Food production and processing material management system

Abstract

The application discloses a kind of food production and processing material control system, it is related to food material control field, including material batch state generation module, formula constraint matching module, material requisition production generation module, feeding risk verification module, material conversion record module, material balance verification module, abnormal range positioning module.In the application, by the formula boundary discrimination mechanism based on MABAC algorithm and support vector data description algorithm fusion, the forced attribute and optimization attribute in product formula constraint data are handled in layers, and the boundary distance expression of candidate material batch is formed by available candidate domain, forbidden candidate domain, available boundary approximate area and forbidden boundary approximate area, so that material formula matching result can simultaneously reflect direct matching, review matching and forbidden matching state, solve the problem that single sorting mode is difficult to identify forbidden risk and boundary candidate, improve the identification accuracy, risk sensitivity and result explainability of material formula matching.

Description

Technical Field

[0001] This invention relates to the field of food material management, and more particularly to a food production and processing material management system. Background Technology

[0002] In the production and processing of cakes, rice crackers, and candies, material control typically revolves around raw material warehousing, inventory records, formula matching, material requisition and production scheduling, material input execution, processing records, and finished product traceability. Existing systems mostly rely on material batches, inventory quantities, production orders, and formula requirements for material recommendation and flow management to ensure the continuity of material supply and the integrity of processing records during the production process.

[0003] Current food production and processing material control relies heavily on inventory records and manual production scheduling, making it difficult to accurately match material batch status with product formula constraints; the lack of unified boundary judgment for mandatory attributes, optimization attributes, and prohibition conditions easily leads to a disconnect between material requisition scheduling and material input verification; the lack of continuous lineage tracking between processing events, output results, and abnormal material identification results in delayed material consumption verification and makes it difficult to accurately locate the range of affected finished products. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a material control system for food production and processing.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a food production and processing material control system, comprising: The material batch status generation module is used to receive basic material data and inventory collection data in food production and processing, perform batch status processing on the basic material data and inventory collection data, and generate material batch status data. The formula constraint matching module is used to receive material batch status data and product formula constraint data, perform material formula boundary approximate matching processing based on the MABAC algorithm on the material batch status data and product formula constraint data, and generate material formula matching data. The material requisition and scheduling generation module is used to receive material formula matching data and production order data, perform material requisition and scheduling generation processing on the material formula matching data and production order data, and generate material requisition and scheduling data. The material feeding risk verification module is used to receive material requisition and production scheduling data, material batch status data and material feeding request data, perform material feeding risk verification processing on the material requisition and production scheduling data, material batch status data and material feeding request data, and generate material feeding permission status data. The material conversion record module is used to receive material feeding permit status data and production processing event data, perform material conversion association processing on the material feeding permit status data and production processing event data, and generate material batch lineage data; The material balance verification module is used to receive material batch lineage data and production output data, perform material consumption consistency verification processing on the material batch lineage data and production output data, and generate material balance verification data. The anomaly range positioning module is used to receive material batch lineage data, material balance verification data, and abnormal material identification data, and to perform abnormal finished product range positioning processing on the material batch lineage data, material balance verification data, and abnormal material identification data to generate affected finished product range data.

[0006] As a further description of the above technical solution: The material batch status generation module receives basic material data and inventory collection data from food production and processing. It collects the corresponding content in the basic material data according to the material batch identifier and connects the corresponding content in the inventory collection data to the collected content according to the same material batch identifier, thereby generating material inventory association data. Perform consistency checks on the corresponding contents of the same material batch identifier in the material inventory association data, and organize the corresponding contents that pass the consistency check according to the inventory location, inventory quantity and inventory status to generate material batch organization data; Connect the material batch sorting data with the corresponding content in the material basic data, and then arrange the connected corresponding content in a unified manner according to the material batch identifier, inventory location, inventory quantity and inventory status to generate material batch status data.

[0007] As a further description of the above technical solution: The formula constraint matching module receives material batch status data and product formula constraint data, and connects the candidate material batches in the material batch status data with the corresponding contents in the product formula constraint data according to the candidate material batches to generate a candidate material batch evaluation matrix. The evaluation matrix of candidate material batches is processed to unify the attribute direction, converting positive and negative attributes into the same evaluation direction, generating a unified matrix of candidate material batches. The unified matrix of candidate material batches is then normalized to generate a normalized matrix of candidate material batches. The weights of the evaluation attributes are determined based on the product formula constraint data, and the normalized matrix of candidate material batches is weighted to generate a weighted normalized matrix of candidate material batches. Based on the mandatory attributes in the product formula constraint data, the candidate material batches corresponding to the weighted normalization matrix of the candidate material batches are divided into available candidate domains and prohibited candidate domains. The available candidate domain is processed by the available domain description algorithm based on the support vector data description algorithm to generate an approximate region of available boundaries. The corresponding content in the forbidden candidate domain is processed by the forbidden domain description algorithm based on the support vector data description algorithm to generate an approximate region of forbidden boundaries. Calculate the available boundary distance of the candidate material batch relative to the approximate area of ​​the available boundary, and calculate the prohibited boundary distance of the candidate material batch relative to the approximate area of ​​the prohibited boundary; Determine the distance content corresponding to the mandatory attribute from the available boundary distances, and generate the mandatory attribute locking distance based on the distance content corresponding to the mandatory attribute; The corresponding content in the product formula constraint data, excluding mandatory attributes and used for comprehensive ranking, is identified as the optimization attribute; Based on the mandatory attribute locking distance control optimization attribute corresponding to the available boundary distance, the comprehensive sorting is used to generate the comprehensive matching distance of candidate material batches; Candidate material batches are classified and processed according to the comprehensive matching distance of the candidate material batches, generating direct matching sorting content, verification matching content, and prohibited matching content; The directly matched sorted content, the verified matched content, and the prohibited matched content are merged to generate material formula matching data.

[0008] As a further description of the above technical solution: Available domain description processing and forbidden domain description processing based on support vector data description algorithms include: Read the rows of the candidate material batch weighted normalized matrix corresponding to the available candidate domain from the candidate material batch weighted normalized matrix to generate the available domain training matrix; read the rows of the candidate material batch weighted normalized matrix corresponding to the prohibited candidate domain from the candidate material batch weighted normalized matrix to generate the prohibited domain training matrix. Each row in the available domain training matrix is ​​used as an available domain sample vector, and each row in the forbidden domain training matrix is ​​used as a forbidden domain sample vector, while maintaining the correspondence between the available domain sample vector, the forbidden domain sample vector and the candidate material batch; Perform availability domain envelope modeling on availability domain sample vectors to determine the availability domain center and availability domain envelope radius; perform forbidden domain envelope modeling on forbidden domain sample vectors to determine the forbidden domain center and forbidden domain envelope radius. Based on the distance from the available domain sample vector to the center of the available domain, the available domain support vectors located on the boundary of the available domain envelope are filtered, and the available domain envelope boundary data is generated based on the center of the available domain, the radius of the available domain envelope, and the available domain support vectors. Based on the distance from the forbidden region sample vector to the center of the forbidden region, the forbidden region support vectors located on the forbidden region envelope boundary are selected, and forbidden region envelope boundary data is generated based on the forbidden region center, forbidden region envelope radius, and forbidden region support vectors. Based on the available domain envelope boundary data, determine the available domain boundary position of the candidate material batch in the weighted normalized attribute space, and write the available domain boundary position into the available boundary approximation region. Based on the forbidden zone envelope boundary data, determine the forbidden zone boundary position of the candidate material batch in the weighted normalized attribute space, and write the forbidden zone boundary position into the forbidden zone approximate region. Calculate the available boundary distance of the candidate material batch relative to the approximate area of ​​the available boundary, and calculate the prohibited boundary distance of the candidate material batch relative to the approximate area of ​​the prohibited boundary.

[0009] As a further description of the above technical solution: The material requisition and scheduling generation module receives material formula matching data and production order data. It connects the corresponding content in the production order data with the direct matching sorting content, verification matching content and prohibited matching content in the material formula matching data according to the candidate material batches to generate order material association data. The candidate material batches in the order material association data are compared with the prohibited matching content, and the candidate material batches that exist in the prohibited matching content are removed from the order material association data to generate order material requisition data; Arrange the candidate material batches in the order material requisition data according to the sorting relationship in the direct matching sorting content, and mark the candidate material batches that exist in the review matching content as review candidate content, and generate candidate material requisition order data; The candidate material batches in the candidate material requisition sequence data are matched with the production sequence in the production order data, and the candidate material batches belonging to the same production order are aggregated according to the production sequence to generate material requisition and production scheduling data that includes production order, candidate material batches, material requisition sequence and review candidate content.

[0010] As a further description of the above technical solution: The material feeding risk verification module receives material requisition and production scheduling data, material batch status data, and material feeding request data. It connects the corresponding content in the material feeding request data with the candidate material batches, material requisition order, and review candidate content in the material requisition and production scheduling data according to the candidate material batches. It then associates the connected corresponding content with the corresponding content in the material batch status data to generate material feeding request association data. The candidate material batches in the material feeding request associated data are checked for consistency with the candidate material batches and material feeding order in the material requisition and production scheduling data. The corresponding contents that fail the consistency check are marked as prohibited material feeding contents, and material feeding order check data is generated. The candidate material batches that pass the consistency check in the feeding sequence check data are checked against the corresponding contents in the material batch status data, and the corresponding contents that fail the status check are marked as prohibited feeding contents, thus generating feeding status check data. The candidate material batches that pass the status check in the material feeding status check data are compared with the review candidate contents in the material requisition and production scheduling data. The corresponding contents existing in the review candidate contents are marked as review feeding contents. The corresponding contents that pass the status check and are not marked as prohibited feeding contents or review feeding contents are marked as direct feeding contents, thus generating material feeding permit status data.

[0011] As a further description of the above technical solution: The material conversion record module receives material feeding permit status data and production processing event data. It connects the corresponding content in the production processing event data with the direct material feeding content, verification material feeding content, and prohibited material feeding content in the material feeding permit status data according to the candidate material batch. It also removes the production processing event data corresponding to the prohibited material feeding content from the corresponding content after connection to generate convertible event data. The production and processing event data corresponding to the direct feeding content and the verified feeding content in the convertible event data are arranged according to the candidate material batch, production order and processing order. The corresponding content after arrangement is connected with the event sequence in the production and processing event data to generate material conversion order data. Candidate material batches belonging to the same production order in the material conversion sequence data are connected sequentially according to the processing order, and the connected candidate material batches are bound to the corresponding content in the production processing event data to generate material conversion association data. The candidate material batches, production orders, and processing order in the material conversion association data are written into the same batch association structure to generate material batch lineage data.

[0012] As a further description of the above technical solution: The material balance verification module receives material batch lineage data and production output data, and connects the corresponding content in the production output data with the candidate material batches, production orders and processing order in the material batch lineage data according to the production order to generate production lineage association data. The candidate material batches belonging to the same production order in the generated bloodline association data are matched with the corresponding contents in the production output data according to the processing order. The matched contents are then collected according to the candidate material batches to generate material output verification data. The consumption consistency of candidate material batches, production orders, and corresponding contents in production output data in the material output verification data is checked. The corresponding contents that pass the consumption consistency check and the corresponding contents that fail the consumption consistency check are marked with status to generate material balance verification data.

[0013] As a further description of the above technical solution: The anomaly range location module receives material batch lineage data, material balance verification data, and abnormal material identification data. It connects the abnormal material identification data with the candidate material batches in the material batch lineage data according to the candidate material batches, and then checks the corresponding content after connection with the corresponding content in the material balance verification data to generate abnormal lineage association data. The candidate material batches that are connected to the abnormal material identification data in the abnormal lineage association data are taken as the abnormal starting content. The candidate material batches and production orders that have a sequential connection with the abnormal starting content are searched along the batch association structure in the material batch lineage data according to the processing order, and abnormal transmission association data is generated. The abnormal transmission correlation data is compared with the material balance verification data, and the corresponding content that fails the consumption consistency verification is marked in the abnormal transmission correlation data to generate abnormal finished product candidate data. The candidate material batches belonging to the same production order in the abnormal finished product candidate data are grouped according to the processing order, and the corresponding content after grouping is linked with the abnormal material identification data to generate abnormal finished product range sorting data. Write the corresponding contents from the production orders, candidate material batches, abnormal material identification data, and material balance verification data in the abnormal finished product range data into the same range location result to generate the affected finished product range data.

[0014] The present invention has the following beneficial effects: 1. In this invention, firstly, a material state stabilization mechanism based on consistent batch identifiers is used to collect, verify, and uniformly arrange basic material data and inventory collection data under the same material batch identifier. This ensures that inventory location, quantity, and status are stably expressed within the same material batch state data, solving the problem of unstable connection between basic material batch information and inventory collection information in food production and processing, and improving the consistency, accuracy, and stability of subsequent formula matching input data. Secondly, a formula boundary discrimination mechanism based on the fusion of the MABAC algorithm and support vector data description algorithm is used to hierarchically process the mandatory and optimization attributes in the product formula constraint data, and candidates are formed through available candidate domains, prohibited candidate domains, approximate regions of available boundaries, and approximate regions of prohibited boundaries. The boundary distance representation of material batches enables material formula matching results to simultaneously reflect direct matching, verification matching, and prohibited matching statuses. This solves the problem that a single sorting method is insufficient to identify prohibited risks and boundary candidates, improving the identification accuracy, risk sensitivity, and interpretability of material formula matching results. Through a material requisition and feeding linkage control mechanism based on the continuity constraints of matching results, the direct matching sorting content, verification matching content, and prohibited matching content in the material formula matching data are extended to the material requisition scheduling data and feeding permission status data. This ensures that the material requisition order, verification candidate content, prohibited feeding content, verification feeding content, and direct feeding content maintain a continuous constraint relationship, solving the problem of disconnect between formula matching and on-site feeding execution, and improving the stability of material requisition scheduling, the accuracy of feeding verification, and linkage control capabilities.

[0015] 2. In this invention, a material lineage balance verification mechanism based on batch association structure is used to connect material input permit status data with production and processing event data to form convertible event data, material conversion sequence data, and material batch lineage data. The material batch lineage data is then compared with production output data for consumption consistency, ensuring a continuous tracking relationship between candidate material batches, production orders, and processing sequence between material consumption and finished product output. This solves the problem of unclear correlation between the material input and processing process and the output verification results, improving the accuracy of material consumption verification, the traceability of the production process, and the interpretability of the verification results. Furthermore, an anomaly range positioning mechanism based on the linkage between lineage transmission and balance verification connects abnormal material identification data with material batch lineage data and material balance verification data. Anomaly transmission correlation data is formed along the batch association structure and processing sequence. This data, combined with the corresponding content that failed the consumption consistency verification, generates the affected finished product range data. This solves the problem of accurately locating the affected finished product range after abnormal materials are discovered, improving the ability to locate abnormal finished products, response time, and decision support capabilities. Attached Figure Description

[0016] Figure 1 This is a system architecture diagram of the present invention. Detailed Implementation

[0017] 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.

[0018] Reference Figure 1 One embodiment of the present invention provides a food production and processing material control system, comprising: The material batch status generation module is used to receive basic material data and inventory collection data in food production and processing, perform batch status processing on the basic material data and inventory collection data, and generate material batch status data. The formula constraint matching module is used to receive material batch status data and product formula constraint data, perform material formula boundary approximate matching processing based on the MABAC algorithm on the material batch status data and product formula constraint data, and generate material formula matching data. The material requisition and scheduling generation module is used to receive material formula matching data and production order data, perform material requisition and scheduling generation processing on the material formula matching data and production order data, and generate material requisition and scheduling data. The material feeding risk verification module is used to receive material requisition and production scheduling data, material batch status data and material feeding request data, perform material feeding risk verification processing on the material requisition and production scheduling data, material batch status data and material feeding request data, and generate material feeding permission status data. The material conversion record module is used to receive material feeding permit status data and production processing event data, perform material conversion association processing on the material feeding permit status data and production processing event data, and generate material batch lineage data; The material balance verification module is used to receive material batch lineage data and production output data, perform material consumption consistency verification processing on the material batch lineage data and production output data, and generate material balance verification data. The anomaly range positioning module is used to receive material batch lineage data, material balance verification data, and abnormal material identification data, and to perform abnormal finished product range positioning processing on the material batch lineage data, material balance verification data, and abnormal material identification data to generate affected finished product range data.

[0019] In this embodiment, the material batch status generation module receives basic material data and inventory collection data from food production and processing. First, it identifies corresponding content with material batch identifiers from the basic material data and uses these identifiers as the aggregation basis, grouping corresponding content under the same material batch identifier into the same aggregated corresponding content. Then, it compares the corresponding content in the inventory collection data with the already aggregated material batch identifiers, allowing content in the inventory collection data that corresponds to the same material batch identifier to be included in the aggregated corresponding content. The corresponding content in the inventory collection data does not change the existing aggregation relationship in the basic material data; instead, it fills in the corresponding positions around the same material batch identifier, connecting the corresponding content in the basic material data and the corresponding content in the inventory collection data under the same material batch identifier. Finally, it saves the already connected corresponding content under the same material batch identifier as associated content of the same batch, generating material inventory association data.

[0020] From the material inventory association data, locate the corresponding content that has formed a connection according to the same material batch identifier, and perform consistency verification on the corresponding content from the material base data and inventory collection data under the same material batch identifier; retain the content that can correspond to each other under the same material batch identifier in the same verification result, and exclude the content that cannot form a consistent correspondence with the same material batch identifier from the corresponding content that has passed the consistency verification; organize the corresponding content that has passed the consistency verification around the inventory location, inventory quantity, and inventory status, so that the inventory location, inventory quantity, and inventory status under the same material batch identifier maintain the same arrangement relationship; inventory location is used to limit the arrangement position of the corresponding content, inventory quantity is used to limit the quantity relationship of the corresponding content, and inventory status is used to limit the status relationship of the corresponding content. All three are kept under the same material batch identifier and are not mixed with the corresponding content under other material batch identifiers, generating material batch organized data.

[0021] The material batch processing data and the corresponding content in the material base data are connected according to the material batch identifier, so that the inventory location, inventory quantity, and inventory status already processed in the material batch processing data are connected to the corresponding content under the same material batch identifier in the material base data. First, the arrangement relationship of inventory location, inventory quantity, and inventory status under the same material batch identifier in the material batch processing data is maintained, and then the corresponding content in the material base data is incorporated into this arrangement relationship, so that the connected corresponding content simultaneously inherits the corresponding content in both the material base data and the material batch processing data. The connected corresponding content is then uniformly arranged around the material batch identifier, and the order of inventory location, inventory quantity, and inventory status is maintained under the same material batch identifier, so that the content corresponding to the same material batch identifier forms a data structure that can be received by the subsequent formula constraint matching module to generate material batch status data.

[0022] In this embodiment, the formula constraint matching module receives material batch status data and product formula constraint data. It reads the corresponding content related to candidate material batches from the material batch status data and reads the content from the product formula constraint data that can establish a correspondence with candidate material batches. The read material batch status data maintains the distinguishing relationship between candidate material batches, and the read product formula constraint data maintains the matching relationship with candidate material batches. When connecting candidate material batches in the material batch status data with the corresponding content in the product formula constraint data, the candidate material batches are used as the common connection object, so that each candidate material batch corresponds to its related content in the product formula constraint data. The connected corresponding content maintains the arrangement relationship between candidate material batches and uses the candidate material batches as the basis for arranging the corresponding content in the matrix, so that candidate material batches and their corresponding product formula constraint data content form corresponding positions in the same matrix, generating a candidate material batch evaluation matrix. The formula for the candidate material batch evaluation matrix is: ; Candidate material batch evaluation matrix : No. The candidate material batch in the first The evaluation value under each evaluation attribute : The number of candidate material batches The number of evaluation attributes. : The sequence number of the candidate material batch. , The index of the evaluation attribute. , : A set of candidate material batch numbers , : A set of evaluation attribute numbers , The first item in the material batch status data One batch of candidate materials : Product formulation constraint data and the first The content corresponding to each evaluation attribute : Batch of candidate materials Corresponding content in the product formulation constraint data The function that retrieves the evaluation value after connection.

[0023] For the content corresponding to the positive attributes and the content corresponding to the negative attributes in the candidate material batch evaluation matrix, the original correspondence between candidate material batches is maintained, and the positive and negative attributes are converted to the same evaluation direction. During the attribute direction unification process, the arrangement relationship between candidate material batches in the candidate material batch evaluation matrix is ​​not changed, nor is the connection relationship between the corresponding content in the product formula constraint data and the candidate material batches. Instead, the positive and negative attributes are made able to undergo subsequent normalization processing under the same evaluation direction, generating a candidate material batch direction unification matrix. After the candidate material batch direction unification matrix is ​​formed, the candidate material batch direction is unified... The corresponding content in a matrix is ​​normalized to ensure that the expression range of different corresponding content in the candidate material batch-oriented unified matrix remains consistent, generating a candidate material batch normalization matrix. When determining the evaluation attribute weights based on product formula constraint data, a correspondence is established between the evaluation attribute weights and the corresponding evaluation attributes in the candidate material batch normalization matrix, and the evaluation attribute weights are applied to the corresponding content in the candidate material batch normalization matrix. After weighting, the corresponding content in the candidate material batch normalization matrix forms a weighted corresponding content while maintaining the candidate material batch arrangement relationship, generating a candidate material batch weighted normalization matrix. The formula for the candidate material batch-oriented unified matrix is: ; ; Candidate material batch direction unification matrix : No. The candidate material batch in the first The evaluation value after all evaluation attributes have been aligned in the same direction. : No. The candidate material batch in the first The evaluation value under each evaluation attribute : A set of positive attribute indices : A set of inverse attribute ordinal numbers The candidate material batch evaluation matrix The maximum rating value for each rating attribute The first candidate material batch evaluation matrix The minimum rating value of each rating attribute. : The sequence number of the candidate material batch. : The sequence number of the evaluation attribute. Formula for the normalized matrix of candidate material batches: ; ; Candidate material batch normalization matrix : No. The candidate material batch in the first The evaluation value after normalization of each evaluation attribute : No. The candidate material batch in the first The evaluation value after all evaluation attributes have been aligned in the same direction. The first in the batch direction unification matrix of candidate materials The maximum value of each evaluation attribute. The first in the batch direction unification matrix of candidate materials The minimum value of each evaluation attribute. : The sequence number of the candidate material batch. : The sequence number of the evaluation attribute. Formula for the batch weighted normalization matrix of candidate materials: ; Batch-weighted normalized matrix of candidate materials : No. The candidate material batch in the first The evaluation value after weighted normalization of each evaluation attribute The first step in determining product formulation constraint data. Each evaluation attribute weight, : No. The candidate material batch in the first The evaluation value after normalization of each evaluation attribute : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0024] Establish a correspondence between the mandatory attributes in the product formula constraint data and the candidate material batches corresponding to the weighted normalization matrix of candidate material batches. When establishing this correspondence, ensure that the mandatory attributes originate from the product formula constraint data and that the candidate material batches originate from the weighted normalization matrix of candidate material batches. Classify candidate material batches in the weighted normalization matrix that satisfy the mandatory attributes into the available candidate domain, and ensure that these batches retain their corresponding content in the weighted normalization matrix. Classify candidate material batches in the weighted normalization matrix that do not satisfy the mandatory attributes into the prohibited candidate domain, and ensure that these batches retain their corresponding content in the weighted normalization matrix. After the candidate material batches are divided, the candidate material batches in the weighted normalization matrix will have a clear correspondence with both the available and prohibited candidate domains. Formulas for the available and prohibited candidate domains: ; ; : No. The candidate material batch in the first A satisfying flag under a mandatory attribute : No. The candidate material batch in the first The evaluation value after weighted normalization of each evaluation attribute The first in the product formula constraint data The constraint content corresponding to each mandatory attribute. : No. The mandatory attributes of each candidate material batch are comprehensively satisfied by the flag. : Set of mandatory attribute numbers Available candidate domains Forbidden candidate domains : No. One batch of candidate materials : A set of candidate material batch numbers : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0025] When performing availability domain description and forbidden domain description processing based on the support vector data description algorithm, the candidate material batch weighted normalization matrix is ​​first read, maintaining the arrangement relationship between candidate material batches in the matrix. Rows in the candidate material batch weighted normalization matrix that correspond to the available candidate domains are then separated, ensuring that each separated row still corresponds to the original candidate material batch. These separated rows are then arranged according to the correspondence between candidate material batches and available candidate domains, ensuring that the corresponding content within the available candidate domain is continuously expressed in the same matrix structure, thus generating availability domain training. The matrix is ​​used to separate the rows of the candidate material batch weighted normalized matrix that correspond to the prohibited candidate domain, ensuring that each separated matrix row still corresponds to the original candidate material batch. These separated matrix rows are then arranged according to the correspondence between candidate material batches and the prohibited candidate domain, maintaining continuous representation of the corresponding content within the same matrix structure, thus generating the prohibited domain training matrix. The available domain training matrix and the prohibited domain training matrix respectively inherit the existing candidate material batch correspondences from the candidate material batch weighted normalized matrix without altering the weighted normalized representation within the candidate material batch weighted normalized matrix. The formula for the available domain training matrix is: Formula for forbidden region training matrix: ; : Usable domain training matrix, Forbidden-entry training matrix : No. The candidate material batch in the first The evaluation value after weighted normalization of each evaluation attribute : No. One batch of candidate materials Available candidate domains Forbidden candidate domains : A set of evaluation attribute numbers : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0026] Each row in the availability domain training matrix is ​​treated as an availability domain sample vector, maintaining the correspondence between this sample vector and the original rows in the availability domain training matrix. Each availability domain sample vector is then mapped to a candidate material batch within the availability candidate domain, ensuring the availability domain sample vector carries the relationship between the availability candidate domain and the candidate material batch. Similarly, each row in the restricted domain training matrix is ​​treated as a restricted domain sample vector, maintaining the correspondence between this sample vector and the original rows in the restricted domain training matrix. Each restricted domain sample vector is then mapped to a candidate material batch within the restricted candidate domain, ensuring the restricted domain sample vector carries the relationship between the restricted candidate domain and the candidate material batch. During the formation of availability domain and restricted domain sample vectors, the row arrangement in the availability domain and restricted domain training matrices remains unchanged, and the correspondence between candidate material batches and availability / restriction domain sample vectors is consistently preserved. The formula for the availability domain sample vector is: , Formula for forbidden region sample vector: ; : No. Available domain sample vectors : No. A forbidden region sample vector, : No. The weighted and normalized evaluation values ​​of each candidate material batch under each evaluation attribute. The number of evaluation attributes. : No. One batch of candidate materials Available candidate domains Forbidden candidate domains : Vector transpose symbol : Serial number of candidate material batch Availability domain sample vectors are modeled using availability domain envelopes. In this modeling process, each availability domain sample vector is treated as a sample within the same availability domain, and the availability domain sample vectors are arranged around their distribution within the candidate availability domains to form availability domain centers. During the determination of availability domain centers, the correspondence between each availability domain sample vector and the candidate material batch is maintained, ensuring that the availability domain centers reflect the overall positional relationship between the availability domain sample vectors. After the availability domain centers are determined, the envelope relationship between the availability domain sample vectors and the availability domain centers is calculated, and the availability domain envelope radius is determined using this relationship. No-entry domain sample vectors are modeled using no-entry domain envelopes. In network modeling, forbidden region sample vectors are used as sample representations within the same forbidden region, and forbidden region centers are formed around their distribution relationships within candidate forbidden regions. During the determination of forbidden region centers, the correspondence between each forbidden region sample vector and candidate material batches is maintained, ensuring that the forbidden region centers reflect the overall positional relationships between forbidden region sample vectors. After determining the forbidden region centers, the envelope relationship between the forbidden region sample vectors and the forbidden region centers is calculated, and the envelope radius is determined using this relationship. The available domain center and available domain envelope radius jointly encompass the available domain sample vectors, and the forbidden region center and forbidden domain envelope radius jointly encompass the forbidden region sample vectors. Formulas for available domain center, available domain envelope radius, forbidden region center, and forbidden region envelope radius are as follows: ; ; :domain The corresponding center; when When it is the center of the availability zone, At that time, it was the center of the restricted area. :domain The corresponding envelope radius; when When is the radius of the available domain envelope, when The radius of the forbidden domain envelope is [time]. : Domain flag, whose value is an available candidate domain Or ban candidate domain , Candidate centers in the optimization process of support vector data description algorithms. Candidate envelope radius during the optimization process of support vector data description algorithms. : No. The envelope relaxation amount corresponding to each batch of candidate materials. :domain The corresponding envelope constraint coefficients, : No. One batch of candidate materials :domain The Middle a sample vector; when Time is the available domain sample vector ,when When is the forbidden region sample vector , Sample vector Feature mapping results in support vector data description algorithms : The distance between the sample vector mapping result and the candidate center The value of the variable that minimizes the objective function. : The sequence number of the candidate material batch. Formulas for available region support vectors and forbidden region support vectors: ; :domain The corresponding set of support vectors; when When is the set of support vectors in the available domain, when The set of support vectors for the forbidden region is given at this time. : No. One batch of candidate materials : Domain flag, whose value is an available candidate domain Or ban candidate domain , :domain The Middle A sample vector, Sample vector Feature mapping results :domain The corresponding center, :domain The corresponding envelope radius, : The sequence number of the candidate material batch.

[0027] The distance between each availability zone sample vector and the availability zone center is calculated, and the relationship between each distance calculation result and the corresponding availability zone sample vector is maintained. The distance from each availability zone sample vector to the availability zone center is mapped to the availability zone envelope radius, so that the availability zone sample vector can determine whether it is located on the availability zone envelope boundary based on its relationship with the availability zone envelope radius. Availability zone sample vectors located on the availability zone envelope boundary are selected as availability zone support vectors, and the correspondence between availability zone support vectors and candidate material batches is maintained. After the availability zone support vectors are determined, the availability zone center, availability zone envelope radius, and availability zone support vectors are connected, so that the availability zone center is used to limit the center correspondence of availability zone support vectors, and the availability zone envelope radius is used to limit the boundary correspondence of availability zone support vectors. Availability zone envelope boundary data is generated through the connection relationship between availability zone center, availability zone envelope radius, and availability zone support vectors.

[0028] Calculate the distance between each forbidden region sample vector and the forbidden region center, maintaining the relationship between each distance calculation result and the corresponding forbidden region sample vector. Map the distance from each forbidden region sample vector to the forbidden region center to the forbidden region envelope radius, allowing the forbidden region sample vector to determine whether it lies on the forbidden region envelope boundary based on its relationship with the forbidden region envelope radius. Select forbidden region sample vectors located on the forbidden region envelope boundary as forbidden region support vectors, maintaining the correspondence between the forbidden region support vectors and candidate material batches. After determining the forbidden region support vectors, connect the forbidden region center, forbidden region envelope radius, and forbidden region support vectors. The forbidden region center is used to define the center correspondence of the forbidden region support vectors, and the forbidden region envelope radius is used to define the boundary correspondence of the forbidden region support vectors. Generate forbidden region envelope boundary data through the connection relationships between the forbidden region center, forbidden region envelope radius, and forbidden region support vectors. Formulas for available domain envelope boundary data and forbidden region envelope boundary data are as follows: ; :domain The corresponding envelope boundary data; when When the available domain envelope boundary data is available, At that time, the data was the boundary of the forbidden domain envelope. : No. One batch of candidate materials :domain The Middle A sample vector, :domain The corresponding center, :domain The corresponding envelope radius, :domain The corresponding set of support vectors, : Domain flag, whose value is an available candidate domain Or ban candidate domain , : The sequence number of the candidate material batch.

[0029] The system maintains the connectivity between the availability domain center, availability domain envelope radius, and availability domain support vectors from the availability domain envelope boundary data, and applies this connectivity to the position representation of candidate material batches in the weighted normalized attribute space. Based on the correspondence between availability domain support vectors and candidate material batches, the relative position between candidate material batches and availability domain envelope boundary data is determined. The boundary representation formed by availability domain support vectors is constrained by the availability domain center and availability domain envelope radius, enabling the determination of the availability domain boundary position of candidate material batches in the weighted normalized attribute space. After the availability domain boundary position is determined, it is written into the availability boundary approximation region, allowing the availability boundary approximation region to inherit the center, radius, and support vector relationships already formed in the availability domain envelope boundary data, and maintaining the correspondence between the availability boundary approximation region and candidate material batches. Formulas for the availability boundary approximation region and the forbidden boundary approximation region are as follows: ; :domain The corresponding boundary approximate region; when When the usable boundary approximates the region, At that time, the area was approximately within the prohibited boundary. : No. Each candidate material batch relative to the domain The boundary position, : A set of candidate material batch numbers : Domain flag, whose value is an available candidate domain Or ban candidate domain , : The sequence number of the candidate material batch.

[0030] This process preserves the connectivity between the forbidden region center, forbidden region envelope radius, and forbidden region support vectors from the forbidden region envelope boundary data, and applies this connectivity to the position representation of candidate material batches in the weighted normalized attribute space. Based on the correspondence between forbidden region support vectors and candidate material batches, the relative position between candidate material batches and forbidden region envelope boundary data is determined. The boundary representation formed by forbidden region support vectors is defined by the forbidden region center and forbidden region envelope radius, enabling the determination of the forbidden region boundary position of candidate material batches in the weighted normalized attribute space. After determining the forbidden region boundary position, it is written into the forbidden region approximation area, allowing the approximation area to inherit the center, radius, and support vector relationships already formed in the forbidden region envelope boundary data, while maintaining the correspondence between the approximation area and candidate material batches. Formulas for available domain boundary position and forbidden region boundary position are provided. ; : No. Each candidate material batch relative to the domain The boundary position; when When is the location of the available domain boundary, when At that time, it was the boundary of the restricted area. :domain The corresponding center, :domain The corresponding envelope radius, : No. Feature mapping results of weighted normalized vectors of candidate material batches : No. The row vectors of each candidate material batch in the weighted normalized matrix of candidate material batches. , : No. The candidate material batch in the first The evaluation value after weighted normalization of each evaluation attribute : Domain flag, whose value is an available candidate domain Or ban candidate domain , : The sequence number of the candidate material batch. The number of evaluation attributes.

[0031] Distance calculations are performed between candidate material batches and the approximate region of the available boundary, maintaining the correspondence between the candidate material batches and the boundary positions of the available domain during the calculation process. Each candidate material batch forms a corresponding distance value relative to the approximate region of the available boundary, which is saved according to the arrangement relationship between candidate material batches, generating the available boundary distance. Simultaneously, distance calculations are performed between candidate material batches and the approximate region of the restricted boundary, maintaining the correspondence between the candidate material batches and the boundary positions of the restricted domain during the calculation process. Each candidate material batch forms a corresponding distance value relative to the approximate region of the restricted boundary, which is saved according to the arrangement relationship between candidate material batches, generating the restricted boundary distance. The available boundary distance and the restricted boundary distance respectively represent the distance relationships between candidate material batches and the approximate regions of the available and restricted boundaries. Available boundary distance formula: , ; Establish distance calculation relationships between candidate material batches and the approximate regions of the available boundary and prohibited boundary, respectively. When calculating the available boundary distance of a candidate material batch relative to the approximate region of the available boundary, maintain the source relationship between the corresponding content in the weighted normalization matrix of the candidate material batches, so that each candidate material batch forms a distance content corresponding to the approximate region of the available boundary, generating the available boundary distance. When calculating the prohibited boundary distance of a candidate material batch relative to the approximate region of the prohibited boundary, maintain the correspondence between the same candidate material batch and the approximate region of the prohibited boundary, so that each candidate material batch forms a distance content corresponding to the approximate region of the prohibited boundary, generating the prohibited boundary distance. The correspondence between candidate material batches is retained for both the available boundary distance and the prohibited boundary distance. Prohibited boundary distance formula: ; : No. Each candidate material batch relative to the domain The boundary distance; when When is the available boundary distance, when At that time, the distance to the restricted boundary was... : No. The candidate material batch in the first Each evaluation attribute is relative to the domain The property boundary distance, : No. Feature mapping results of weighted normalized vectors of candidate material batches : No. The row vectors of each candidate material batch in the weighted normalized matrix of candidate material batches. :domain The corresponding center, :domain The corresponding envelope radius, : No. The candidate material batch in the first The evaluation value after weighted normalization of each evaluation attribute : No. Each candidate material batch relative to the domain The boundary position of the first The components corresponding to each evaluation attribute : Domain flag, whose value is an available candidate domain Or ban candidate domain , : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0032] The distance content corresponding to the mandatory attribute is determined from the available boundary distance. During this process, the mandatory attribute originates from product formula constraint data, and the distance content originates from the available boundary distance. Distance content corresponding to the mandatory attribute within the same candidate material batch is grouped into the same correspondence. After aggregating the distance content corresponding to the mandatory attribute, the correspondence between candidate material batches and mandatory attributes remains unchanged, and the distance content corresponding to the mandatory attribute is used to form the distance expression controlling subsequent comprehensive ranking. When generating the mandatory attribute locking distance, the mandatory attribute locking distance is assigned to candidate material batches, and the mandatory attribute locking distance inherits the distance content corresponding to the mandatory attribute from the available boundary distance, thus generating the mandatory attribute locking distance. The formula for the mandatory attribute locking distance is: ; : No. The mandatory attribute locking distance corresponding to each batch of candidate materials. : Set of mandatory attribute numbers : No. The candidate material batch in the first A satisfying flag under a mandatory attribute : No. The candidate material batch in the first The distance of an evaluation attribute relative to the boundary of the available candidate domain. : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0033] In the product formula constraint data, excluding mandatory attributes, the corresponding content used for comprehensive ranking is identified as optimization attributes. When determining optimization attributes, the source relationship between mandatory attributes and other corresponding content in the product formula constraint data remains unchanged. Instead, the corresponding content (excluding mandatory attributes) participating in comprehensive ranking is separated from the product formula constraint data, while maintaining its correspondence with candidate material batches. After the optimization attributes are determined, the source relationship between the optimization attributes and the product formula constraint data is maintained, and the connection relationship required for comprehensive ranking is maintained between the optimization attributes and candidate material batches, thus generating the optimization attributes. Optimization attribute formula: Optimize the attribute index set. The index of the evaluation attribute. : A set of evaluation attribute numbers : Set of mandatory attribute numbers The first in the product formula constraint data A marker indicating whether an evaluation attribute is used for comprehensive ranking; Indicates the first One evaluation attribute is used for comprehensive ranking.

[0034] The mandatory attribute locking distance is connected to the available boundary distance corresponding to the optimized attribute, allowing the mandatory attribute locking distance to control whether the available boundary distance corresponding to the optimized attribute participates in the comprehensive ranking. During the control process, the mandatory attribute locking distance maintains its correspondence with candidate material batches, and the available boundary distance corresponding to the optimized attribute maintains its correspondence with the same candidate material batch. The available boundary distances corresponding to the optimized attributes allowed to participate in the comprehensive ranking are aggregated with the mandatory attribute locking distance within the same candidate material batch, while maintaining the distinction between candidate material batches. After the comprehensive ranking is completed, each candidate material batch generates a corresponding comprehensive matching distance, thus generating the candidate material batch comprehensive matching distance formula: ; : No. Overall matching distance of each candidate material batch : No. The mandatory attributes of each candidate material batch are comprehensively satisfied by the flag. Optimize the attribute index set. The first step in determining product formulation constraint data. Each evaluation attribute weight, : No. The candidate material batch in the first The distance of an evaluation attribute relative to the boundary of the available candidate domain. : No. The distance of each candidate material batch relative to the boundary of the prohibited candidate domain. : No. The mandatory attribute locking distance corresponding to each batch of candidate materials. : The sequence number of the candidate material batch. : The index of the evaluation attribute.

[0035] Candidate material batches are categorized based on their comprehensive matching distance. During this categorization process, a one-to-one correspondence is maintained between the comprehensive matching distance and the candidate material batches themselves. Batches with a direct matching relationship are written into the direct matching ranking content. Batches requiring verification in subsequent processing are written into the verification matching content. Batches that cannot proceed to subsequent material requisition and production processes are written into the prohibited matching content. The direct matching ranking content, verification matching content, and prohibited matching content maintain their respective source relationships with the comprehensive matching distance of the candidate material batches and collectively utilize this distance to generate the direct matching ranking content, verification matching content, and prohibited matching content. The formula for direct matching ranking content is: Formula for verifying matching content: Forbidden matching formula: ;: Prohibited content matching : Review the matching content, : Directly match sorted content, : No. One batch of candidate materials : No. The mandatory attributes of each candidate material batch are comprehensively satisfied by the flag. : No. The distance of each candidate material batch relative to the boundary of the prohibited candidate domain. : No. Overall matching distance of each candidate material batch A sorting function that arranges items in descending order. : The sequence number of the candidate material batch.

[0036] The direct matching, verification matching, and prohibited matching content are merged around the candidate material batch. During the merging process, the ranking relationship within the direct matching content, the verification relationship within the verification matching content, and the prohibited relationship within the prohibited matching content are maintained, ensuring that the three types of content do not have conflicting relationships within the same candidate material batch. After merging, the direct matching, verification matching, and prohibited matching content corresponding to the candidate material batch form data results that can be received by the material requisition generation module, generating material formula matching data. Material formula matching data formula: ; Material formulation matching data : Directly match sorted content, : Review the matching content, Content matching is prohibited.

[0037] In this embodiment, the material requisition and production generation module receives material formula matching data and production order data. It maintains the classification relationship between directly matched, reviewed, and prohibited matching content in the material formula matching data, and also maintains the correspondence between corresponding content in the production order data and the production orders. When connecting corresponding content in the production order data with directly matched content, candidate material batches are used as the common connection basis, ensuring that content in the production order data corresponding to a candidate material batch is connected to content in the same candidate material batch in the directly matched content. Similarly, when connecting corresponding content in the production order data with reviewed matching content, candidate material batches are still used as the common connection basis, ensuring that content in the reviewed matching content corresponding to the same candidate material batch is connected. The content can be linked to the corresponding content in the production order data; when linking the corresponding content in the production order data with the prohibited matching content, the candidate material batch is still used as the linking basis, so that the content of the same candidate material batch in the prohibited matching content can be linked to the corresponding content in the production order data; during the linking process, the direct matching sorting content, the verification matching content, and the prohibited matching content do not replace each other, and the corresponding content in the production order data does not deviate from the original correspondence between the production order and the candidate material batch; after the linking is completed, the corresponding content in the production order data and the direct matching sorting content, verification matching content, and prohibited matching content in the material formula matching data form the same association relationship around the candidate material batch, generating order material association data.

[0038] The process involves comparing candidate material batches in the order material association data with prohibited matching content. During comparison, the connection between candidate material batches in the order material association data and their corresponding content in the production order data is maintained, and candidate material batches already formed in the prohibited matching content are used as the comparison basis. When a candidate material batch in the order material association data can form a corresponding relationship in the prohibited matching content, the candidate material batch, along with its corresponding content in the production order data already connected to the order material association data, is removed from the order material association data. When a candidate material batch in the order material association data cannot form a corresponding relationship in the prohibited matching content, the candidate material batch, along with its corresponding content in the production order data already connected to the order material association data, is retained. During the removal and retention process, the correspondence between the candidate material batches not removed from the order material association data and the directly matched sorting content and the verified matching content is not changed. After the comparison and removal are completed, the candidate material batches retained in the order material association data continue to maintain a connection with their corresponding content in the production order data, generating order material requisition data.

[0039] The process involves mapping candidate material batches in the order material availability data to their sorting relationships in the direct matching sorting content, ensuring that these batches inherit the established order from the direct matching sorting content. During sorting, the connection between candidate material batches in the order material availability data and their corresponding content in the production order data is maintained, without altering the existing material availability relationships. The process also involves comparing candidate material batches in the order material availability data with the verification matching content, ensuring that candidate material batches present in the verification matching content form a verification correspondence in the order material availability data. Candidate material batches present in the verification matching content are marked as verification candidates, and the connection between the verification candidates and their corresponding content in the production order data is maintained. For candidate material batches not present in the verification matching content, the sorting relationship in the direct matching sorting content is maintained. After completing the sorting and verification marking, candidate material batches in the order material availability data establish a material issuance sequence relationship, and the verification candidates are connected to their corresponding batches, generating candidate material issuance order data.

[0040] The candidate material batches in the candidate material requisition sequence data are matched with the production sequence in the production order data. During the matching process, the requisition sequence of the candidate material batches in the candidate material requisition sequence data is maintained, and the connection between the production order and the production sequence in the production order data is maintained, so that the candidate material batches simultaneously inherit the production sequence and the material requisition sequence under the corresponding production order. Candidate material batches belonging to the same production order are aggregated according to the production sequence, and the correspondence between the candidate material batches, the material requisition sequence, and the review candidate content is maintained during the aggregation process. For candidate material batches with review candidate content, the review candidate content is retained under the production order corresponding to the candidate material batch. For candidate material batches without review candidate content, the correspondence between the candidate material batch and the material requisition sequence is maintained. After the matching and aggregation are completed, the candidate material batches, the material requisition sequence, and the review candidate content under the same production order form a data result that can be received by the material feeding risk verification module, generating material feeding scheduling data that includes the production order, candidate material batches, the material requisition sequence, and the review candidate content.

[0041] In this embodiment, the material feeding risk verification module receives material requisition and production scheduling data, material batch status data, and material feeding request data. It first maintains the existing correspondence between candidate material batches, requisition order, and review candidate content in the material requisition and production scheduling data, and also maintains the existing correspondence between the corresponding content in the material batch status data and the candidate material batches. When connecting the corresponding content in the material feeding request data with the candidate material batches in the material requisition and production scheduling data, the candidate material batches are used as the common connection basis, ensuring that the content in the material feeding request data corresponding to the candidate material batches is connected to the content of the same candidate material batches in the material requisition and production scheduling data. When connecting the corresponding content in the material feeding request data with the requisition order in the material requisition and production scheduling data, the correspondence between candidate material batches and requisition order is maintained, ensuring that the corresponding content in the material feeding request data is connected to the content of the same candidate material batches in the material requisition and production scheduling data. The material batch is assigned to the material requisition sequence. When connecting the corresponding content in the material feeding request data with the review candidate content in the material requisition scheduling data, the correspondence between the review candidate content and the candidate material batch is maintained, so that the corresponding content in the material feeding request data is assigned to the review candidate content under the same candidate material batch. When associating the connected corresponding content with the corresponding content in the material batch status data, the candidate material batch is still used as the common association basis, so that the corresponding content in the material feeding request data, the candidate material batch in the material requisition scheduling data, the material requisition sequence, the review candidate content, and the corresponding content in the material batch status data form the same associated content under the same candidate material batch. After the association is completed, the correspondence between the material feeding request data, the material requisition scheduling data, and the material batch status data around the candidate material batch is retained, and material feeding request association data is generated.

[0042] The process involves several steps: First, verifying the consistency between the candidate material batches in the material feeding request association data and the candidate material batches in the material requisition and production scheduling data, while maintaining the connection between the candidate material batches in the material feeding request association data and the corresponding content in the material feeding request data. Second, verifying the consistency between the candidate material batches in the material feeding request association data and the material requisition sequence in the material requisition and production scheduling data, ensuring that the candidate material batches in the material feeding request association data correspond to the material requisition sequence under the same candidate material batch in the material requisition and production scheduling data. During the consistency verification process, it is possible to simultaneously match the candidate material batches and material requisition sequences in the material requisition and production scheduling data. The correspondence between candidate material batches and material requisition order is maintained; corresponding content that fails the consistency check is marked as prohibited material feeding content, and the connection between prohibited material feeding content and corresponding candidate material batches is maintained; corresponding content that passes the consistency check continues to maintain the connection between candidate material batches in the material feeding request association data, corresponding content in the material feeding request data, and material requisition order; after completing the consistency check and marking of prohibited material feeding content, the corresponding content that passes the consistency check and the corresponding content marked as prohibited material feeding content are kept together in the same data result to generate material feeding order verification data.

[0043] The candidate material batches that pass the consistency check in the material feeding sequence verification data are compared with the corresponding contents in the material batch status data. During the status verification process, the connection between the candidate material batches in the material feeding sequence verification data and the corresponding contents in the material feeding request data are maintained, as well as the correspondence between the corresponding contents in the material batch status data and the candidate material batches. Candidate material batches that can form a status verification relationship with the corresponding contents in the material batch status data continue to retain the connection between the candidate material batches already formed in the material feeding sequence verification data, the corresponding contents in the material feeding request data, and the material requisition sequence. The connection relationship is maintained; the corresponding content that fails the status check is marked as prohibited from feeding, and the connection relationship between the prohibited content and the corresponding candidate material batch is maintained; the corresponding content that has been marked as prohibited from feeding in the feeding sequence check data continues to maintain the prohibited content status and does not participate in the connection of candidate material batches that have passed the status check; after completing the status check and the marking of prohibited content, the corresponding content of the candidate material batches that have passed the status check, the corresponding content that has failed the status check and has been marked as prohibited from feeding, and the original prohibited content are all kept in the same data result to generate feeding status check data.

[0044] The candidate material batches that pass the status verification in the material feeding status verification data are compared with the review candidate contents in the material requisition and production scheduling data. During the comparison, the connection relationship between the candidate material batches in the material feeding status verification data and the corresponding contents in the material feeding request data is maintained, as is the correspondence relationship between the review candidate contents and the candidate material batches in the material requisition and production scheduling data. The corresponding contents existing in the review candidate contents are marked as review feeding contents, and the connection relationship between the review feeding contents and the corresponding candidate material batches is maintained. The corresponding contents that pass the status verification and are not marked as prohibited feeding contents or review feeding contents are marked as direct feeding contents, and the connection relationship between the direct feeding contents and the corresponding candidate material batches is maintained. The corresponding contents that have been marked as prohibited feeding contents continue to maintain the prohibited feeding contents status and are distinguished from the direct feeding contents and review feeding contents. After the comparison and marking are completed, the direct feeding contents, review feeding contents, and prohibited feeding contents respectively inherit the connection relationship of the corresponding candidate material batches in the material feeding status verification data, and together form the data result used to represent the material feeding permission status, generating material feeding permission status data.

[0045] In this embodiment, the material conversion record module receives material feeding permit status data and production processing event data. It maintains the distinction between direct material feeding content, verification material feeding content, and prohibited material feeding content in the material feeding permit status data, and maintains the existing correspondence between the corresponding content in the production processing event data and the candidate material batches. It connects the corresponding content in the production processing event data with the direct material feeding content in the material feeding permit status data according to the candidate material batches, so that the content in the production processing event data corresponding to the candidate material batches is linked with the direct material feeding content under the same candidate material batch. It also connects the corresponding content in the production processing event data with the verification material feeding content in the material feeding permit status data according to the candidate material batches, so that the content in the production processing event data corresponding to the candidate material batches is linked with the verification material feeding content under the same candidate material batch. The process involves connecting the corresponding content in the production and processing event data with the prohibited material feeding content in the material feeding permit status data according to candidate material batches. This ensures that the content in the production and processing event data corresponding to a candidate material batch is linked to the prohibited material feeding content within the same candidate material batch. After the connection is established, the direct material feeding content, the reviewed material feeding content, and the prohibited material feeding content maintain their connection relationships with their corresponding content in the production and processing event data. The production and processing event data corresponding to the prohibited material feeding content is removed from the linked corresponding content, while the production and processing event data corresponding to the direct material feeding content and the reviewed material feeding content are retained. During the removal process, the correspondence between the prohibited material feeding content and the candidate material batch is not written into the retained corresponding content, while the correspondence between the direct material feeding content and the reviewed material feeding content and the candidate material batch remains retained, generating convertible event data.

[0046] Arrange the production and processing event data corresponding to the direct material input content in the convertible event data according to the candidate material batch, so that the corresponding content in the direct material input content, candidate material batch, and production and processing event data maintains the same arrangement relationship; arrange the production and processing event data corresponding to the review material input content in the convertible event data according to the candidate material batch, so that the corresponding content in the review material input content, candidate material batch, and production and processing event data maintains the same arrangement relationship; continue to arrange the production and processing event data already arranged by candidate material batch according to production order, so that the candidate material batches under the same production order are retained in the same correspondence relationship; The production processing event data arranged according to the production order is further arranged according to the processing sequence, so that the candidate material batches under the same production order form a sequential relationship according to the processing sequence. After the arrangement is completed, the corresponding content after the arrangement is connected with the event sequence relationship in the production processing event data, so that the candidate material batches, production orders, processing sequence, and event sequence relationship are continuously connected in the corresponding content of the same production processing event data. During the connection process, the production processing event data corresponding to the direct material feeding content and the production processing event data corresponding to the verified material feeding content maintain the original connection relationship with the candidate material batches, generating material conversion sequence data.

[0047] The process involves locating candidate material batches belonging to the same production order within the material conversion sequence data, while maintaining the processing order of these batches. Under the same production order, candidate material batches with earlier processing orders are linked to those with later processing orders, creating a continuous connection along the processing order. During this linking, the correspondence between candidate material batches and production orders in the material conversion sequence data is maintained, as is the connection between candidate material batches and corresponding content in the production processing event data. After linking, the linked candidate material batches are bound to their corresponding content in the production processing event data, ensuring the continuity between candidate material batches, production orders, and processing order. During binding, the corresponding content in the production processing event data remains connected to its corresponding candidate material batch, and the candidate material batch remains connected to its corresponding production order. The processing order continues to define the linking relationship between candidate material batches, generating material conversion association data.

[0048] Candidate material batches from the material conversion association data are written into the same batch association structure, maintaining the existing sequential connections between candidate material batches in the material conversion association data. Production orders from the material conversion association data are written into the same batch association structure, maintaining the correspondence between production orders and candidate material batches. The processing sequence from the material conversion association data is written into the same batch association structure, maintaining the limiting effect of the processing sequence on the sequential connections between candidate material batches. During the writing process, candidate material batches, production orders, and processing sequences jointly inherit the corresponding content in the material conversion association data and continue to maintain the binding relationship with the corresponding content in the production and processing event data. After being written, candidate material batches are used to express batch objects in the same batch association structure, production orders are used to express order correspondence in the same batch association structure, and processing sequences are used to express sequential connections in the same batch association structure. After writing is completed, the same batch association structure simultaneously retains the binding relationships between candidate material batches, production orders, processing sequences, and the corresponding content in the production and processing event data, generating material batch lineage data.

[0049] In this embodiment, the material balance verification module receives material batch lineage data and production output data. First, it maintains the existing correspondence between candidate material batches, production orders, and processing sequences in the material batch lineage data, and also maintains the correspondence between corresponding content in the production output data and production orders. When connecting corresponding content in the production output data with production orders in the material batch lineage data, the production orders are used as the common connection basis, ensuring that content in the production output data corresponding to a production order is connected to content under the same production order in the material batch lineage data. When connecting corresponding content in the production output data with candidate material batches in the material batch lineage data, the correspondence between candidate material batches and production orders is maintained, ensuring that content in the production output data corresponding to a production order is connected to content under the same production order in the material batch lineage data. The content should be able to support candidate material batches under the same production order; when connecting the corresponding content in the production output data with the processing sequence in the material batch lineage data, the limiting effect of the processing sequence on the sequential relationship between candidate material batches should be maintained, so that the corresponding content in the production output data can continue to support the processing sequence under the same production order; during the connection process, the corresponding content in the production output data should not be separated from its corresponding production order, the candidate material batch should not be separated from its corresponding production order, and the processing sequence should not be separated from its corresponding candidate material batch; after the connection is completed, the corresponding content in the production output data, the candidate material batch in the material batch lineage data, the production order, and the processing sequence form the same association around the production order, generating lineage association data.

[0050] Locate candidate material batches belonging to the same production order in the generated bloodline association data, while maintaining the processing order of these batches within the data. When verifying the correspondence between candidate material batches under the same production order and their corresponding content in the production output data, first maintain the correspondence between the candidate material batches and the production order, then maintain the connection between the corresponding content in the production output data and the production order, ensuring that candidate material batches and their corresponding content in the production output data can form a verification relationship under the same production order. During the verification process, the processing order further restricts the verification order between candidate material batches and their corresponding content in the production output data, ensuring that the processing order... Candidate material batches with earlier processing order are first checked against their corresponding contents in the production output data, and candidate material batches with later processing order are checked against their corresponding contents in the production output data. After the corresponding checks are completed, the checked corresponding contents are aggregated according to the candidate material batches, so that the corresponding contents in the production output data of the same candidate material batch are kept in the same aggregation relationship. During the aggregation process, the corresponding contents in the production output data continue to maintain the connection relationship with the production order, and the candidate material batches continue to maintain the connection relationship with the processing order. The checked corresponding contents are not mixed with the corresponding contents of other candidate material batches, thus generating material output check data.

[0051] The candidate material batches, production orders, and corresponding content in the production output data are kept in the same verification relationship, and consumption consistency verification is performed on the corresponding content in the production output data around the candidate material batches. During consumption consistency verification, candidate material batches are used to maintain the batch affiliation of corresponding content in the material output verification data, production orders are used to maintain the order affiliation of corresponding content in the material output verification data, and corresponding content in the production output data is used to inherit the verification results from the generation of lineage association data and the material output verification data. For corresponding content that passes consumption consistency verification, a status flag indicating that consumption consistency verification has been passed is written to that corresponding content, and the connection between that corresponding content and the candidate material batches and production orders is maintained. For content that does not pass consumption consistency verification... By using the corresponding content of the consumption consistency check, the status flag of the failure to pass the consumption consistency check is written into the corresponding content, and the connection relationship between the corresponding content and the candidate material batch and the production order is maintained. During the status flag writing process, the corresponding content of the consumption consistency check and the corresponding content of the failure to pass the consumption consistency check maintain their respective connection relationships with the corresponding content in the candidate material batch, production order and production output data, without changing the candidate material batch aggregation relationship already formed in the material output check data. After the status flag is completed, the corresponding content of the consumption consistency check and the corresponding content of the failure to pass the consumption consistency check jointly take over the corresponding content in the candidate material batch, production order and production output data in the material output check data to generate material balance verification data.

[0052] In this embodiment, the anomaly range location module receives material batch lineage data, material balance verification data, and anomaly material identification data. It first maintains the existing correspondence between candidate material batches, batch association structures, processing order, and production orders in the material batch lineage data, and also maintains the existing correspondence between the corresponding content in the material balance verification data and the candidate material batches and production orders. When connecting the anomaly material identification data to the candidate material batches in the material batch lineage data, the candidate material batches are used as the common connection basis, enabling the anomaly material identification data to connect to the corresponding content under the same candidate material batch in the material batch lineage data. The anomaly material identification data does not deviate from the candidate material batches, and the candidate material batches do not deviate from the batch association structure in the material batch lineage data. The structure maintains the correspondence between the processing sequence and production orders. After the abnormal material identification data is linked with the candidate material batch, the linked corresponding content is checked against the corresponding content in the material balance verification data, so that the corresponding content in the material balance verification data can be accessed under the same candidate material batch and production order. The corresponding content in the material balance verification data continues to maintain the connection relationship with the candidate material batch and production order, and the abnormal material identification data continues to maintain the connection relationship with the candidate material batch. After the verification is completed, the abnormal material identification data, the candidate material batch in the material batch lineage data, the batch association structure, the processing sequence, the production order, and the corresponding content in the material balance verification data form the same connection result around the candidate material batch, generating abnormal lineage association data.

[0053] Candidate material batches that are linked to abnormal material identification data in the abnormal lineage association data are taken as the starting content of the anomaly, and the connection relationship between the starting content of the anomaly and the abnormal material identification data is maintained. When searching for candidate material batches that have a sequential connection relationship with the starting content of the anomaly along the batch association structure in the material batch lineage data, the position of the starting content of the anomaly in the batch association structure is taken as the starting point, and the identity of the starting content of the anomaly as a candidate material batch is maintained. The batch association structure is used to maintain the sequential connection relationship between candidate material batches, the processing order is used to limit the search continuation direction between candidate material batches, and the production order is used to maintain the correspondence between the production orders to which the candidate material batches belong. The search continues along the processing order. During the search, the correspondence between the anomaly initiation content and the processing sequence is maintained first. Then, candidate material batches with sequential relationships to the anomaly initiation content are identified along the processing sequence, ensuring that the found candidate material batches remain within the batch association structure of the material batch lineage data. When searching for production orders, candidate material batches with sequential relationships to the anomaly initiation content are mapped to production orders in the material batch lineage data, ensuring that candidate material batches and production orders follow the same search path. After the search is completed, the anomaly initiation content, candidate material batches with sequential relationships to the anomaly initiation content, production orders, and anomaly material identification data are kept in the same data result, generating anomaly propagation association data.

[0054] The abnormal transmission correlation data is reconciled with the material balance verification data to ensure that the candidate material batches and production orders in the abnormal transmission correlation data can be linked to the corresponding content in the material balance verification data. First, the sequential connection relationship formed along the batch association structure between the abnormal origination content, candidate material batches, and production orders in the abnormal transmission correlation data is maintained. Then, the connection relationship between the corresponding content in the material balance verification data and the candidate material batches and production orders is maintained. When incorporating the corresponding content from the material balance verification data into the abnormal transmission correlation data, the candidate material batches and production orders are used as the common verification basis to ensure that the corresponding content in the material balance verification data can be linked to the content under the same candidate material batch and production order in the abnormal transmission correlation data. When marking content that fails the consumption consistency check in the abnormal transmission association data, the connection between the content that fails the consumption consistency check and the material balance verification data should be maintained, as well as the connection between the content that fails the consumption consistency check and the candidate material batches and production orders in the abnormal transmission association data. During the marking process, the content that fails the consumption consistency check should not be separated from the preceding and following connections in the abnormal transmission association data, and the candidate material batches should not be separated from the corresponding production orders. After the verification and marking are completed, the candidate material batches and production orders that are connected to the content that fails the consumption consistency check in the abnormal transmission association data should be retained in the same data result to generate abnormal finished product candidate data.

[0055] Candidate material batches belonging to the same production order in the abnormal finished product candidate data are grouped according to the processing sequence, maintaining the existing correspondence between candidate material batches and production orders. During the grouping process, the production order is used first as the grouping basis to keep candidate material batches belonging to the same production order within the same grouping range. Then, the processing sequence is used as the sorting basis to maintain the sequential connection between candidate material batches under the same production order. When connecting the grouped corresponding content with the abnormal material identification data, the candidate material batches under the same production order continue to inherit the connection between the abnormal starting content and the abnormal material identification data. In the grouped corresponding content, candidate material batches continue to maintain the connection with the corresponding content in the material balance verification data, production orders continue to maintain the correspondence with candidate material batches, and abnormal material identification data continues to maintain the connection with the abnormal starting content. After completing the grouping and connection, the candidate material batches, processing sequence, abnormal material identification data, and corresponding content in the material balance verification data under the same production order are kept in the same organized result, generating abnormal finished product range organized data.

[0056] The process involves writing production orders from the abnormal finished product range data into the same range positioning result, maintaining the correspondence between production orders and candidate material batches in the abnormal finished product range data; writing candidate material batches from the abnormal finished product range data into the same range positioning result, maintaining the connection between candidate material batches and abnormal material identification data, while also maintaining the sequential connection between candidate material batches formed under the processing sequence; writing abnormal material identification data from the abnormal finished product range data into the same range positioning result, maintaining the connection between abnormal material identification data and the abnormal starting content; writing corresponding content from the material balance verification data into the same range positioning result, maintaining the connection between corresponding content from the material balance verification data and candidate material batches and production orders; during the writing process, the corresponding content in production orders, candidate material batches, abnormal material identification data, and material balance verification data all inherit the correspondence in the abnormal finished product range data, without changing the candidate material batch aggregation relationship already formed in the abnormal finished product candidate data; after the writing is completed, the same range positioning result simultaneously contains the corresponding content in production orders, candidate material batches, abnormal material identification data, and material balance verification data, and forms the affected finished product range data through this same range positioning result.

[0057] In this embodiment, to verify the feasibility of the invention, it is applied to a food processing company that simultaneously produces cakes, rice crackers, and candies. This company is characterized by parallel production of multiple product categories, numerous batches of raw materials and auxiliary materials, strict formula constraints, frequent material requisitions and feeding, and difficulties in tracing abnormal batches. The production site involves materials such as flour, egg liquid, cream, syrup, glutinous rice flour, puffed base material, chopped nuts, fruit-flavored syrup, gelatin, and food coloring. The daily production order volume is 356 orders, with 1680 candidate material batches. Inventory locations are distributed in ambient temperature warehouses, cold storage warehouses, auxiliary material temporary storage areas, and workshop temporary storage areas. The original material management method mainly relies on inventory records, manual formula verification, and material requisition form review, which can complete basic inbound and outbound management. However, when multiple batches of the same raw material exist, the same batch is distributed in multiple inventory locations, mandatory and optimized attributes coexist in the formula, and temporary material feeding requests increase on-site, problems easily arise such as inconsistent batch status, unclear formula matching boundaries, disconnect between material requisition scheduling and material feeding verification, and difficulty in closing the gap between processing records and output results. To solve these problems, enterprises can integrate the food production and processing material control system of this invention into the inventory collection terminal, formula management terminal, production order terminal, material feeding request terminal, production and processing event recording terminal, and production output recording terminal.

[0058] During system operation, the material batch status generation module receives basic material data and inventory collection data from food production and processing. Using material batch identifiers as the connection basis, it aggregates basic material data and inventory collection data under the same material batch identifier into the same batch relationship, and performs consistency checks on inventory location, quantity, and status to generate material batch status data. Before system integration, the company sampled 1680 candidate material batches and found 137 batches with inconsistencies in inventory location, quantity, or status, resulting in a batch status consistency rate of 91.85%. After system integration, only 19 batches of the same size required manual verification, increasing the batch status consistency rate to 98.87%. In cake production, egg liquid and cream need to be kept refrigerated. The system can simultaneously write the refrigerated inventory location, quantity, and status into the material batch status data, avoiding the recommendation of unsuitable candidate material batches based solely on total inventory quantity.

[0059] After receiving material batch status data and product formula constraint data, the formula constraint matching module concatenates candidate material batches with corresponding content in the product formula constraint data to generate a candidate material batch evaluation matrix. This matrix is ​​then processed through attribute direction unification, normalization, and weighting to form a weighted normalized matrix for candidate material batches. The system divides candidate material batches into available and prohibited candidate domains based on mandatory attributes. Then, through available and prohibited domain description processing based on support vector data description algorithms, it generates approximate regions for available and prohibited boundaries, calculates the distance between available and prohibited boundaries, and the distance to mandatory attribute locking, ultimately generating direct matching ranking content, verification matching content, and prohibited matching content. In continuous production verification, the company used 21 sets of product formula constraint data, accumulating a total of 48,620 rows of weighted normalized matrix for candidate material batches. Before the system was implemented, the manual review and rejection rate for recipe matching was 12.8%, and the number of batches recommended for material requisition despite not meeting mandatory attributes was 26 per thousand batches. After the system was implemented, the review and rejection rate dropped to 3.1%, and the number of incorrect recommendations for mandatory attributes dropped to 3 per thousand batches. For batches of fruit-flavored syrups in candy production that meet the color requirements but not the shelf life requirements, the system classifies these batches into prohibited matching content through mandatory attribute locking distance, avoiding incorrect material requisition due to high optimization attribute scores.

[0060] The material requisition and scheduling generation module receives material formula matching data and production order data. It extends the directly matched sorting content, the verified matching content, and the prohibited matching content into the production order data, generating order material association data. The system removes candidate material batches corresponding to prohibited matching content, forming order material requisition data, and generates candidate material requisition order data based on the directly matched sorting content and the verified matching content, ultimately generating material requisition and scheduling data. Previously, the average time from material confirmation to material requisition result for a single production order was 26 minutes, with 64 modifications to the material requisition form per day. After the system was implemented, the average material requisition and scheduling time for a single production order decreased to 9 minutes, the number of modifications to the material requisition form per day decreased to 15, and the first-pass yield rate for material requisition and scheduling increased from 82.5% to 95.7%. For scenarios where both cake base and frosting production are required under the same production order, the system can determine the priority material requisition batches based on the directly matched sorting content and retain verified candidate content, facilitating verification before material requisition.

[0061] After receiving material requisition and production scheduling data, material batch status data, and material feeding request data, the material feeding request data is linked with the candidate material batches, requisition order, and review candidate content in the material requisition and production scheduling data. It then combines this with the material batch status data to verify the material feeding order and status, generating material feeding permission status data. During continuous operation, the system processed 21,750 material feeding request data entries, including 19,862 direct feeding entries, 1,344 review feeding entries, and 544 prohibited feeding entries. Before system operation, the number of instances where non-compliant batches were found after material requisition was 39 per 10,000 feeding attempts; after system operation, this decreased to 6 per 10,000 feeding attempts, and the average verification time for material feeding requests decreased from 220 seconds to 27 seconds. In cases where puffed base material has entered the material requisition and production scheduling data but its inventory status has temporarily changed during rice cracker production, the system can mark the corresponding material feeding request as prohibited feeding content based on the material batch status data, reducing the risk of incorrect on-site feeding.

[0062] After receiving material feeding permit status data and production processing event data, the material conversion record module removes the production processing event data corresponding to prohibited material feeding content. It then arranges the production processing event data corresponding to directly fed and verified material feeding content according to candidate material batches, production orders, and processing sequence, and connects them with the pre- and post-event relationships to generate material batch lineage data. Before the system was implemented, tracing a single abnormal batch took an average of 47 minutes, requiring manual searching of material requisition forms, feeding records, production logs, and output sheets. After the system was implemented, the average time for tracing a single batch decreased to 5 minutes, and the batch lineage record completeness rate increased from 86.2% to 99.1%. When multiple cream batches are added sequentially to the whipping and decorating processes in a cake order, the system can retain the continuous relationship between candidate material batches, production orders, and processing sequence, providing a basis for subsequent material balance verification.

[0063] After receiving material batch lineage data and production output data, the material balance verification module connects the production output data with candidate material batches, production orders, and processing sequence to generate lineage correlation data. It then performs consumption consistency checks on candidate material batches to generate material balance verification data. The system has verified a total of 9860 production orders, identifying 217 orders with discrepancies between material consumption and production output. These discrepancies included 104 orders with missing material input records, 78 orders with delayed output data registration, and 35 orders with mixed candidate material batches. Before system implementation, the average delay in detecting consumption anomalies was 6.5 production shifts; after system implementation, this decreased to 0.8 production shifts, and the material consumption verification accuracy improved from 88.7% to 97.6%.

[0064] After receiving material batch lineage data, material balance verification data, and abnormal material identification data, the anomaly range localization module connects the abnormal material identification data with candidate material batches and searches for associated candidate material batches and production orders along the batch association structure and processing sequence. It then combines this with the material balance verification data to generate affected finished product range data. The system verified 12 instances of abnormal material identification data, generating affected finished product range data in an average of 200 seconds, while manual methods took an average of 112 minutes. The system identified 139 affected production orders, of which 134 were confirmed through manual review, resulting in a range localization accuracy of 96.4%. Manual methods typically expanded the affected product range to an average of 7200 finished products per shift, while the system-generated affected finished product range data averaged 1300 finished products per shift, representing a range compression ratio of 81.9%. This embodiment demonstrates that the present invention can improve material batch status consistency, material formula matching and identification accuracy, material requisition and production scheduling stability, material input verification accuracy, material batch traceability integrity, material consumption verification accuracy, and anomaly range localization capabilities.

[0065] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A material control system for food production and processing, characterized in that: include: The material batch status generation module is used to receive basic material data and inventory collection data in food production and processing, perform batch status processing on the basic material data and inventory collection data, and generate material batch status data. The formula constraint matching module is used to receive material batch status data and product formula constraint data, perform material formula boundary approximate matching processing based on the MABAC algorithm on the material batch status data and product formula constraint data, and generate material formula matching data. The material requisition and scheduling generation module is used to receive material formula matching data and production order data, perform material requisition and scheduling generation processing on the material formula matching data and production order data, and generate material requisition and scheduling data. The material feeding risk verification module is used to receive material requisition and production scheduling data, material batch status data and material feeding request data, perform material feeding risk verification processing on the material requisition and production scheduling data, material batch status data and material feeding request data, and generate material feeding permission status data. The material conversion record module is used to receive material feeding permit status data and production processing event data, perform material conversion association processing on the material feeding permit status data and production processing event data, and generate material batch lineage data; The material balance verification module is used to receive material batch lineage data and production output data, perform material consumption consistency verification processing on the material batch lineage data and production output data, and generate material balance verification data. The anomaly range positioning module is used to receive material batch lineage data, material balance verification data, and abnormal material identification data, and to perform abnormal finished product range positioning processing on the material batch lineage data, material balance verification data, and abnormal material identification data to generate affected finished product range data.

2. The food production and processing material control system according to claim 1, characterized in that: The material batch status generation module receives basic material data and inventory collection data from food production and processing. It collects the corresponding content in the basic material data according to the material batch identifier and connects the corresponding content in the inventory collection data to the collected content according to the same material batch identifier, thereby generating material inventory association data. Perform consistency checks on the corresponding contents of the same material batch identifier in the material inventory association data, and organize the corresponding contents that pass the consistency check according to the inventory location, inventory quantity and inventory status to generate material batch organization data; Connect the material batch sorting data with the corresponding content in the material basic data, and then arrange the connected corresponding content in a unified manner according to the material batch identifier, inventory location, inventory quantity and inventory status to generate material batch status data.

3. The food production and processing material control system according to claim 1, characterized in that: The formula constraint matching module receives material batch status data and product formula constraint data, and connects the candidate material batches in the material batch status data with the corresponding contents in the product formula constraint data according to the candidate material batches to generate a candidate material batch evaluation matrix. The evaluation matrix of candidate material batches is processed to unify the attribute direction, converting positive and negative attributes into the same evaluation direction, generating a unified matrix of candidate material batches. The unified matrix of candidate material batches is then normalized to generate a normalized matrix of candidate material batches. The weights of the evaluation attributes are determined based on the product formula constraint data, and the normalized matrix of candidate material batches is weighted to generate a weighted normalized matrix of candidate material batches. Based on the mandatory attributes in the product formula constraint data, the candidate material batches corresponding to the weighted normalization matrix of the candidate material batches are divided into available candidate domains and prohibited candidate domains. The available candidate domain is processed by the available domain description algorithm based on the support vector data description algorithm to generate an approximate region of available boundaries. The corresponding content in the forbidden candidate domain is processed by the forbidden domain description algorithm based on the support vector data description algorithm to generate an approximate region of forbidden boundaries. Calculate the available boundary distance of the candidate material batch relative to the approximate area of ​​the available boundary, and calculate the prohibited boundary distance of the candidate material batch relative to the approximate area of ​​the prohibited boundary; Determine the distance content corresponding to the mandatory attribute from the available boundary distances, and generate the mandatory attribute locking distance based on the distance content corresponding to the mandatory attribute; The corresponding content in the product formula constraint data, excluding mandatory attributes and used for comprehensive ranking, is identified as the optimization attribute; Based on the mandatory attribute locking distance control optimization attribute corresponding to the available boundary distance, the comprehensive sorting is used to generate the comprehensive matching distance of candidate material batches; Candidate material batches are classified and processed according to the comprehensive matching distance of the candidate material batches, generating direct matching sorting content, verification matching content, and prohibited matching content; The directly matched sorted content, the verified matched content, and the prohibited matched content are merged to generate material formula matching data.

4. The food production and processing material control system according to claim 3, characterized in that: Available domain description processing and forbidden domain description processing based on support vector data description algorithms include: Read the rows of the candidate material batch weighted normalized matrix corresponding to the available candidate domain from the candidate material batch weighted normalized matrix to generate the available domain training matrix; read the rows of the candidate material batch weighted normalized matrix corresponding to the prohibited candidate domain from the candidate material batch weighted normalized matrix to generate the prohibited domain training matrix. Each row in the available domain training matrix is ​​used as an available domain sample vector, and each row in the forbidden domain training matrix is ​​used as a forbidden domain sample vector, while maintaining the correspondence between the available domain sample vector, the forbidden domain sample vector and the candidate material batch; Perform availability domain envelope modeling on availability domain sample vectors to determine the availability domain center and availability domain envelope radius; perform forbidden domain envelope modeling on forbidden domain sample vectors to determine the forbidden domain center and forbidden domain envelope radius. Based on the distance from the available domain sample vector to the center of the available domain, the available domain support vectors located on the boundary of the available domain envelope are filtered, and the available domain envelope boundary data is generated based on the center of the available domain, the radius of the available domain envelope, and the available domain support vectors. Based on the distance from the forbidden region sample vector to the center of the forbidden region, the forbidden region support vectors located on the forbidden region envelope boundary are selected, and forbidden region envelope boundary data is generated based on the forbidden region center, forbidden region envelope radius, and forbidden region support vectors. Based on the available domain envelope boundary data, determine the available domain boundary position of the candidate material batch in the weighted normalized attribute space, and write the available domain boundary position into the available boundary approximation region. Based on the forbidden zone envelope boundary data, determine the forbidden zone boundary position of the candidate material batch in the weighted normalized attribute space, and write the forbidden zone boundary position into the forbidden zone approximate region. Calculate the available boundary distance of the candidate material batch relative to the approximate area of ​​the available boundary, and calculate the prohibited boundary distance of the candidate material batch relative to the approximate area of ​​the prohibited boundary.

5. A food production and processing material control system according to claim 3, characterized in that: The material requisition and scheduling generation module receives material formula matching data and production order data. It connects the corresponding content in the production order data with the direct matching sorting content, verification matching content and prohibited matching content in the material formula matching data according to the candidate material batches to generate order material association data. The candidate material batches in the order material association data are compared with the prohibited matching content, and the candidate material batches that exist in the prohibited matching content are removed from the order material association data to generate order material requisition data; Arrange the candidate material batches in the order material requisition data according to the sorting relationship in the direct matching sorting content, and mark the candidate material batches that exist in the review matching content as review candidate content, and generate candidate material requisition order data; The candidate material batches in the candidate material requisition sequence data are matched with the production sequence in the production order data, and the candidate material batches belonging to the same production order are aggregated according to the production sequence to generate material requisition and production scheduling data that includes production order, candidate material batches, material requisition sequence and review candidate content.

6. The food production and processing material control system according to claim 5, characterized in that: The material feeding risk verification module receives material requisition and production scheduling data, material batch status data, and material feeding request data. It connects the corresponding content in the material feeding request data with the candidate material batches, material requisition order, and review candidate content in the material requisition and production scheduling data according to the candidate material batches. It then associates the connected corresponding content with the corresponding content in the material batch status data to generate material feeding request association data. The candidate material batches in the material feeding request associated data are checked for consistency with the candidate material batches and material feeding order in the material requisition and production scheduling data. The corresponding contents that fail the consistency check are marked as prohibited material feeding contents, and material feeding order check data is generated. The candidate material batches that pass the consistency check in the feeding sequence check data are checked against the corresponding contents in the material batch status data, and the corresponding contents that fail the status check are marked as prohibited feeding contents, thus generating feeding status check data. The candidate material batches that pass the status check in the material feeding status check data are compared with the review candidate contents in the material requisition and production scheduling data. The corresponding contents existing in the review candidate contents are marked as review feeding contents. The corresponding contents that pass the status check and are not marked as prohibited feeding contents or review feeding contents are marked as direct feeding contents, thus generating material feeding permit status data.

7. A food production and processing material control system according to claim 6, characterized in that: The material conversion record module receives material feeding permit status data and production processing event data. It connects the corresponding content in the production processing event data with the direct material feeding content, verification material feeding content, and prohibited material feeding content in the material feeding permit status data according to the candidate material batch. It also removes the production processing event data corresponding to the prohibited material feeding content from the corresponding content after connection to generate convertible event data. The production and processing event data corresponding to the direct feeding content and the verified feeding content in the convertible event data are arranged according to the candidate material batch, production order and processing order. The corresponding content after arrangement is connected with the event sequence in the production and processing event data to generate material conversion order data. Candidate material batches belonging to the same production order in the material conversion sequence data are connected sequentially according to the processing order, and the connected candidate material batches are bound to the corresponding content in the production processing event data to generate material conversion association data. The candidate material batches, production orders, and processing order in the material conversion association data are written into the same batch association structure to generate material batch lineage data.

8. A food production and processing material control system according to claim 7, characterized in that: The material balance verification module receives material batch lineage data and production output data, and connects the corresponding content in the production output data with the candidate material batches, production orders and processing order in the material batch lineage data according to the production order to generate production lineage association data. The candidate material batches belonging to the same production order in the generated bloodline association data are matched with the corresponding contents in the production output data according to the processing order. The matched contents are then collected according to the candidate material batches to generate material output verification data. The consumption consistency of candidate material batches, production orders, and corresponding contents in production output data in the material output verification data is checked. The corresponding contents that pass the consumption consistency check and the corresponding contents that fail the consumption consistency check are marked with status to generate material balance verification data.

9. A food production and processing material control system according to claim 8, characterized in that: The anomaly range location module receives material batch lineage data, material balance verification data, and abnormal material identification data. It connects the abnormal material identification data with the candidate material batches in the material batch lineage data according to the candidate material batches, and then checks the corresponding content after connection with the corresponding content in the material balance verification data to generate abnormal lineage association data. The candidate material batches that are connected to the abnormal material identification data in the abnormal lineage association data are taken as the abnormal starting content. The candidate material batches and production orders that have a sequential connection with the abnormal starting content are searched along the batch association structure in the material batch lineage data according to the processing order, and abnormal transmission association data is generated. The abnormal transmission correlation data is compared with the material balance verification data, and the corresponding content that fails the consumption consistency verification is marked in the abnormal transmission correlation data to generate abnormal finished product candidate data. The candidate material batches belonging to the same production order in the abnormal finished product candidate data are grouped according to the processing order, and the corresponding content after grouping is linked with the abnormal material identification data to generate abnormal finished product range sorting data. Write the corresponding contents from the production orders, candidate material batches, abnormal material identification data, and material balance verification data in the abnormal finished product range data into the same range location result to generate the affected finished product range data.

Images