality exception keyword library, and a keyword matching value is generated according to the ratio of the number of matched keywords to the total number of keywords. The matching value, time proximity value and keyword matching value are then weighted and summed according to preset weights to obtain the comprehensive ranking value of each terminal feedback stage node. Multiple terminal feedback stage nodes are arranged from high to low according to the comprehensive ranking value, and the terminal feedback stage node with the highest ranking result is determined as the product terminal feedback node and written into the terminal feedback node table. After identifying the product's terminal feedback node, the system uses the stage sub-code corresponding to that node as the starting point for reverse safety backtracking. It queries the corresponding upstream stage sub-code in the product's dynamic traceability code quality data and, following the upstream stage sub-code's pointing relationship, traces back step-by-step through the sales and distribution stage, logistics and transportation stage, warehousing and storage stage, quality inspection stage, production and processing stage, and raw material receiving stage, sequentially reading the quality data of each stage to form end-to-end traceability quality data. Then, it performs quality cross-comparison and anomaly identification and location on the read end-to-end traceability quality data to determine candidate nodes with anomalies, defining them as product quality and safety traceability nodes.Finally, the product root code, stage sub-code, stage node number, anomaly field, anomaly cause, responsible entity code, upstream associated node, downstream affected node, and two-way traceability task number corresponding to the product quality and safety traceability node are written into the quality and safety node positioning table. Through the above process, the downstream distribution range and terminal feedback location of the target product can be determined first, and then the traceability can be traced back from the terminal feedback node to the specific anomaly stage, achieving precise positioning of the product quality and safety traceability node.

[0032] Furthermore, the reverse security tracing of the product dynamic traceability code quality data along the product terminal feedback node, to locate the product quality and safety traceability node, includes: Using the product terminal feedback node as the starting point for backtracking, reverse security tracing is performed on the product dynamic traceability code quality data to retrieve the full-link traceability quality data; based on the full-link traceability quality data, quality cross-comparison and anomaly identification and location are performed to determine the product quality and safety traceability node.

[0033] Optionally, after determining the product terminal feedback node, the product root code, stage sub-code, traceability identifier code, feedback number, feedback anomaly type, and feedback time corresponding to that product terminal feedback node are read, and the stage sub-code is written into the reverse backtracking task table as the backtracking starting point for reverse security traceability. Subsequently, the upstream stage sub-code corresponding to the backtracking starting point is queried in the product dynamic traceability code quality data, and the sales and distribution stage, logistics and transportation stage, warehousing and storage stage, quality inspection stage, production and processing stage, and raw material receiving stage are traced forward step by step according to the pointing relationship of the upstream stage sub-code. At each stage node, the traceability identifier code, stage node number, stage quality data, quality inspection results, environmental parameters, collection time, responsible entity code, and node acceptance field corresponding to that stage node are read, and the read data is written into the full-link traceability quality data table according to the stage sequence number, thereby forming full-link traceability quality data covering the terminal feedback node to the raw material receiving node. After retrieving the entire supply chain data, the system reads the quality standard thresholds, environmental allowable ranges, and field integrity rules corresponding to the target product category and stage node number from the stage quality standard table. It then performs a quality cross-comparison on the supply chain traceability quality data and marks stage nodes with out-of-limit quality inspection results, out-of-limit process parameters, out-of-limit environmental parameters, missing key fields, or inconsistent fields between adjacent nodes as candidate abnormal nodes, adding them to a preliminary candidate abnormal node set. Next, the system evaluates the node abnormality impact weight coefficient of these preliminary candidate abnormal nodes and identifies and locates the abnormalities according to this coefficient, determining the product quality and safety traceability node. Through this process, the system can trace back the entire supply chain quality data of the target product from the terminal feedback node, and through cross-comparison of standard thresholds, environmental parameters, process parameters, and node connection relationships, locate quality and safety issues to specific abnormal stage nodes, improving the accuracy and feasibility of reverse traceability.

[0034] Furthermore, based on the aforementioned end-to-end traceability quality data, quality cross-comparison and anomaly identification and location are performed to determine product quality and safety traceability nodes, including: By combining product quality standard thresholds, normal sample data from the same batch, and historical quality risk cases of the product, a quality cross-comparison is performed on the full-chain traceability quality data to obtain a preliminary set of candidate abnormal nodes; the set of node abnormality impact weight coefficients of the preliminary set of candidate abnormal nodes is evaluated and obtained, and anomaly identification and location are performed based on the set of node abnormality impact weight coefficients to determine the product quality and safety traceability node.

[0035] Preferably, after obtaining the end-to-end traceability quality data, the product sub-category code, product batch number, and stage node number of the target product are used as search fields. The quality standard thresholds for each stage are read from the quality standard threshold table, normal sample data that has not triggered quality anomalies under the same product batch is read from the normal sample table of the same batch, and historical quality risk cases of the same category as the target product are read from the historical risk case table. Subsequently, the end-to-end traceability quality data is cross-checked according to the stage node number. In this process, the quality inspection results, process parameters, warehousing environment parameters, and transportation environment parameters of each stage are compared with the quality standard thresholds to obtain the standard threshold deviation value; the quality data of each stage is compared with the mean of the normal sample data of the same batch to obtain the sample deviation value of the same batch; and the anomaly fields, anomaly types, and stage node numbers of each stage are matched with the risk fields, risk types, and risk occurrence stages in the historical risk cases to obtain the historical risk matching value. The system writes the standard threshold deviation, the deviation of samples from the same batch, and the historical risk matching value into a quality cross-comparison table. It then marks any stage node whose comparison result exceeds the preset anomaly judgment criteria as a preliminary anomaly candidate node, thus obtaining a set of preliminary anomaly candidate nodes. Next, the standard threshold deviation is normalized to obtain the first influence factor, the deviation of samples from the same batch is normalized to obtain the second influence factor, and the historical risk matching value is normalized to obtain the third influence factor. The ratio of the number of downstream distributed nodes corresponding to this node to the total number of nodes in the entire chain is used as the fourth influence factor. The first, second, third, and fourth influence factors are then weighted and summed according to preset weights to obtain the node anomaly influence weight coefficient for that node. The system writes the node anomaly influence weight coefficients corresponding to each preliminary anomaly candidate node into a node influence weight table, forming a set of node anomaly influence weight coefficients. Finally, the preliminary anomaly candidate node set is sorted from high to low according to the node anomaly influence weight coefficient, and the preliminary anomaly candidate node with the highest node anomaly influence weight coefficient is determined as the product quality and safety traceability node. Through the above process, the quality data of the entire chain can be cross-compared in multiple dimensions by combining standard thresholds, normal samples from the same batch, and historical risk cases. The traceability nodes most likely to cause quality and safety abnormalities can be screened according to the weight coefficient of the impact of node anomalies, thereby improving the accuracy and reliability of product quality and safety traceability node identification.

[0036] For example, the target product is a cold chain dairy product with batch number B20240618001. During the reverse traceability process, preliminary anomaly candidate nodes include node N003 in the quality inspection stage and node N005 in the logistics and transportation stage. Taking node N003 in the quality inspection stage as an example, the total bacterial count at this node is 1600 CFU / g, corresponding to a product quality standard threshold of 1000 CFU / g. Therefore, the deviation from the standard threshold is (1600-1000) / 1000=0.60. The average total bacterial count of normal samples in the same batch is 950 CFU / g, so the deviation of samples in the same batch is (1600-950) / 950=0.68. The historical risk case table retrieved similar historical risk cases of cold chain dairy products. The risk fields "total bacterial count," risk type "microbial contamination," and risk occurrence stage "quality testing stage" all matched the current node. Therefore, the historical risk matching value was set to 1.00. Furthermore, this node is located in the quality testing stage, and its downstream involves four distributed nodes: warehousing, logistics, sales, and terminal feedback. The entire chain has seven stages, so the downstream impact range value is 4 / 7 = 0.57. The system presets the standard threshold deviation value weight as 0.35, the batch sample deviation value weight as 0.25, the historical risk matching value weight as 0.25, and the downstream impact range value weight as 0.15. Therefore, the node anomaly impact weight coefficient for node N003 is 0.35 × 0.60 + 0.25 × 0.68 + 0.25 × 1.00 + 0.15 × 0.57 = 0.7155. The weighting coefficient for the abnormal impact of node N005 in the logistics and transportation stage is calculated in the same way, and it is 0.4620. Since the weighting coefficient for the abnormal impact of node N003 is higher than that of N005, the system determines node N003 in the quality inspection stage as the product quality and safety traceability node.

[0037] like Figure 2 The diagram illustrates the distribution of weighted coefficients for anomalies at various stages of a product batch, including raw material receiving, production and processing, quality inspection, warehousing, logistics and transportation, sales and distribution, and end-user feedback. The color intensity and numerical values ​​reflect the degree of influence of each stage on quality and safety anomalies; a higher weighted coefficient indicates that the stage is more likely to be a source of anomalies or have a greater impact on the subsequent spread of quality risks. This diagram allows for the intuitive identification of high-risk stages in different batches, providing data for subsequent reverse tracing and dynamic optimization of link node weights.

[0038] Based on the product quality and safety traceability nodes, the traceability processing records and dynamic closed-loop optimizations are performed on the multi-stage traceability chain of the target.

[0039] In one embodiment, after locating the product quality and safety traceability nodes, these nodes are used as the objects of handling. Their corresponding anomalies, traceability results, and handling processes are recorded, forming a complete product quality and safety traceability handling process data. This ensures that the entire process of quality and safety issues, from discovery, tracing, confirmation to handling, is fully preserved. Subsequently, based on this complete product quality and safety traceability handling process data, feedback analysis is performed on the abnormal impact of relevant nodes in the target multi-stage traceability chain. If a node exhibits high abnormal correlation or significant quality impact during the handling process, its abnormal impact weight is increased accordingly. If a node shows low abnormal correlation after handling verification, its abnormal impact weight is appropriately adjusted. This allows the target multi-stage traceability chain to be dynamically updated based on actual handling results. Through this process, the handling results of quality and safety traceability nodes can be used in reverse to influence the traceability chain, achieving dynamic closed-loop optimization of node abnormal impact weights and improving the accuracy and adaptability of subsequent quality and safety traceability judgments.

[0040] Furthermore, based on the product quality and safety traceability nodes, the multi-stage traceability chain of the target is subjected to traceability processing records and dynamic closed-loop optimization, including: Based on the product quality and safety traceability nodes, traceability and handling records are made to obtain full-process data on product quality and safety traceability and handling; dynamic closed-loop optimization is performed on the node anomaly impact weight of the target multi-stage traceability link according to the full-process data on product quality and safety traceability and handling.

[0041] Preferably, after determining the product quality and safety traceability node, the product root code, stage sub-code, stage node number, anomaly field, anomaly type, anomaly impact scope, responsible entity code, and location time corresponding to that node are read. Using the product root code and stage node number as the primary key, a traceability and handling task is generated in the traceability and handling task table. This task includes the handling task number, abnormal product batch number, abnormal stage node, abnormal responsible entity, anomaly field, anomaly type, handling requirements, task generation time, and task execution status. Subsequently, according to the handling task number, the handling personnel code, handling start time, handling actions, handling basis, handling object, number of affected products, handling process description, and handling completion time are collected during the handling execution process, and this data is written into the handling process collection table. After the handling is completed, handling result data is collected, including re-inspection results, isolation quantity, recall quantity, data correction records, responsibility confirmation results, anomaly cause confirmation conclusion, and handling completion status, and written into the handling result confirmation table. The system integrates and links data corresponding to the same disposal task number in the source tracing and disposal task table, disposal process collection table, and disposal result confirmation table. It generates full-process data for product quality and safety source tracing and disposal in the order of "anomaly location information - disposal task information - disposal execution information - disposal result information - responsibility confirmation information," and writes this data into the full-process data table. When the responsibility confirmation result indicates that the current product quality and safety source tracing node is the actual source of the anomaly, the system reads the original node anomaly impact weight coefficient of that stage node from the node anomaly impact weight coefficient set. Based on the anomaly impact range, the number of affected products, and the anomaly cause confirmation conclusion, a weight increment is generated. The original node anomaly impact weight is added to the weight increment to obtain the updated node anomaly impact weight. When the disposal result indicates that the anomaly generated by this node has been disposed of, the disposal completion status and disposal completion time are written as a closed-loop identifier into the closed-loop optimization record table. The system synchronously writes the updated node anomaly impact weight, disposal task number, weight adjustment reason, weight before adjustment, weight after adjustment, and closed-loop identifier into the node anomaly impact weight coefficient set and the closed-loop optimization record table, and writes the adjusted node anomaly impact weight back to the corresponding stage node in the target multi-stage source tracing link. Through the above process, the abnormal location results, handling execution process, and handling confirmation results of the product quality and safety traceability nodes can be transformed into writeable link optimization data. This allows the impact weight of node abnormalities in the target multi-stage traceability link to be updated with the actual handling results, thereby improving the accuracy of subsequent identification and location of similar quality and safety anomalies and the closed-loop management capability.

[0042] In summary, the embodiments of this application have at least the following technical effects: First, a big data linkage engine is introduced to construct a multi-stage product traceability link library. Based on this library, a target multi-stage traceability link for the target product is matched. Then, according to the target multi-stage traceability link, the multi-stage quality-related data stream of the target product is collected, and traceability coding is assigned and stored for this data stream, generating dynamic traceability code quality data for the product. Next, a two-way traceability triggering mechanism is constructed to activate the dynamic traceability code quality data for bidirectional security tracing, locating product quality and safety traceability nodes. Finally, based on these traceability nodes, traceability handling records and dynamic closed-loop optimization are performed on the target multi-stage traceability link. This solves the technical problems in existing product quality and safety traceability processes, such as insufficient multi-stage quality data correlation and fragmented traceability links, which lead to difficulties in quickly locating quality and safety anomalies and untimely handling feedback. It achieves the technical effect of improving the accuracy of product quality and safety traceability location and the efficiency of closed-loop management by dynamically coding and storing multi-stage quality-related data and implementing bidirectional security tracing.

[0043] Example 2, based on the same inventive concept as the multi-stage dynamic traceability method for quality and safety based on big data linkage in the previous examples, such as... Figure 3 As shown, this application provides a multi-stage dynamic traceability system for quality and safety based on big data linkage, wherein the system includes: Link Matching Module 11: Introduces a big data linkage engine to construct a multi-stage product traceability link library, and matches the target multi-stage traceability link of the target product based on the multi-stage product traceability link library; Code Assignment and Storage Module 12: Collects multi-stage quality-related data streams of the target product according to the target multi-stage traceability link, assigns traceability codes to the multi-stage quality-related data streams, and generates dynamic traceability code quality data for the product; Security Traceability Module 13: Constructs a two-way traceability triggering mechanism, activates the two-way traceability triggering mechanism to perform two-way security traceability of the dynamic traceability code quality data of the product, and locates the product quality and safety traceability node; Traceability Handling Module 14: Based on the product quality and safety traceability node, performs traceability handling records and dynamic closed-loop optimization of the target multi-stage traceability link.

[0044] Furthermore, the link matching module 11 is used to perform the following method: A big data linkage engine is introduced to access industry product classification directories and multi-industry product association data. This multi-industry product association data includes product circulation history data, product quality traceability data, and product quality testing reference data. The multi-industry product association data is classified and labeled according to the industry product classification directories to obtain product association data for different industries. Data cleaning and traceability stage decomposition are performed on the different industry product association data to obtain multi-stage traceability sets for different industries. Based on these multi-stage traceability sets, multi-industry product traceability link templates are generated and integrated to construct a multi-stage product traceability link library.

[0045] Furthermore, the encoding storage module 12 is used to perform the following method: A hierarchical coding mechanism is constructed, and the multi-stage quality-related data stream is traced and coded using the hierarchical coding mechanism to obtain the target product traceability identifier code; a multi-stage data mapping relationship is established between the multi-stage quality-related data stream and the sub-codes of each stage in the target product traceability identifier code; the multi-stage quality-related data stream is encrypted and stored based on the multi-stage data mapping relationship to generate product dynamic traceability code quality data.

[0046] Furthermore, the encoding storage module 12 is used to perform the following method: Obtain the basic attribute information of the target product, design coding rules based on the basic attribute information, and generate product root code coding rules; perform feature extraction and coding parsing on each stage node in the target multi-stage traceability link to obtain product stage sub-code coding rules; merge the product root code coding rules and the product stage sub-code coding rules to construct a hierarchical coding mechanism.

[0047] Furthermore, the security traceability module 13 is used to perform the following methods: According to the bidirectional traceability triggering mechanism, the product quality traceability triggering conditions are obtained; when the product quality traceability triggering conditions are triggered, the bidirectional traceability triggering mechanism is activated, which includes forward quality traceability and reverse quality traceability; the bidirectional traceability triggering mechanism is used to perform forward and reverse security traceability on the product dynamic traceability code quality data to locate the product quality security traceability node.

[0048] Furthermore, the security traceability module 13 is used to perform the following methods: The bidirectional traceability triggering mechanism is used to perform forward security traceability of the product dynamic traceability code quality data to obtain a set of distributed nodes involved in the product; based on the set of distributed nodes involved in the product, the product terminal feedback node is determined; and the product dynamic traceability code quality data is performed in reverse security traceability along the product terminal feedback node to locate the product quality and safety traceability node.

[0049] Furthermore, the security traceability module 13 is used to perform the following methods: Using the product terminal feedback node as the starting point for backtracking, reverse security tracing is performed on the product dynamic traceability code quality data to retrieve the full-link traceability quality data; based on the full-link traceability quality data, quality cross-comparison and anomaly identification and location are performed to determine the product quality and safety traceability node.

[0050] Furthermore, the security traceability module 13 is used to perform the following methods: By combining product quality standard thresholds, normal sample data from the same batch, and historical quality risk cases of the product, a quality cross-comparison is performed on the full-chain traceability quality data to obtain a preliminary set of candidate abnormal nodes; the set of node abnormality impact weight coefficients of the preliminary set of candidate abnormal nodes is evaluated and obtained, and anomaly identification and location are performed based on the set of node abnormality impact weight coefficients to determine the product quality and safety traceability node.

[0051] Furthermore, the source tracing and handling module 14 is used to perform the following methods: Based on the product quality and safety traceability nodes, traceability and handling records are made to obtain full-process data on product quality and safety traceability and handling; dynamic closed-loop optimization is performed on the node anomaly impact weight of the target multi-stage traceability link according to the full-process data on product quality and safety traceability and handling.

[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A multi-stage dynamic traceability method for quality and safety based on big data linkage, characterized in that: The method includes: A big data linkage engine is introduced to build a multi-stage product traceability link library. Based on the multi-stage product traceability link library, the target multi-stage traceability link of the target product is matched. According to the target multi-stage traceability link, the multi-stage quality-related data stream of the target product is collected, the multi-stage quality-related data stream is traced, coded and stored, and product dynamic traceability code quality data is generated. A two-way traceability triggering mechanism is constructed, and the two-way traceability triggering mechanism is activated to perform two-way security traceability of the product dynamic traceability code quality data and locate the product quality and safety traceability node; Based on the product quality and safety traceability nodes, the traceability processing records and dynamic closed-loop optimizations are performed on the multi-stage traceability chain of the target.

2. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 1, characterized in that, By introducing a big data linkage engine, a multi-stage product traceability link library is constructed, including: A big data linkage engine is introduced to access industry product classification catalogs and multi-industry product association data. The multi-industry product association data includes product circulation history data, product quality traceability data, and product quality testing reference data. The multi-industry product association data is classified and identified according to the industry product classification catalog to obtain product association data of different industries. Data cleaning and traceability stage decomposition are performed on the product-related data of different industries to obtain a multi-stage set of product traceability for different industries. Based on the multi-stage set of product traceability in different industries, generate multi-industry product traceability link templates, and integrate the multi-industry product traceability link templates to construct a multi-stage product traceability link library.

3. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 1, characterized in that, The multi-stage quality-related data stream is traced, coded, and stored to generate dynamic product traceability code quality data, including: A hierarchical coding mechanism is constructed, and the hierarchical coding mechanism is used to perform traceability coding on the multi-stage quality-related data stream to obtain the traceability identification code of the target product. Establish a multi-stage data mapping relationship between the multi-stage quality-related data stream and the stage sub-codes in the target product traceability identification code; Based on the multi-stage data mapping relationship, the multi-stage quality-related data stream is encrypted and stored to generate product dynamic traceability code quality data.

4. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 3, characterized in that, Constructing a layered coding mechanism includes: Obtain the basic attribute information of the target product, design coding rules based on the basic attribute information, and generate product root code assignment rules; Feature extraction and encoding parsing are performed on each stage node in the multi-stage traceability link of the target to obtain the product stage sub-code assignment rules; The product root code assignment rules and the product stage sub-code assignment rules are merged to construct a hierarchical coding mechanism.

5. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 1, characterized in that, Activating the bidirectional traceability trigger mechanism enables bidirectional security traceability of the product's dynamic traceability code quality data, locating the product quality and safety traceability node, including: Based on the bidirectional traceability triggering mechanism, the product quality traceability triggering conditions are obtained; When the product quality traceability triggering condition is triggered, the two-way traceability triggering mechanism is activated, which includes forward quality traceability and reverse quality traceability. The bidirectional traceability triggering mechanism is used to perform forward and reverse security traceability of the product dynamic traceability code quality data, and to locate the product quality and safety traceability node.

6. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 5, characterized in that, The bidirectional traceability triggering mechanism is used to perform forward and reverse security traceability of the product dynamic traceability code quality data, and to locate the product quality and safety traceability node, including: The bidirectional traceability triggering mechanism is used to perform forward security traceability of the product's dynamic traceability code quality data to obtain the set of distributed nodes involved in the product; Based on the set of distributed nodes involved in the product, determine the product terminal feedback node; The product's dynamic traceability code quality data is traced back to the product terminal feedback node to locate the product quality and safety traceability node.

7. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 6, characterized in that, Reverse security tracing of the product dynamic traceability code quality data along the product terminal feedback node, locating the product quality and safety traceability node, including: Using the product terminal feedback node as the starting point for backtracking, reverse security tracing is performed on the product dynamic traceability code quality data to retrieve the full-link traceability quality data; Based on the full-chain traceability quality data, cross-comparison of quality and anomaly identification and location are performed to determine the traceability nodes for product quality and safety.

8. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 7, characterized in that, Based on the aforementioned end-to-end traceability quality data, cross-comparison of quality data and anomaly identification and location are performed to determine product quality and safety traceability nodes, including: By combining product quality standard thresholds, normal sample data from the same batch, and historical quality risk cases of the product, a quality cross-comparison is performed on the full-chain traceability quality data to obtain a preliminary set of candidate abnormal nodes. The set of node anomaly impact weight coefficients obtained from the preliminary set of candidate anomalies is evaluated, and anomaly identification and location are performed based on the set of node anomaly impact weight coefficients to determine the traceability node for product quality and safety.

9. The multi-stage dynamic traceability method for quality and safety based on big data linkage as described in claim 1, characterized in that, Based on the product quality and safety traceability nodes, the multi-stage traceability chain of the target is traceable and its handling records are dynamically optimized, including: Based on the product quality and safety traceability nodes, traceability and handling records are made to obtain full-process data on product quality and safety traceability and handling. Based on the data from the entire product quality and safety traceability process, the impact weights of node anomalies in the multi-stage traceability chain of the target are dynamically optimized in a closed loop.

10. A multi-stage dynamic traceability system for quality and safety based on big data linkage, characterized in that: The system is used to implement the multi-stage dynamic traceability method for quality and safety based on big data linkage as described in any one of claims 1-9, the system comprising: Link matching module: Introduces a big data linkage engine to build a multi-stage product traceability link library, and matches the target multi-stage traceability link of the target product based on the multi-stage product traceability link library; The coding and storage module: according to the target multi-stage traceability link, it collects the multi-stage quality association data stream of the target product, assigns traceability codes to the multi-stage quality association data stream, and generates dynamic traceability code quality data for the product; Security traceability module: Constructs a two-way traceability triggering mechanism, activates the two-way traceability triggering mechanism to perform two-way security traceability of the product dynamic traceability code quality data, and locates the product quality and safety traceability node; Source tracing and handling module: Based on the product quality and safety source tracing nodes, the source tracing and handling module records and dynamically optimizes the multi-stage source tracing chain of the target.

Images