A method for producing a soy protein-dietary fiber complex food product

Abstract

The present application relates to the technical field of food production traceability, and discloses a soybean protein-dietary fiber compound food production traceability method, which comprises the following steps: step 1, generating a batch number and calculating a standardized proportion index to form a raw material event; step 2, setting a window matrix containing pH, proportion, temperature and time based on a process target and generating a window signature; step 3, generating a process trajectory abstract and forming a compliance judgment to obtain a process event; step 4, generating a fingerprint ternary vector and a tolerance band thereof to form a content event; step 5, generating a double-anchor coupled code and forming a chain event through chain connection; step 6, generating a two-dimensional carrier code and issuing an offline silent verification rule; and step 7, verifying data consistency and determining an abnormal minimum cut time period when an exception occurs. The present application realizes traceable management of the whole production process of soybean protein-dietary fiber compound food and accurate positioning of an abnormal time period.

Description

Technical Field

[0001] This invention belongs to the field of food production traceability technology, specifically relating to a method for tracing the production of soybean protein-dietary fiber complex food. Background Technology

[0002] Soy protein-dietary fiber complex foods have gradually become an important product type in the food industry due to their high value in nutritional structure, functional characteristics, and application scenarios. In the production process of this type of food, there is a strong correlation between raw material ratios, process parameters, and product characteristics. The production process typically involves multiple batches of raw material input, multi-stage process control, and multi-index testing. With the expansion of production scale and the extension of the product distribution chain, how to effectively trace the entire production process has become a key issue for enterprises in quality management and process control.

[0003] Current food production traceability methods often focus on batch identification records or simple correlations of key node information, typically using production time, production line number, or test results as traceability basis. In practical applications, these methods often fail to fully reflect the inherent correspondence between raw material ratios, process objectives, process execution status, and product characteristics. This is especially true when quality anomalies or disputes occur, making it difficult to accurately pinpoint the specific stage and timeframe of the anomaly. Furthermore, some traceability systems rely on online systems for real-time verification; in offline environments or on-site inspection scenarios, data consistency verification capabilities are limited, affecting the reliability of traceability results. In addition, during the production of soybean protein-dietary fiber complex foods, process parameters fluctuate within a certain range, and product physicochemical indicators also exhibit natural dispersion. Simply judging based on whether standards are exceeded easily overlooks the comprehensive correlation between process and content, making it difficult to generate refined and recalculated anomaly analysis results. Summary of the Invention

[0004] This invention provides a method for tracing the production of soybean protein-dietary fiber complex food, which solves the technical problems in related technologies where raw material ratio information, process constraint information, process execution data and product detection data are recorded in a scattered manner and lack a unified correlation mechanism during the production of multiple batches of food, making it difficult to verify the consistency of production data and accurately locate the time period of anomalies.

[0005] This invention provides a method for traceability in the production of soybean protein-dietary fiber complex foods, comprising the following steps:

[0006] Step 1: Collect batch serial numbers to generate batch numbers, collect dry-based soybean protein and dietary fiber quality to generate standardized ratio indices, and form raw material events;

[0007] Step 2: Based on the process objectives of this batch, set a window matrix including pH, standardized ratio index, temperature and time, and generate window signature;

[0008] Step 3: Collect pH, temperature and time interval data to generate a process trajectory summary, generate compliance judgments based on standardized ratio index and window matrix, and form process events;

[0009] Step 4: Collect nitrogen solubility index, product water retention capacity and product isokinetic viscosity to generate a fingerprint ternary vector and the tolerance band for each component of the fingerprint ternary vector to form a content event;

[0010] Step 5: Generate a dual-anchor coupling code and a check bit based on the batch number, window signature, process trajectory summary, and fingerprint ternary vector to form a coupling event, and chain the coupling events into a chain event;

[0011] Step 6: Generate a two-dimensional carrier code based on the chained events and issue offline silent verification rules. The offline silent verification rules include: reading the two-dimensional carrier code and obtaining the window signature based on the chained event index, recalculating the verification bits based on the two-dimensional carrier code and the window signature, and comparing them.

[0012] Step 7: Retest the fingerprint ternary vector and recalculate the dual-anchor coupling code and check bit, and compare it with the chain event to form a verification event; when the verification event indicates an anomaly, construct a constraint set including compliance judgment and tolerance zone judgment, determine the minimum cut time period of the anomaly, and form an anomaly event.

[0013] The beneficial effects of this invention are as follows: The soybean protein-dietary fiber complex food production traceability method provided by this invention achieves unified identification and associated storage of key information throughout the entire production process by structurally managing raw material events, window events, process events, content events, and their chain-like relationships. This allows data generated at different stages to be consistently referenced and verified within the same batch dimension. By solidifying process objectives and tolerances in the form of a window matrix and coupling them with process trajectory summaries and fingerprint ternary vectors, the correspondence between production process states and product characteristics is clearly expressed, avoiding the information fragmentation problem caused by relying solely on single detection results for traceability. Furthermore, this invention continuously binds multi-stage data through dual-anchor coupling codes and chain-like event structures, improving the integrity and verifiability of traceability data in the time series. Combined with two-dimensional code carriers and offline silent verification rules, traceability data can still be verified for consistency in non-real-time network environments, enhancing the flexibility of production management and on-site verification. When an anomaly occurs, this invention constructs a constraint set based on verification results, compliance judgment, and tolerance zone judgment, and determines the minimum cut time period of the anomaly through a chain-like event timestamp sequence, extending anomaly localization from result judgment to process analysis, thereby improving the level of precision in food production traceability at the management decision-making and anomaly analysis levels. Attached Figure Description

[0014] Figure 1 This is a flowchart of a method for tracing the production of soybean protein-dietary fiber complex food according to the present invention. Detailed Implementation

[0015] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0016] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of the present invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in one or more embodiments of the present invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0017] like Figure 1 As shown, a method for traceability in the production of soybean protein-dietary fiber complex food includes the following steps:

[0018] Step 1: Collect batch serial numbers to generate batch numbers, collect dry-based soybean protein and dietary fiber quality to generate standardized ratio indices, and form raw material events;

[0019] Step 2: Based on the process objectives of this batch, set a window matrix including pH, standardized ratio index, temperature and time, and generate window signature;

[0020] Step 3: Collect pH, temperature and time interval data to generate a process trajectory summary, generate compliance judgments based on standardized ratio index and window matrix, and form process events;

[0021] Step 4: Collect nitrogen solubility index, product water retention capacity and product isokinetic viscosity to generate a fingerprint ternary vector and the tolerance band for each component of the fingerprint ternary vector to form a content event;

[0022] Step 5: Generate a dual-anchor coupling code and a check bit based on the batch number, window signature, process trajectory summary, and fingerprint ternary vector to form a coupling event, and chain the coupling events into a chain event;

[0023] Step 6: Generate a two-dimensional carrier code based on the chained events and issue offline silent verification rules. The offline silent verification rules include: reading the two-dimensional carrier code and obtaining the window signature based on the chained event index, recalculating the verification bits based on the two-dimensional carrier code and the window signature, and comparing them.

[0024] Step 7: Retest the fingerprint ternary vector and recalculate the dual-anchor coupling code and check bit, and compare it with the chain event to form a verification event; when the verification event indicates an anomaly, construct a constraint set including compliance judgment and tolerance zone judgment, determine the minimum cut time period of the anomaly, and form an anomaly event.

[0025] In one embodiment of the present invention, a batch number is generated by collecting batch serial numbers, and a standardized ratio index is generated by collecting dry-based soybean protein content and dietary fiber quality, forming a raw material event, including:

[0026] Step 11: Collect batch serial numbers used to identify the production sequence. These batch serial numbers are numerical strings representing the order of production and are used to distinguish different production batches. After collection, the batch serial numbers undergo a numerical string format verification to ensure they contain only numeric characters and meet a preset length requirement. Simultaneously, the currently collected batch serial number is compared with the batch serial numbers in existing raw material events to complete a uniqueness verification, thereby preventing the same batch serial number from being reused and ensuring the uniqueness and determinism of batch identification in the production traceability chain.

[0027] Step 12: After completing the batch serial number verification, generate a batch number based on the batch serial number. The batch number serves as a unified index identifier for this batch throughout the entire production traceability process. The batch number consists of the batch serial number and a verification field calculated using a verification algorithm based on the batch serial number. The verification field is used to verify the integrity and consistency of the batch number to prevent batch identification errors caused by miswriting or tampering during data storage or transmission. The generated batch number is written into the raw material event and used as a unified reference key in subsequent window events, process events, content events, and chained events.

[0028] Step 13: Collect the dry-based soybean protein and dietary fiber mass used in this batch of production. The dry-based soybean protein mass refers to the mass of soybean protein raw material after moisture removal, and the dietary fiber mass refers to the mass of dietary fiber raw material involved in the formation of the complex. To ensure the certainty of subsequent calculations, the dry-based soybean protein and dietary fiber mass are both checked for values ​​greater than zero to ensure that the collected mass data are valid. After passing the check, the dry-based soybean protein and dietary fiber mass are added together as the denominator, and the dry-based soybean protein mass is used as the numerator. The ratio of the numerator to the denominator is calculated to obtain the standardized proportion index.

[0029] The standardized proportioning index is used to characterize the relative proportions of dry-based soybean protein in the soybean protein-dietary fiber complex system. By converting the raw material proportions into a standardized proportioning index, different batches can still use a unified proportioning representation even with varying total raw material amounts, thus facilitating consistent reference in subsequent process constraint settings, compliance assessments, and anomaly verification. After calculation, the standardized proportioning index is written into the raw material event, forming the core data content of the raw material event together with the batch number.

[0030] Through the above methods, this invention standardizes batch identification, raw material quality data, and proportioning relationships at the initial stage of the production traceability process, and stores them in a structured manner as raw material events. On the one hand, this ensures the uniqueness and verifiability of batch identification throughout the entire traceability chain; on the other hand, it achieves a unified expression of raw material proportioning data through standardized proportioning indices, providing a stable data foundation for subsequent traceability management based on process windows, process trajectories, and content fingerprints, thereby improving the controllability and consistency of the soybean protein-dietary fiber complex food production process in terms of information management and traceability control.

[0031] In one embodiment of the present invention, based on the process objectives of this batch, a window matrix including pH, standardized ratio index, temperature, and time is set and a window signature is generated, including:

[0032] Step 21: Obtain the batch number and standardized proportion index corresponding to this batch from the raw material event. When generating the window matrix, use the batch number as the batch identifier of the window matrix, and write the standardized proportion index into the proportion-related position in the window matrix to ensure a unique correspondence between the window matrix and the raw material event at the batch level.

[0033] The window matrix is ​​a data structure used to describe the process constraints of this batch. Its rows correspond to four types of constraint objects: pH, standardized proportioning index, temperature, and time, used to distinguish the constraint dimensions of different process parameters. Its columns include target information columns and tolerance information columns, used to describe the target values ​​and allowable deviation ranges of each process parameter, respectively. Matrix elements are numerical values ​​corresponding to the target information or tolerance information. Specifically, the target information column corresponds to the target fields for pH, standardized proportioning index, temperature, and time, used to store the expected target values ​​of the process parameters for this batch; the tolerance information column corresponds to the tolerance fields for pH, standardized proportioning index, temperature, and time, used to store the allowable deviation ranges corresponding to each target value. Through this row and column structure, the targets and tolerances of different process parameters are represented in a one-to-one correspondence within the same matrix, thereby avoiding confusion between parameters.

[0034] Step 22: After initializing the window matrix structure, obtain the process objectives for this batch. These process objectives refer to the preset process control requirements for the production of this batch of soybean protein-dietary fiber complex food. Specifically, they include target information and corresponding tolerance information for pH, target information and corresponding tolerance information for standardized ratio index, target information and corresponding tolerance information for temperature, and target information and corresponding tolerance information for time. Write the above target information and corresponding tolerance information into the corresponding target and tolerance fields in the window matrix, so that the window matrix fully reflects the target status and allowable deviations of this batch in each process dimension.

[0035] Step 23: After the window matrix is ​​constructed, it is serialized according to a preset fixed field order. This fixed field order ensures consistent serialization results when processing the window matrix at different times, on different devices, or on different nodes. Subsequently, a hash calculation is performed on the serialization result to obtain the window signature. The window signature is a unique checksum determined by the window matrix and is used to verify the consistency of the window matrix content in subsequent processes. Finally, the window matrix and its corresponding window signature together form a window event, which is then stored in association with the batch number.

[0036] Through the above method, this invention structurally solidifies the process objectives and tolerance ranges of this batch in the form of a window matrix, and achieves verifiable identification of process constraint information through window signatures. On the one hand, it enables the management of multi-dimensional process parameters involved in food production under a unified data structure; on the other hand, it provides a reliable process constraint benchmark for subsequent production traceability based on process trajectory summaries, fingerprint ternary vectors, and chain events, thereby improving the traceability and consistency of the soybean protein-dietary fiber complex food production process at the information management level.

[0037] In one embodiment of the present invention, a process trajectory summary is generated by collecting pH, temperature, and time interval data. Compliance judgments are then generated based on a standardized ratio index and a window matrix, forming process events, including:

[0038] Step 31: Obtain the batch number and standardized proportion index corresponding to this batch from the raw material events, and obtain the window events associated with this batch based on the batch number. Read the window matrix and window signature from the window events. This method ensures that the collection and judgment of process data are always based on the established process objectives and tolerance ranges, and maintains a one-to-one correspondence between the data and window events.

[0039] Step 32: After obtaining the process constraint baseline, the pH, temperature, and time intervals generated during the production process are collected to form a collection record sequence. To avoid interference from instantaneous fluctuations in the process description, the pH values ​​collected in each step are weighted with their corresponding time intervals. This is done by summing the products of each pH value and its corresponding time interval, then dividing by the sum of all time intervals to obtain a weighted average pH value. Similarly, the products of each temperature value and its corresponding time interval are summed, then divided by the sum of all time intervals to obtain a weighted average temperature value. Simultaneously, all time intervals are accumulated to obtain the total time reflecting the duration of the process in this batch. The weighted average pH value, the weighted average temperature value, and the total time together constitute a process trajectory summary. The process trajectory summary is a compressed representation of the overall process state, used to accurately reflect the key characteristics of the process without retaining all original collected data, thus facilitating subsequent traceability, verification, and anomaly location.

[0040] Step 33: After generating the process trajectory summary, a compliance determination is made on the process trajectory summary and the standardized proportioning index formed in the raw material stage based on the window matrix. Specifically, the weighted average value of pH in the process trajectory summary is compared with the target pH field and pH tolerance field in the window matrix to determine whether the pH falls within the range defined by the target value and tolerance, thus obtaining a pH compliance determination; the weighted average value of temperature is compared with the target temperature field and temperature tolerance field to obtain a temperature compliance determination; the total time is compared with the target time field and time tolerance field to obtain a time compliance determination; simultaneously, the standardized proportioning index is compared with the target standardized proportioning index field and standardized proportioning index tolerance field in the window matrix to obtain a proportioning compliance determination. The boundary comparison refers to determining whether the actual value is within the allowable range defined by the target information as the center and the tolerance information as the boundary. The above pH compliance determination, temperature compliance determination, time compliance determination, and proportioning compliance determination are summarized to obtain the compliance determination result. Finally, the batch number, process trace summary, compliance judgment, and window signature are encapsulated together to form a process event, which is then stored in association with the batch number.

[0041] Through the above methods, this invention transforms continuously and repeatedly collected process parameters during production into structured process trajectory summaries, and completes compliance determination under a unified window matrix constraint. On the one hand, it reduces the complexity of storing and processing production process data in the traceability system; on the other hand, it enables the recording and traceability of whether the production process meets the predetermined process objectives in a deterministic and recalcible manner, providing reliable process-side evidence for subsequent verification and anomaly analysis based on fingerprint ternary vectors, dual-anchor coupling codes, and chain events, thereby enhancing the technical effectiveness of this invention in business rule management and process consistency control.

[0042] In one embodiment of the present invention, a fingerprint ternary vector is generated by collecting nitrogen solubility index, product water retention capacity, and product isokinetic viscosity, and the tolerance bands for each component of the fingerprint ternary vector are used to form a content event, including:

[0043] Step 41: Obtain the batch number corresponding to this batch from the raw material event, and test the produced product based on the batch number, collecting the nitrogen solubility index, product water retention capacity, and product isokinetic viscosity. The nitrogen solubility index is used to characterize the solubility characteristics of proteins in the aqueous phase, the product water retention capacity is used to characterize the product's ability to retain water, and the product isokinetic viscosity is used to characterize the rheological properties of the product under specific shear conditions. The collected nitrogen solubility index, product water retention capacity, and product isokinetic viscosity are associated and stored with the batch number, so that the test data can be matched one-to-one with the specific production batch in the traceability system.

[0044] Step 42: After completing the content detection data acquisition, the nitrogen solubility index, product water retention capacity, and product isokinetic viscosity are processed according to preset numerical solidification rules to generate a fingerprint ternary vector. The numerical solidification rules refer to rules for converting continuous detection values ​​into a uniform precision and representation format. For example, by retaining a fixed number of precision digits or using a preset numerical range mapping method, the impact of differences in detection equipment precision or minor numerical fluctuations on subsequent comparisons can be avoided. After processing, the nitrogen solubility index corresponds to the first component of the fingerprint ternary vector, the product water retention capacity corresponds to the second component, and the product isokinetic viscosity corresponds to the third component, thus obtaining a fingerprint ternary vector used to characterize the features of this batch of products.

[0045] Step 43: After forming the fingerprint ternary vector, further read the pre-set tolerance thresholds for each component of the fingerprint ternary vector. These tolerance thresholds limit the allowable fluctuation range of each fingerprint component under normal production conditions. Using the value of each component in the fingerprint ternary vector as the center value and the corresponding tolerance threshold as the offset, a tolerance band for each component of the fingerprint ternary vector is generated. This tolerance band describes the allowable value range of each fingerprint component and is an important basis for boundary comparison and anomaly detection in subsequent verification stages. The batch number, the fingerprint ternary vector, and the tolerance bands for each component of the fingerprint ternary vector are encapsulated together to form a content event and stored in association with the batch number.

[0046] Through the above method, this invention transforms multidimensional detection data at the product level into a structured fingerprint ternary vector and clearly defines the normal fluctuation range through tolerance bands. On the one hand, this enables the content characteristics of different batches of products to be compared and traced under a unified data structure; on the other hand, it provides a stable content-side judgment basis for subsequent production traceability based on dual-anchor coupling codes, verification events, and minimum cut time periods for anomalies.

[0047] In one embodiment of the present invention, a dual-anchor coupling code and a check bit are generated based on the batch number, window signature, process trajectory summary, and fingerprint ternary vector to form a coupling event, and the coupling events are chained together into a chain of events, including:

[0048] Step 51: Obtain the batch number corresponding to this batch from the raw material event, and based on the batch number, obtain the window event, process event, and content event associated with this batch. Read the window signature from the window event, the process trajectory summary from the process event, and the fingerprint ternary vector from the content event. This ensures that all types of data involved in the coupled calculation originate from the same batch and represent the key features of the process constraint side, process execution side, and product content side, respectively.

[0049] Step 52: After obtaining all the input data required for the coupling calculation, the batch number, window signature, process trajectory summary, and fingerprint ternary vector are serialized according to a preset fixed field order. This fixed field order ensures consistent serialization results when processing the same data at different times or nodes. Subsequently, a hash calculation is performed on the serialization result to obtain a dual-anchor coupling code. This dual-anchor coupling code is a checksum determined by the batch identifier, process constraint information, process trajectory information, and content fingerprint information. It is used to simultaneously anchor both process-side and content-side data, thereby reflecting the coupling relationship between the production process and product characteristics of this batch. After generating the dual-anchor coupling code, a verification algorithm is executed using the dual-anchor coupling code as input to obtain a checksum. This checksum is used for rapid comparison of the integrity and consistency of the dual-anchor coupling code during subsequent offline verification or validation stages. Finally, the batch number, dual-anchor coupling code, and checksum are encapsulated together to form a coupling event, which is then stored in association with the batch number.

[0050] Step 53: After forming the coupled events, to ensure the continuity and verifiability of production traceability data in the time dimension, the coupled events are further linked to generate chain events. Specifically, the hash value of the preceding chain event associated with the batch number is obtained. The hash value of the preceding chain event refers to the hash verification value corresponding to the chain event formed before the current coupled event, used to represent the previous node in the chain structure. After obtaining the hash value of the preceding chain event, the hash value of the preceding chain event and the current coupled event are serialized in the same fixed field order as described above, and a hash calculation is performed on the serialization result to obtain the hash value of the current chain event. The hash value of the current chain event is used to uniquely identify the position of the current chain event in the chain structure. Subsequently, the coupled event, the hash value of the preceding chain event, and the hash value of the current chain event are jointly encapsulated to form a chain event, and stored in association with the batch number.

[0051] Through the above method, this invention uniformly binds key data generated at different stages in the form of dual-anchor coupling codes and continuously associates them in a time series through a chain event structure. On the one hand, this makes each batch of production data logically form an indivisible traceability link; on the other hand, by associating the hash values ​​of previous chain events with the hash values ​​of current chain events, the chain events have a clear sequential relationship and verifiability, thus providing a reliable data foundation for subsequent two-dimensional code generation, offline silent verification, event verification, and determination of the minimum cut time period for anomalies.

[0052] In one embodiment of the present invention, a two-dimensional carrier code is generated based on chained events, and offline silent verification rules are issued. The offline silent verification rules include: reading the two-dimensional carrier code and obtaining the window signature based on the chained event index; recalculating the verification bits based on the two-dimensional carrier code and the window signature and comparing them, including:

[0053] Step 61: Obtain the batch number corresponding to this batch from the raw material events, and obtain the chain events associated with this batch based on the batch number. After obtaining the chain events, generate a chain event index. The chain event index is used to locate specific chain event nodes in the subsequent verification process and is a key identifier for establishing a mapping relationship between the two-dimensional code and the chain events. Subsequently, extract the batch number, process trajectory summary, fingerprint ternary vector, check bit, and chain event index from the chain events to form a two-dimensional code field set. The two-dimensional code field set is used to describe core information directly related to the production process and product characteristics of this batch. Serialize the two-dimensional code field set according to a preset fixed field order, and encode the serialization result to generate a two-dimensional code. By using a fixed field order and deterministic encoding method, the parsing results of the same two-dimensional code are kept consistent under different reading scenarios.

[0054] Step 62: After generating the two-dimensional carrier code, an offline silent verification rule is issued. This offline silent verification rule refers to the processing rules for verifying the information carried by the two-dimensional carrier code without requiring real-time interaction with the central system. According to this rule, the two-dimensional carrier code is first read and parsed to obtain the batch number, process trajectory summary, fingerprint ternary vector, checksum, and chained event index. Subsequently, the target chained event is located based on the chained event index, and the window signature is obtained from the target chained event.

[0055] Step 63: After completing the two-dimensional carrier code parsing and window signature acquisition, the batch number, window signature, process trajectory summary, and fingerprint ternary vector are serialized according to the fixed field order consistent with the coupling event generation stage. A hash calculation is then performed on the serialization result to obtain the recalculated dual-anchor coupling code. Subsequently, the recalculated dual-anchor coupling code is used as input to execute a verification algorithm to obtain the recalculated check bit. The recalculated check bit is compared with the check bit parsed from the two-dimensional carrier code to determine whether the data corresponding to the two-dimensional carrier code is consistent with the data recorded in the chained events.

[0056] Through the above methods, this invention achieves consistency verification of production traceability data without relying on an online system. On the one hand, the two-dimensional code centrally carries key information such as batch identifiers, process trajectory summaries, and content fingerprints, facilitating rapid reading during production, distribution, or verification stages. On the other hand, the offline silent verification rules, through the combined use of chained event indexes and window signatures, enable the verification process to both reference process constraint information and verify whether the data has been tampered with or replaced.

[0057] In one embodiment of the present invention, the fingerprint ternary vector is retested and the dual-anchor coupling code and check bit are recalculated, and compared with the chained events to form a verification event; when the verification event indicates an anomaly, a constraint set including compliance judgment and tolerance band judgment is constructed to determine the minimum cut time period of the anomaly, forming an anomaly event, including:

[0058] Step 71: Based on the batch number, obtain the chain of events associated with that batch. Read the double-anchor coupling code and check bit from the chain of events as the benchmark values ​​for verification and comparison. Simultaneously, based on the batch number, re-detect the product, collecting the nitrogen solubility index, product water retention capacity, and product isokinetic viscosity. Process the detection results according to the same numerical solidification rules used in the content events to form a re-tested fingerprint ternary vector. This re-tested fingerprint ternary vector reflects the content characteristics of the product at the current verification moment and maintains comparability with previously recorded fingerprint information. Furthermore, based on the batch number, obtain the window signature from the window event and the process trajectory summary from the process event, respectively, as process-side input data for the recalculation coupling calculation.

[0059] Step 72: After acquiring all verification input data, the batch number, window signature, process trajectory summary, and retested fingerprint ternary vector are serialized according to the fixed field order consistent with the coupling event generation stage. A hash calculation is then performed on the serialization result to obtain the recalculated dual-anchor coupling code. Subsequently, a verification algorithm is executed using the recalculated dual-anchor coupling code as input to obtain the recalculated check bits. The recalculated dual-anchor coupling code is compared with the dual-anchor coupling codes recorded in the chained events, and the recalculated check bits are also compared with the check bits recorded in the chained events. A verification event is formed based on the comparison results. This verification event records whether the data is consistent in this verification and serves as the triggering basis for subsequent anomaly handling.

[0060] Step 73: When the verification event indicates an anomaly, the anomaly analysis process begins. Specifically, compliance judgments are obtained from process events based on the batch number. These compliance judgments are obtained by comparing the process trajectory summary and standardized proportioning index using a window matrix, reflecting whether the production process meets the predetermined process objectives. Simultaneously, tolerance bands for each component of the fingerprint ternary vector are obtained from content events based on the batch number, and the retested fingerprint ternary vector is compared with these tolerance bands at the component level to obtain tolerance band judgments. These tolerance band judgments reflect whether the characteristics of the retested product are within the normal fluctuation range.

[0061] The aforementioned compliance and tolerance zone determinations are jointly constructed into a constraint set. This constraint set is used to uniformly describe the process-side and content-side determination conditions used in anomaly analysis. Subsequently, based on the timestamp sequence of chained events associated with the batch number, the satisfaction of the constraint set in the time dimension is analyzed to determine the minimum cut time period for the anomaly. The minimum cut time period for the anomaly refers to the smallest continuous time interval in the chained event time sequence where the constraint set first changes from a satisfied state to a non-satisfied state and continues until recovery, representing the time range in which the anomaly is most likely to occur. The batch number, verification event, constraint set, and minimum cut time period for the anomaly are collectively encapsulated to form an anomaly event, which is then stored in association with the batch number.

[0062] Through the above methods, this invention not only enables consistency verification of key data in the production traceability chain, but also allows for the precise and recalculated location of the timeframe of anomalies when they are detected. On one hand, it improves the accuracy of anomaly identification and analysis during the traceability process of soybean protein-dietary fiber complex food production; on the other hand, it enhances the technical effectiveness of the production traceability method in business management and anomaly handling by basing the anomaly handling process on explicit data constraints and time series analysis.

[0063] In one embodiment of the present invention, determining the minimum cut-off time period for the anomaly when a verification event indicates an anomaly includes:

[0064] Step 81: When the verification event indicates an anomaly, first obtain the chain of events associated with that batch number. Sort the obtained chain of events according to their timestamp sequence to form a chain of event timestamp sequence, and read the chain of events sequentially. For each chain of events, read the batch number, process trajectory summary, fingerprint ternary vector, and check bit one by one, and obtain the window signature based on the chain of events index corresponding to that chain of events.

[0065] Step 82: After completing the chain event data reading, recalculation and verification processing is performed on each chain event. Specifically, the batch number, window signature, process trajectory summary, and fingerprint ternary vector are serialized according to a preset fixed field order, and hash calculation is performed on the serialization result to obtain the recalculated double-anchor coupling code. Subsequently, the verification algorithm is executed using the recalculated double-anchor coupling code as input to obtain the recalculated verification bit. The recalculated verification bit is compared with the verification bit recorded in the current chain event, and if the comparison result is inconsistent, the timestamp of the corresponding chain event is recorded. The inconsistent chain event timestamp is used to identify that at that time point, there is a difference between the data recorded in the chain event and the data recalculated based on the current input.

[0066] Step 83: After verifying all chain events, analyze the timestamps of the recorded inconsistent chain events. Determine the start point of the minimum cut-off time period for the anomaly by identifying the earliest occurrence of inconsistency and the end point by identifying the last occurrence of inconsistency. The resulting minimum cut-off time period represents the smallest continuous time interval within which the abnormal state persists in the production traceability chain. Finally, the minimum cut-off time period is written into the anomaly event generated in step 73, ensuring that the anomaly event includes not only the anomaly determination result but also clear time location information.

[0067] Through the above methods, this invention, when verifying events indicating anomalies, can not only confirm the existence of the anomaly but also definitively locate the time range of the anomaly based on the timestamp sequence of chained events. On the one hand, it avoids the problem of difficulty in determining the duration of an anomaly based solely on single-point anomaly information; on the other hand, it binds the anomaly analysis results to the production traceability chain in the form of time periods, which is beneficial for subsequent tracing, analysis, and handling of anomalies at the business management level.

[0068] The embodiments of the present invention have been described above, but the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of the present embodiments, all of which are within the protection scope of the present embodiments.

Claims

1. A method for traceability in the production of soybean protein-dietary fiber complex foods, characterized in that, Includes the following steps: Step 1: Collect batch serial numbers to generate batch numbers, collect dry-based soybean protein and dietary fiber quality to generate standardized ratio indices, and form raw material events; Step 2: Based on the process objectives of this batch, set a window matrix including pH, standardized ratio index, temperature and time, and generate a window signature; Step 3: Collect pH, temperature and time interval data to generate a process trajectory summary, generate compliance judgments based on standardized ratio index and window matrix, and form process events; Step 4: Collect nitrogen solubility index, product water retention capacity and product isokinetic viscosity to generate a fingerprint ternary vector and the tolerance band for each component of the fingerprint ternary vector to form a content event; Step 5: Generate a dual-anchor coupling code and check bit based on the batch number, window signature, process trajectory summary, and fingerprint ternary vector to form a coupled event. Then, chain these coupled events together to form a chain of events, including: Step 51: Obtain the window signature in the window event, the process trajectory summary in the process event, and the fingerprint ternary vector in the content event based on the batch number; Step 52: Serialize the batch number, window signature, process trajectory summary and fingerprint ternary vector according to the fixed field order, perform hash calculation on the serialization result to obtain the double anchor coupling code, and use the double anchor coupling code as input to perform the verification algorithm to calculate the verification bit, and form a coupling event with the batch number, double anchor coupling code and verification bit; Step 53: Obtain the hash value of the preceding chain event associated with the batch number, serialize the hash value of the preceding chain event and the coupled event in a fixed field order, perform hash calculation on the serialization result to obtain the hash value of the current chain event, and form a chain event with the coupled event, the hash value of the preceding chain event and the hash value of the current chain event and store it in association with the batch number. Step 6: Generate a two-dimensional carrier code based on the chained events and issue offline silent verification rules. The offline silent verification rules include: reading the two-dimensional carrier code and obtaining the window signature based on the chained event index, recalculating the verification bits based on the two-dimensional carrier code and the window signature, and comparing them. Step 7: Retest the fingerprint ternary vector and recalculate the dual-anchor coupling code and check bit, and compare it with the chain event to form a verification event; when the verification event indicates an anomaly, construct a constraint set including compliance judgment and tolerance zone judgment, determine the minimum cut time period of the anomaly, and form an anomaly event.

2. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, Batch numbers are generated by collecting batch serial numbers, and standardized ratio indices are generated by collecting dry-basis soybean protein and dietary fiber quality data, forming raw material events, including: Step 11: Collect batch serial number, verify the batch serial number for digital string format, and compare the batch serial number with the batch serial number in the formed raw material event to complete the uniqueness verification. Step 12: Generate a batch number based on the batch serial number. The batch number consists of the batch serial number and a verification field calculated by a verification algorithm based on the batch serial number. Write the batch number into the raw material event. Step 13: Collect the dry-based soybean protein mass and dietary fiber mass, and perform a range check on the dry-based soybean protein mass and dietary fiber mass that are greater than zero. Add the dry-based soybean protein mass and dietary fiber mass as the denominator, and use the dry-based soybean protein mass as the numerator. Calculate the ratio of the numerator to the denominator to obtain the standardized ratio index, and write the standardized ratio index into the raw material event.

3. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, Based on the process objectives for this batch, a window matrix including pH, standardized ratio index, temperature, and time is set, and a window signature is generated, including: Step 21: Obtain the batch number and standardized ratio index from the raw material event, use the batch number as the batch identifier of the window matrix, and write the standardized ratio index into the window matrix. Step 22: Obtain the process objectives for this batch. The process objectives for this batch include target information and corresponding tolerance information for pH, target information and corresponding tolerance information for standardized proportion index, target information and corresponding tolerance information for temperature, and target information and corresponding tolerance information for time. Write each target information and corresponding tolerance information into the pH target field, pH tolerance field, standardized proportion index target field, standardized proportion index tolerance field, temperature target field, temperature tolerance field, time target field, and time tolerance field of the window matrix, respectively. Step 23: Serialize the window matrix according to the fixed field order, perform hash calculation on the serialization result to obtain the window signature, and combine the window matrix and the window signature to form a window event and store it in association with the batch number.

4. The method for traceability of soybean protein-dietary fiber complex food production according to claim 3, characterized in that, The rows of the window matrix correspond to four types of constraints: pH, standardized ratio index, temperature, and time. The columns include target information columns and tolerance information columns. The matrix elements are the values ​​of the corresponding target information and tolerance information. The target information column corresponds to the target fields of pH, standardized ratio index, temperature, and time. The tolerance information column corresponds to the tolerance fields of pH, standardized ratio index, temperature, and time.

5. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, Collect pH, temperature, and time interval data to generate a process trajectory summary. Based on a standardized ratio index and window matrix, generate compliance judgments to form process events, including: Step 31: Obtain the batch number and standardized ratio index from the raw material event, and obtain the window matrix and window signature from the window event based on the batch number; Step 32: Collect pH, temperature and time intervals to form a collection record sequence. Sum the products of each pH and the corresponding time interval and divide by the sum of all time intervals to obtain the pH weighted average. Sum the products of each temperature and the corresponding time interval and divide by the sum of all time intervals to obtain the temperature weighted average. Accumulate all time intervals to obtain the total time. Generate a process trajectory summary based on the pH weighted average, the temperature weighted average and the total time. Step 33: Perform boundary comparison between the acid-base weighted average value in the process trajectory summary and the acid-base target field and acid-base tolerance field in the window matrix to obtain the acid-base compliance judgment; perform boundary comparison between the temperature weighted average value and the temperature target field and temperature tolerance field to obtain the temperature compliance judgment; perform boundary comparison between the total time and the time target field and time tolerance field to obtain the time compliance judgment; perform boundary comparison between the standardized proportion index and the standardized proportion index target field and standardized proportion index tolerance field to obtain the proportion compliance judgment; summarize to obtain the compliance judgment; and form the process event by combining the batch number, process trajectory summary, compliance judgment, and window signature.

6. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, The nitrogen solubility index, product water retention capacity, and product isokinetic viscosity are collected to generate a fingerprint ternary vector. The tolerance bands for each component of the fingerprint ternary vector are then used to form content events, including: Step 41: Collect nitrogen solubility index, product water retention capacity and product isokinetic viscosity based on batch number, and store nitrogen solubility index, product water retention capacity and product isokinetic viscosity associated with batch number; Step 42: Process the nitrogen solubility index according to the preset numerical solidification rule to obtain the first component of the fingerprint ternary vector; process the water retention capacity of the product according to the preset numerical solidification rule to obtain the second component of the fingerprint ternary vector; process the isokinetic viscosity of the product according to the preset numerical solidification rule to obtain the third component of the fingerprint ternary vector, and form the fingerprint ternary vector. Step 43: Read the preset tolerance threshold for each component of the fingerprint ternary vector, generate a tolerance band for each component of the fingerprint ternary vector with each component as the center value and the corresponding preset tolerance threshold as the offset, and form a content event with the batch number, the fingerprint ternary vector and the tolerance band for each component of the fingerprint ternary vector and store it in association with the batch number.

7. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, A two-dimensional code is generated based on chained events, and offline silent verification rules are issued. These rules include: reading the two-dimensional code and obtaining the window signature based on the chained event index; recalculating and comparing the checksum based on the two-dimensional code and the window signature; and more. Step 61: Based on the batch number, obtain the chain events associated with the batch number, generate the chain event index, extract the batch number, process trajectory summary, fingerprint ternary vector, check bit and chain event index from the chain events as a two-dimensional code field set, and serialize and encode the two-dimensional code field set according to a fixed field order to generate a two-dimensional code. Step 62: Issue offline silent verification rules. The offline silent verification rules include: reading the two-dimensional code to obtain the batch number, process trajectory summary, fingerprint ternary vector, verification bit and chain event index, and locating the target chain event based on the chain event index and obtaining the window signature in the target chain event. Step 63: Serialize the batch number, window signature, process trajectory summary and fingerprint ternary vector according to the fixed field order. Perform hash calculation on the serialization result to obtain the recalculated double-anchor coupling code. Use the recalculated double-anchor coupling code as input to execute the verification algorithm to calculate the recalculated check bit. Compare the recalculated check bit with the check bit in the two-dimensional carrier code.

8. The method for traceability of soybean protein-dietary fiber complex food production according to claim 1, characterized in that, The fingerprint ternary vector is retested and the dual-anchor coupling code and check bit are recalculated, then compared with the chained events to form a verification event. When the verification event indicates an anomaly, a constraint set including compliance judgment and tolerance band judgment is constructed to determine the minimum cut time period of the anomaly, forming an anomaly event, including: Step 71: Based on the batch number, obtain the double-anchor coupling code and check bit in the chain event; based on the batch number, collect the nitrogen solubility index, product water retention capacity and product isokinetic viscosity and form a fingerprint ternary vector for retesting; based on the batch number, obtain the window signature in the window event and the process trajectory summary in the process event. Step 72: Serialize the batch number, window signature, process trajectory summary and retest fingerprint ternary vector according to a fixed field order. Perform hash calculation on the serialization result to obtain the recalculated double-anchor coupling code. Use the recalculated double-anchor coupling code as input to execute the verification algorithm to calculate the recalculated check bit. Compare the recalculated double-anchor coupling code with the double-anchor coupling code in the chain event. Compare the recalculated check bit with the check bit in the chain event. Based on the comparison result, form a verification event. Step 73: When the verification event indicates an anomaly, obtain the compliance judgment in the process event based on the batch number, obtain the tolerance zone for each component of the fingerprint ternary vector in the content event based on the batch number, and compare the boundary of the retested fingerprint ternary vector with the tolerance zone to obtain the tolerance zone judgment. Construct a constraint set based on the compliance judgment and the tolerance zone judgment, and determine the minimum cut time period of the anomaly based on the timestamp sequence of the chained events. Combine the batch number, verification event, constraint set and minimum cut time period of the anomaly to form an anomaly event.

9. A method for traceability of soybean protein-dietary fiber complex food production according to claim 8, characterized in that, When verifying an event indicating an anomaly, determine the minimum cut time period for the anomaly, including: Step 81: When the verification event indicates an anomaly, obtain the chain of events associated with the batch number based on the batch number, sort the chain of events according to the timestamp sequence, read the batch number, process trajectory summary, fingerprint ternary vector and check bit in each chain of events one by one, and obtain the window signature based on the chain of events index. Step 82: For each chain event, serialize the batch number, window signature, process trajectory summary and fingerprint ternary vector in a fixed field order. Perform hash calculation on the serialization result to obtain the recalculated double-anchor coupling code. Use the recalculated double-anchor coupling code as input to perform a verification algorithm to calculate the recalculated check bit. Compare the recalculated check bit with the check bit in the chain event and record the timestamp of the inconsistent chain event. Step 83: Take the earliest inconsistent chain event timestamp as the starting point of the minimum cut time period of the anomaly, take the latest inconsistent chain event timestamp as the ending point of the minimum cut time period of the anomaly, and write the minimum cut time period of the anomaly into the anomaly event formed in step 73.

Images