Andunna active peptide production whole cycle tracking and management method based on internet of things

Abstract

The application relates to the field of data processing management, and discloses a giant salamander active peptide production whole-cycle tracing and management method based on the Internet of Things, which comprises the following steps: constructing a batch cumulative compliance matrix containing cumulative compliance state numerical indicators in a storage; determining a process stage and loading a judgment rule on a time axis constructed by a context flow according to an environmental parameter physical timestamp; calculating a risk increment and updating the matrix; when a process change event with an occurrence time earlier than processed data is received, performing backtracking recalculation: deducting an original contribution value, recalculating according to a new rule and accumulating a new contribution value; and through performing a canceling-recomputing atomic operation on the batch cumulative compliance matrix, the application solves the rule mis-matching problem caused by heterogeneous data flow time sequence dislocation, and avoids the interference of network delay on the consistency of compliance judgment logic.

Description

Technical Field

[0001] This invention relates to a method for full-cycle traceability and management of the production of active peptides from giant salamanders based on the Internet of Things, belonging to the field of management data processing technology. Background Technology

[0002] In the current industrial manufacturing of biopharmaceuticals and high-value-added functional foods, ensuring that the entire production process complies with pharmaceutical production quality management standards or food safety standards is a basic industry principle. For bioactive substances such as giant salamander active peptides, which are highly sensitive to environmental factors, the production process involves multiple continuous steps with vastly different environmental requirements, including cleaning, enzymatic hydrolysis, inactivation, filtration, concentration, and drying. To achieve real-time monitoring of the production environment, modern factories generally deploy Internet of Things (IoT) sensor networks to collect key physical parameters such as temperature, pH, and dissolved oxygen, and combine this with the production execution system to issue process instructions to manage compliance. Current research focuses on optimizing enzymatic hydrolysis processes or improving composition formulations to obtain end products with specific bioactivity, but this also depends on the precise control of environmental factors in the production process.

[0003] The challenges faced by heterogeneous data flow management models in actual industrial production network environments include the transmission of process switching instructions from the production execution system and real-time environmental data from sensor networks through different physical or logical channels. Affected by network congestion, protocol conversion, or differences in device response, these two data sources are prone to timing misalignment upon arrival at the data processing terminal. This timing asynchrony leads to the management system incorrectly using old process judgment rules to evaluate the environmental data generated by the new process at critical process switching nodes, resulting in numerous compliance misjudgments and false positive or false negative records, severely undermining the logical authenticity and auditability of electronic batch records. Furthermore, the loss of bioactive peptide activity is a cumulative result of environmental stress, not a single-moment instantaneous effect of parameter exceeding limits. Most existing management systems employ threshold-based instantaneous alarm mechanisms, lacking the ability to quantitatively assess the cumulative heat dose or stress fatigue process risks. The process switching from enzymatic hydrolysis to inactivation involves a significant temperature difference, and the thermal inertia of the physical environment within the reactor causes gradual changes in physical parameters. However, rule switching is set as an instantaneous logical jump. The inherent contradiction between gradual physical change and abrupt logical change makes normal process data during the transition period easily judged as non-compliant by the system, requiring managers to perform extensive manual data cleaning, reducing management efficiency and increasing the risk of data tampering.

[0004] Therefore, the technical problem to be solved by this invention is to construct a data processing method that automatically adapts to the temporal misalignment of heterogeneous data, quantifies the accumulated risks of the process in real time, and resolves the conflict between physical inertia and rules, so as to generate a logically rigorous, authentic, and reliable full-cycle compliant record of production. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of this invention is as follows: A method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things, the method running on a data processing system, including the following steps:

[0006] A batch cumulative compliance matrix is ​​constructed in memory for the production batch of giant salamander active peptide. The batch cumulative compliance matrix contains numerical indicators to characterize the cumulative compliance status of the production batch throughout the entire production cycle.

[0007] The system receives environmental data streams and context streams in parallel. The environmental data streams contain environmental parameters with physical timestamps, and the context streams contain process change events with timestamps indicating when they occurred.

[0008] Based on the physical timestamp of the environmental parameters, determine the first process stage to which the environmental parameters belong on the context timeline constructed by the context flow, and load the first judgment rule corresponding to the first process stage;

[0009] Based on the first judgment rule, the first compliance risk increment generated by environmental parameters is calculated and added to the numerical index of the batch cumulative compliance matrix;

[0010] The method also includes a backtracking recalculation step: when the timestamp of the received process change event is earlier than the physical timestamp of the environmental parameter that has been processed according to the first judgment rule, the batch cumulative compliance matrix is ​​locked and the processed environmental parameter is retrieved;

[0011] Based on the first judgment rule, calculate the original contribution value of the processed environmental parameters to the numerical indicators, and deduct the original contribution value from the batch cumulative compliance matrix;

[0012] Based on the second judgment rule corresponding to the second process stage indicated by the process change event, the second compliance risk increment of the processed environmental parameters is recalculated and added to the batch cumulative compliance matrix.

[0013] Preferably, the backtracking recalculation step further includes: after completing the deduction of the original contribution value and the accumulation of the second compliance risk increment, generating a status correction log, which records the original contribution value, the second compliance risk increment, and the system time of executing the backtracking recalculation step; associating the status correction log with the electronic batch record of the production batch; unlocking the batch cumulative compliance matrix and restoring the processing of the real-time environmental data stream.

[0014] Preferably, the step of calculating the first compliance risk increment generated by environmental parameters follows the following calculation logic: ,in This represents the first increase in compliance risk. These are the measured values ​​of environmental parameters. The compliance benchmark value specified in the first judgment rule. The time difference between the environmental parameters and the previous sampling point. This is the risk weighting coefficient associated with the first process stage.

[0015] Preferably, the method further includes a dual-track judgment step for the process transition period: when the context flow indicates that the process stage is switching from the first process stage to the second process stage, a logical transition time window covering the switching time is determined based on preset process physical inertia parameters; for environmental parameters whose physical timestamps are located within the logical transition time window, a first judgment rule and a second judgment rule are loaded; the compliance status of the environmental parameters is calculated according to the first judgment rule and the second judgment rule respectively; only when the environmental parameters are determined to generate compliance risk increments under both the first judgment rule and the second judgment rule are the compliance risk increments updated to the batch cumulative compliance matrix.

[0016] Preferably, the dual-track judgment steps for the process transition period further include: if the environmental parameter is determined to have not generated an incremental compliance risk under either the first or second judgment rule, an exemption mark is generated; the exemption mark is associated with the environmental parameter; the risk value generated by the environmental parameter is prohibited from being accumulated to the numerical indicator; and the transition period exemption count is added to the batch cumulative compliance matrix.

[0017] Preferably, the method further includes a parameter logic topology verification step: pre-setting parameter logic association rules in the rule base, the parameter logic association rules defining the numerical linkage relationship or state mutual exclusion relationship that must be satisfied between the first type of environmental parameters and the second type of environmental parameters; before performing the step of calculating the first compliance risk increment, obtaining the associated environmental parameters that have the same physical timestamp as the currently processed environmental parameters; determining whether the combination of the environmental parameters and associated environmental parameters satisfies the parameter logic association rules; if not, intercepting the environmental parameters, prohibiting the updating of the batch cumulative compliance matrix, and synchronously accumulating the data credibility anomaly count in the batch cumulative compliance matrix.

[0018] Preferably, the numerical indicators in the batch cumulative compliance matrix include a cumulative heat dose indicator and a cumulative pH stress indicator; the cumulative heat dose indicator is used to characterize the total value of the thermal degradation risk borne by the giant salamander active peptide throughout the entire production cycle; the cumulative pH stress indicator is used to characterize the total integral value of the acid-base environment sudden pressure borne by the giant salamander active peptide during the enzymatic hydrolysis and inactivation process; the step of calculating the first compliance risk increment generated by the environmental parameters includes calculating the corresponding compliance risk increment according to the type of environmental parameter and adding it to the cumulative heat dose indicator or the cumulative pH stress indicator.

[0019] Preferably, the method further includes a compliance status invalidation step: after each update of the batch cumulative compliance matrix, the numerical indicator is compared with a preset management termination threshold; if the numerical indicator exceeds the management termination threshold, a batch compliance invalidation instruction is generated; the status identifier of the production batch is modified to invalidation status, and the compliance calculation process for subsequent data of the production batch is terminated.

[0020] Preferably, the context stream is generated and sent by the production execution system, and the process change events include cleaning start event, enzymatic hydrolysis start event, inactivation start event and filtration start event; the environmental data stream is collected and sent by an Internet of Things sensor network distributed on the production site, and the environmental parameters include reactor temperature, solution pH value and stirring motor current.

[0021] Preferably, in the parameter logic topology verification step, the first type of environmental parameter is the stirring motor current, and the second type of environmental parameter is the dissolved oxygen concentration; the parameter logic association rule is limited to the fact that when the stirring motor current indicator is in the off state, the rate of change of dissolved oxygen concentration must not exceed the preset natural diffusion fluctuation threshold; if the stirring motor current is zero and the rate of change of dissolved oxygen concentration exceeds the natural diffusion fluctuation threshold, then it is determined that the combination of environmental parameter and associated environmental parameter does not satisfy the parameter logic association rule.

[0022] Compared with the prior art, the beneficial effects of the present invention are:

[0023] 1. Establish a rigorous and auditable compliance judgment logic in an asynchronous data environment. Construct a backtracking and recalculation mechanism based on logical timestamps to solve the problem of mismatched management rules caused by the misalignment of process context signals and environmental sensor data transmission in the industrial IoT environment. When a process change event occurs earlier than the current time is received, the processing unit automatically locks the corresponding batch's cumulative compliance matrix, retrieves and cancels the compliance status contribution value calculated according to the old process rules within the affected period, and recalculates the data compliance risk increment of the period according to the new process rules. The data processing logic of first canceling and then correcting forces the compliance judgment basis to be anchored to the objective time of the physical event, rather than the random time when the data arrives at the server. This eliminates compliance false alarms and omissions caused by network delays or out-of-order data processing system, ensuring that the final batch compliance record is logically strictly consistent with the actual production process. This provides a legally traceable and authentic data ledger for the strict management scenarios of Good Manufacturing Practice (GMP) for pharmaceuticals.

[0024] 2. This invention enables quantitative management of cumulative risks in the production process of bioactive substances by maintaining a batch cumulative compliance matrix in memory. It upgrades the instantaneous threshold alarm mode to a process cumulative risk measurement mode. Based on the judgment rules, it calculates the product of the magnitude and duration of environmental parameters deviating from the compliance range in real time, and adds the incremental compliance risk to the corresponding index in the matrix. The data structure captures and records the hidden risks of heat dose accumulation and pH stress fatigue. Although the values ​​at a single moment do not exceed the limit, the long-term accumulation leads to the loss of activity. Based on the integral logic state calculation method, the management system can identify the latent quality degradation trend in the production process. It expands the supervision dimension of the thermosensitive biological assets of giant salamander active peptides from simple parameter monitoring to full-cycle cumulative effect control, and improves the ability of management data to characterize the final quality attributes of the product.

[0025] 3. To resolve the logical conflict between the inertia of physical processes and the rigidity of rule switching, a dual-track judgment logic is introduced during the process switching stage. This eliminates the mismatch between the gradual change of physical parameters of production equipment and the instantaneous switching of management rules, which can lead to false violation records. Within the logical transition time window determined by process parameters, the system loads the judgment rules for both processes before and after, and calculates the compliance status under different rules for environmental parameters. When both sets of rules determine that an increase in risk has occurred, the processing unit will include the data in the violation accumulation; otherwise, it will be identified as a reasonable process transition fluctuation and exempted. The processing logic does not rely on manual intervention to screen data and automatically adapts to the physical thermal inertia of the reactor heating or cooling process. This ensures the objectivity and accuracy of the output conclusions of the fully automated traceability system during the complex process connection stage, and avoids abnormal noise caused by non-quality factors in the compliance report. Attached Figure Description

[0026] Figure 1 This is a flowchart of the production lifecycle traceability method including a backtracking and recalculation mechanism according to the present invention.

[0027] Figure 2 This is a comparison diagram of parameters and process status under the heterogeneous data stream timing misalignment scenario of the present invention;

[0028] Figure 3 This is the state transition diagram of the data processing system that combines backtracking and judgment mechanisms according to the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this invention clear, the invention will be described in detail below with reference to specific embodiments. The embodiments and descriptions of this invention are only for explaining this invention and are not intended to limit this invention.

[0030] This invention discloses a method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things (IoT), running on a data processing system with data reception, storage, and processing capabilities. The method constructs a batch cumulative compliance matrix for each batch of giant salamander active peptide production and establishes data processing logic. This logic includes receiving environmental data streams and context streams, performing dynamic rule matching and compliance risk increment calculation on the environmental data based on the context, and including a backtracking recalculation mechanism to correct the calculated cumulative compliance status when a data stream timing misalignment is detected, ensuring the logical consistency of management records. When the data processing system receives a new batch production instruction, it stores... The memory contains an instantiated data structure for the production batch, namely the batch cumulative compliance matrix. The memory can be a cache or a database. This matrix is ​​used to cumulatively quantify compliance risks and includes at least one numerical indicator. These indicators include a cumulative heat dose indicator and a cumulative pH stress indicator. The cumulative heat dose indicator characterizes the total thermal degradation risk experienced by the batch throughout its entire lifecycle, and the cumulative pH stress indicator characterizes the integral total value of the acid-base environment abrupt stress experienced. All indicators are initialized to zero or a baseline value when the batch is created. To support rapid retrieval and contribution deduction of processed environmental parameters during the backtracking recalculation step, the data processing system stores a separate data structure for each batch in memory. Each production batch maintains an independent data index, which bidirectionally associates the unique identifier of the batch's cumulative compliance matrix with the original environmental parameter records in the time-series database. The batch's cumulative compliance matrix exists in memory as key-value pairs, where the key is the cumulative indicator type, such as the cumulative heat dose indicator, and the value is the current floating-point cumulative value. When the matrix is ​​locked during the backtracking recalculation step, the update operation of the key-value pair is locked, without blocking the parallel reading and new data writing of the time-series database. During runtime, the data receiving module receives two types of data streams in parallel through different logical channels: the first type is the environmental data stream, generated by the IoT sensor network deployed on the production site. The first type is network-generated, characterized by high frequency and continuity. Each data packet contains a physical timestamp, equipment identifier, and environmental parameters. Taking the production process of giant salamander active peptides as an example, environmental parameters may include reactor temperature, solution pH value, stirring motor current, or dissolved oxygen concentration. The second type is context stream, generated and sent by the production execution system. It has low frequency and event-driven characteristics. The data packet contains the event occurrence timestamp, batch identifier, and process change event. Taking the enzymatic hydrolysis process as an example, process change events may include cleaning start event, enzymatic hydrolysis start event, inactivation start event, or filtration start event, etc. The occurrence timestamp marks the actual switching time of the process.

[0031] To achieve rule matching, the processing unit dynamically maintains a context timeline in memory for each batch based on the received context stream. This timeline is a time sequence divided by the timestamps of process change events. When the processing unit receives a data packet from the environmental data stream, it extracts the physical timestamp of the data packet, searches it on the context timeline, and determines the process stage to which the physical timestamp belongs. This process stage is the first process stage. The processing unit loads the first judgment rule corresponding to the first process stage from the rule repository. This rule defines the compliance range, benchmark value, or risk weight of each environmental parameter under this process. The context timeline constructed by the context stream is implemented in the data processing system as a linked list of process stages arranged in ascending order of occurrence timestamps. Each node in this linked list contains the start time of the process stage, the process stage identifier, and the storage address pointing to the first judgment rule. When an environmental data stream is received, the processing unit uses its physical timestamp to search the process stage linked list through binary search or interval tree to determine the time range to which it belongs and loads the corresponding first judgment rule. After loading the first judgment rule, the processing unit calculates the first compliance risk increment generated by the environmental parameters. The calculation follows the formula below: ,in, These are the measured values ​​of environmental parameters. The first judgment rule specifies the compliance benchmark value or interval boundary value for this process. This is the physical timestamp of the current environmental parameter data packet and the time difference between the sampling point of the previous processed data. This is a risk weighting coefficient associated with the first process stage. This coefficient is preset based on the sensitivity of different processes to product quality. The processing unit will... The numerical indicators, such as the cumulative heat dose indicator, are added to the batch cumulative compliance matrix.

[0032] The transmission delay between the context stream and the environmental data stream is not synchronized. Often, the timestamp of a process change event occurs earlier than the physical timestamp of the processed environmental parameters. For example, if the system receives an enzymatic hydrolysis start event with a timestamp of 10:00 at 10:05, but multiple temperature data points have already been processed according to the old cleaning rules between 10:00 and 10:05, this method includes a backtracking recalculation step: when the above timing misalignment is detected, the processing unit locks the batch cumulative compliance matrix for that batch and suspends new data writing; the system retrieves all processed environmental parameters from the 10:00 to 10:05 period; based on the old first judgment rule, i.e. The system performs a cleaning process, calculating the original contribution value generated by these data and deducting it from the corresponding numerical indicators in the batch cumulative compliance matrix. Based on the second process stage indicated by the newly received process change event (i.e., enzymatic hydrolysis), it loads the corresponding second judgment rule, recalculates the second compliance risk increment of all environmental parameters within this period according to the new rule, and adds the new contribution value to the batch cumulative compliance matrix. After completing the deduction and accumulation, the system generates a status correction log, which records the original contribution value, the second compliance risk increment, and the system time of the backtracking recalculation, and is associated with the electronic batch record. The processing unit unlocks the matrix and resumes real-time data processing.

[0033] In the production of giant salamander active peptides, when switching from enzymatic hydrolysis to high-temperature inactivation, physical parameters such as reactor temperature exhibit thermal inertia and cannot change instantaneously, while rule switching is instantaneous. This conflict between gradual physical change and abrupt logical change can easily lead to false violation records. To address this, this method also includes a dual-track judgment step for the process transition period: When a process switching event is received, the processing unit determines the logical transition time window covering the switching moment based on preset process physical inertia parameters, which can be set to 5 minutes after the switch. For environmental parameters whose physical timestamps fall within this logical transition time window, the processing unit loads the first judgment rule before the switch and the second judgment rule after the switch. The system calculates the compliance status based on these two sets of rules and executes the judgment logic. Only when the environmental parameter is determined to generate an increase in compliance risk under both sets of rules is it included in the violation accumulation and updated to the batch accumulation compliance matrix. If the parameter is determined to be compliant under either rule, the system determines it as a physical transition, generates an exemption mark, prohibits risk accumulation, and adds a transition period exemption count to the matrix.

[0034] To ensure the authenticity of the data entering the database and prevent compliant but false data from contaminating records due to sensor failure or data forgery, this method may also include a parameter logic topology verification step. This step is performed before calculating the incremental compliance risk and uses the inherent causal relationships between different physical parameters for cross-validation. The system presets parameter logic association rules in the rule base, defining the numerical linkage or state mutual exclusion relationships that must be satisfied between different types of environmental parameters. Taking the stirring process in the production of giant salamander active peptides as an example, the parameter logic association rule is limited to the relationship between the first type of environmental parameter, the stirring motor current, and the second type of environmental parameter, the dissolved oxygen concentration. When the stirring motor current indicator is in the off state, there is a lack of mechanical mixing in the reactor, and the rate of change of dissolved oxygen concentration must not exceed the preset natural diffusion fluctuation threshold. When the processing unit receives dissolved oxygen data, it obtains the stirring motor current data at the same physical timestamp and determines whether the data combination is valid. If the stirring motor current is found to be zero, but the dissolved oxygen concentration change rate exceeds the threshold, the combination is determined to violate the rule. The system will intercept this dissolved oxygen data, prohibit the updating of regular numerical indicators in the batch cumulative compliance matrix, and simultaneously accumulate an independent data credibility anomaly count in the matrix to ensure the data foundation for management. During the operation of the method, after each update of the numerical indicators in the batch cumulative compliance matrix, the data processing system compares the current cumulative value of the indicator with the preset management termination threshold. The termination threshold is a management threshold set based on the critical point of irreversible loss of the quality attributes of the giant salamander active peptides, such as the total cumulative calorific value not exceeding a certain specific value. If any numerical indicator exceeds its corresponding management termination threshold, the system generates a batch compliance failure instruction, changes the status identifier of the production batch to a failure status, terminates the compliance calculation process of subsequent data for that batch, or sends a warning to the management personnel.

[0035] Example 1: In the management of batch DS-2025-04A of giant salamander active peptide production, the data processing system constructs a batch cumulative compliance matrix in memory for this batch. This matrix includes a cumulative heat dose index with an initial value of 0. At T=09:15:00, the production execution system issues an enzymatic hydrolysis start command. This command is received and recorded by the context stream event data processing system. The system then loads the corresponding first judgment rule, i.e., the enzymatic hydrolysis rule, which requires the reactor temperature to be... Maintain at 45 Otherwise, calculate the incremental compliance risk. During the period from T=09:15:00 to T=09:30:00, the IoT sensor network stably transmits environmental data streams, and all temperature readings are above 40°C. Up to 42 Within the specified interval, the processing unit determines data compliance based on enzymatic hydrolysis rules, and the cumulative heat dose index in the batch cumulative compliance matrix remains at 0. At T=09:30:00, the production execution system issues an inactivation start command, and the reactor begins to heat up. The timestamp of this command is T=09:30:00. Due to local congestion in the industrial network, this critical context flow event fails to reach the data processing system in a timely manner. The high-frequency environmental data flow remains uninterrupted. From T=09:30:05, the processing system begins to receive environmental parameters after the heating up. At T=09:30:05... =50 At T=09:30:10 =55 .

[0036] Due to context stream delay, the data processing system's internally maintained context timeline still showed the batch in the enzymatic digestion process between T=09:30:05 and T=09:31:00, causing the system to incorrectly continue using the required temperature below 45°C. The first judgment rule is the enzymatic hydrolysis rule, which determines the reaction time when T=09:30:05. =50 and at T=09:30:10 =55 The system calculates the compliance risk increment based on the error rule and continuously adds it to the cumulative heat dose index of the batch cumulative compliance matrix, causing the status of the batch to be incorrectly marked as non-compliant in the electronic record, even though the physical heating process meets the process requirements. At T=09:31:01, the data processing system receives the inactivation start process change event with a delayed occurrence timestamp of T=09:30:00. The system detects the occurrence timestamp of the event T=09:30:00 and processes the environmental parameters with physical timestamps between T=09:30:05 and T=09:31:00 according to the first judgment rule. The system immediately triggers the backtracking recalculation step: the processing unit locks the batch cumulative compliance matrix of batch number DS-2025-04A and retrieves all processed environmental parameters within the physical timestamp period. The system calculates all data (including 50) within this period according to the original enzymatic digestion rule. 55 The original contribution value generated by the reading is deducted from the cumulative heat dose index in the batch cumulative compliance matrix, restoring the index to the state value of 0 at T=09:30:00; the processing unit, based on the second process stage indicated by the process change event at T=09:30:00, i.e., inactivation, loads the corresponding second judgment rule, which requires a temperature Greater than 90 The system uses this new rule to recalculate environmental parameters (including 50) for the period from T=09:30:00 to T=09:31:00. 55 The second compliance risk increment (of the readings); because none of these temperature values ​​reached 90. If the dual-track judgment step during the process transition period is not enabled, the second compliance risk increment calculated by the system is still a non-zero value. The system adds these newly calculated increment values ​​to the batch cumulative compliance matrix. The system generates a status correction log to record the detailed process of this reversal and recalculation, and unlocks the matrix. This method automatically corrects compliance misjudgments caused by network latency, ensuring that the final value recorded in the batch cumulative compliance matrix is ​​a logically consistent result calculated based on the correct process context.

[0037] Example 2: This example verifies the role of the retrospective recalculation mechanism of the present invention in maintaining the logical consistency of the batch cumulative compliance matrix when faced with context stream network transmission delays. The experimental environment is set up as a data processing system simulation platform, which can replay heterogeneous data streams according to predetermined timestamps and has the function of simulating arbitrary delays in the context stream. The experimental data source includes two parts: one is the real historical record of the high-frequency environmental data stream of the entire cycle of the giant salamander active peptide production batch XJ-2025-07B, which is sampled at 1-second intervals and contains complete temperature change data from the enzymatic hydrolysis process to the inactivation process; the other is the corresponding context stream, which contains T=1200s. The experiment included two treatment groups: a control group, which used standard data processing logic based on data arrival timestamps and lacked a backtracking recalculation mechanism; and an experimental group using the complete method described in the specific implementation, which included a backtracking recalculation mechanism. For the inactivation start event occurring at T=1500s, different transmission delays were artificially injected to simulate the temporal misalignment intensity gradient commonly found in industrial networks. The cumulative heat dose indices calculated from the two groups were compared. The judgment rules for the two groups were as follows: the first judgment rule, i.e., the enzymatic hydrolysis rule, required the temperature to be below 45°C. The excess is included in the cumulative heat dose; the second criterion, the inactivation rule, requires the temperature to be above 90°C. The portion below this threshold is included in the cumulative heat dose; historical data at T=1500s shows that the temperature begins to rise from 42... The temperature increased, and remained at 50°C between T=1500s and T=1560s. Up to 75 The heating range; the test operation and data acquisition process are as follows: network delays of 0 seconds, 30 seconds and 60 seconds were injected into the control sample group and the sample group of the present invention, respectively, and the values ​​of the final cumulative heat dose index in the cumulative compliance matrix of each sample group were recorded after the simulation operation was completed. The results are summarized in Table 1.

[0038] Table 1: Comparison of Cumulative Heat Dose Indicators under Different Network Latencies

[0039]

[0040] Referring to Table 1, in the absence of network delay, the baseline cumulative heat dose calculated for control group A was 41.5 units; when the inactivation start event was delayed by 30 seconds (control group B) and 60 seconds (control group C), the control groups incorrectly applied the old enzymatic digestion rule (requiring <45). ) Determine 50 seconds after T=1500s Up to 75 The temperature rise data caused a large amount of temperature rise data that should have been calculated according to the new rules to be misjudged as non-compliant, resulting in the final cumulative heat dose indicators increasing to 1840.8 and 3652.1 respectively, which deviated from the physical facts. Under the same delay conditions, sample group A (delay of 30 seconds) and sample group B (delay of 60 seconds) of this invention triggered the backtracking recalculation mechanism after receiving the delayed context event. The system automatically canceled the contribution value that was incorrectly calculated according to the old rules during the delay period (T=1500s to T=1530s or T=1560s), and recalculated the compliance risk increment for this period according to the second judgment rule starting from T=1500s. The final cumulative heat dose indicators were 41.9 and 42.3 respectively, which were basically consistent with the benchmark value of 41.5. The slight difference was due to floating-point operations or boundary handling in the recalculation process.

[0041] Example 3: This example combines Figures 1 to 3 This document describes a method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things (IoT). Figure 1As shown, the data stream includes physical timestamps and environmental parameters. After data input, it enters the parameter logic topology verification step. This step intercepts data that does not meet mutual exclusion or linkage rules based on the causal relationship of physical parameters, such as current and dissolved oxygen. After the logic verification is passed, it enters the context timeline matching stage. In this stage, it combines the context stream from the production execution system, which includes process change events and their timestamps. Based on the physical timestamps, it determines the process stage and loads the corresponding judgment rules, then enters the compliance risk increment calculation step. It calculates the deviation and duration of the measured value from the benchmark value to quantify the cumulative process risk. The calculation results are updated to the batch cumulative compliance matrix stored in the cache or database, specifically added to the cumulative heat dose index or p The H-stress index process includes two special branches: the first is the dual-track judgment step during the process transition period, which is triggered by the process switching logic transition time window. Within the time window, two sets of rules are loaded. If there is a double violation, it is counted; otherwise, it is exempted. The second is the core mechanism of the backtracking recalculation step, which is triggered when a time sequence misalignment is detected, i.e., the process change event is earlier than the processed data. The process involves locking the matrix and deducting the original contribution value, recalculating and accumulating the new contribution value according to the new rules, and generating a status correction log. Finally, the process enters the compliance status failure step, which determines whether the accumulated value exceeds the management termination threshold. If it exceeds the threshold, a failure instruction is generated and termination or failure is executed. If the process complies with the rules, an electronic batch record or batch compliance report is finally generated.

[0042] like Figure 2 As shown, the horizontal axis represents the time range from 09:29:30 to 09:31:00, and the left vertical axis represents the reactor temperature in units of... The right vertical axis represents the process identifiers, with enzymatic hydrolysis corresponding to a value of 0 and inactivation corresponding to a value of 1. The solid line in the graph represents the temperature change trajectory of the reactor, showing that the temperature began to rise around 09:30:00, from 42... Gradually climbed to 75 The dashed lines in the diagram represent the state transition of a process stage, showing that the process identifier changed from 0 to 1 at 09:31:00. The area with a shaded background is clearly marked as the period during which the erroneous rule was applied; this area covers the time interval when the temperature rises but the process identifier remains at 0. Figure 3As shown, the system state transition logic begins with receiving production instructions and entering the initialization state. In this state, the batch cumulative compliance matrix is ​​instantiated and the numerical indicators are reset to zero. As the environmental data stream arrives, the system enters the real-time compliance calculation state, performing operations such as loading the current process judgment rules, parameter logic topology verification, and calculating and accumulating the compliance risk increment. When a process change event occurs, the system enters the process transition period judgment state, performing parallel loading of the two sets of rules and dual-track compliance judgment, and exiting after the logical transition time window ends or after generating the exemption mark. When a data stream timing misalignment is detected, the system enters the backtracking locking and recalculation state, sequentially performing operations such as locking the batch cumulative compliance matrix, retrieving processed environmental parameters, canceling the original contribution value, recalculating and accumulating the new contribution value, and generating a status correction log. After correction, it returns to the real-time compliance calculation state. If the accumulated value exceeds the management termination threshold, the system enters the failure state, generates a batch compliance failure instruction, and terminates subsequent compliance calculations. If the entire production cycle ends and is compliant, the final output is an electronic batch record or a batch compliance report.

[0043] Example 4: This example describes a standardized engineering procedure for offline calibration and systematic configuration of key parameters required for the method of the present invention. This eliminates the empirical setting of judgment rules in the data processing system, ensuring that all management logic has a reproducible engineering and data foundation. When the data processing system is first deployed on a specific giant salamander active peptide production line, the key judgment parameters in the rule repository include the logical transition time window required for the dual-track judgment step during the process transition period, the data credibility threshold required for the parameter logic topology verification step, and the calculation of compliance risk increments. The risk weighting coefficient used Both the management termination threshold and the calibration process are in a state of pending calibration, requiring the construction of a calibration procedure to determine these parameters. The procedure for calibrating the logical transition time window, taking the switch from enzymatic hydrolysis to inactivation as an example, involves the physical process of the reactor temperature decreasing from 45°C. The following has been increased to 90. The above describes a process where, under standard materials and volume, the physical heating process for switching steps is repeated 10 times. Each time, the time from the moment the production execution system issues the inactivation start command to the actual measured value by the temperature sensor inside the reactor is recorded. It reached and remained stable at 90.0 for the first time. The above represents the actual time consumption. Ten sets of time consumption data were collected, specifically 305 seconds, 312 seconds, 308 seconds, 315 seconds, 309 seconds, 311 seconds, 318 seconds, 310 seconds, 306 seconds, and 313 seconds. The average value of this set of data was calculated to be 310.7 seconds, and the standard deviation was 3.8 seconds. The logical transition time window was set to the average value plus three standard deviations. The seconds are rounded down to 325 seconds and stored in the rule base. The threshold of the association rule between the stirring motor current and the rate of change of dissolved oxygen concentration in the calibration parameter logic topology verification step is used. The procedure is to retrieve the full-cycle environmental data stream of 100 confirmed quality qualified production batches from the historical database.

[0044] The system screened all data points where the measured current of the stirring motor was below the shutdown threshold of 0.1 amperes, and calculated the absolute value of the dissolved oxygen concentration change rate corresponding to these data points. Statistical distribution analysis was performed on the 51,320 selected change rate values, determining the 99.9th percentile to be 0.045 mg / L / s. The system set the natural diffusion fluctuation threshold in this association rule to 0.05 mg / L / s and stored it in the rule base. The risk weighting coefficient was calibrated. The procedure for managing the termination threshold is as follows: Select 50 historical production batches, each with complete environmental data stream records and end-product quality inspection reports, including the percentage of active peptide purity; set benchmarks for the data processing system. The value is 1.0, based on The formula back-calculates the total raw risk score accumulated for each batch throughout the entire cycle; a regression model is established between the total raw risk score (as the independent variable) and the purity of the active peptide (as the dependent variable); through this regression model, the critical value of the total raw risk score that causes the predicted purity value to fall below 98.0% of the quality specification lower limit is determined, and this critical value is set as the management termination threshold; the slope value of the regression model in the interval near the quality specification lower limit is used to determine the risk weighting coefficient for this process. .

[0045] Example 5: In the enzymatic hydrolysis process of the giant salamander active peptide production batch with batch number MJ-2025-10C, the data processing system continuously receives environmental data streams according to the first judgment rule. Before calculating the incremental compliance risk, a parameter logic topology verification step is executed. The preset parameter logic association rule in this step, based on the calibration results, limits that when the current of the stirring motor (a first-type environmental parameter) is lower than the shutdown threshold of 0.1 amperes, the absolute value of the rate of change of the dissolved oxygen concentration (a second-type environmental parameter) must not exceed the natural diffusion fluctuation threshold of 0.05 mg / L / s. At T=14:30:05, the system receives a data snapshot, showing that the stirring motor current is 0.05 amperes, indicating... The system indicated a shutdown, with a dissolved oxygen concentration of 6.2 mg / L at the same timestamp. At T=14:30:06, the system received the next set of data, showing a stirring motor current of 0.04 amperes, still indicating a shutdown state, but the dissolved oxygen concentration reading was 6.8 mg / L. When processing the data at T=14:30:06, the processing unit calculated the dissolved oxygen concentration change rate to be 0.6 mg / L / s, which is greater than the calibrated natural diffusion fluctuation threshold of 0.05 mg / L / s. At this time, the motor current of 0.04 amperes indicates a shutdown state. This data combination shows a situation where the dissolved oxygen concentration fluctuates drastically under shutdown conditions, which does not meet the preset parameter logic association rules.

[0046] The system determined that the dissolved oxygen data at T=14:30:06 had a logical authenticity defect, which could stem from a momentary jump in the sensor signal or data interference. The system accordingly executed an interception operation, prohibiting the updating of this dissolved oxygen data and any potential compliance risk increments it generated to the regular numerical indicators in the batch cumulative compliance matrix. Simultaneously, the system incremented the independent data credibility anomaly count by 1 in the batch cumulative compliance matrix for batch number MJ-2025-10C. This verification utilizes physical logical mutual verification between parameters to automatically review and isolate suspicious data before it enters the compliance calculation process, preventing invalid data from contaminating electronic batch records and cumulative risk quantification results, thus ensuring the credibility of the monitoring data.

[0047] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things, characterized in that, The method runs on a data processing system and includes the following steps: A batch cumulative compliance matrix is ​​constructed in memory for the production batch of giant salamander active peptide. The batch cumulative compliance matrix contains numerical indicators to characterize the cumulative compliance status of the production batch throughout the entire production cycle. The system receives environmental data streams and context streams in parallel. The environmental data streams contain environmental parameters with physical timestamps, and the context streams contain process change events with timestamps indicating when they occurred. Based on the physical timestamp of the environmental parameters, determine the first process stage to which the environmental parameters belong on the context timeline constructed by the context flow, and load the first judgment rule corresponding to the first process stage; Based on the first judgment rule, the first compliance risk increment generated by environmental parameters is calculated and added to the numerical index of the batch cumulative compliance matrix; The method also includes a backtracking recalculation step: when the timestamp of the received process change event is earlier than the physical timestamp of the environmental parameters that have been processed according to the first judgment rule, the batch cumulative compliance matrix is ​​locked and the environmental parameters that have been processed since the timestamp of the process change event are retrieved. Based on the first judgment rule, calculate the original contribution value of the environmental parameters that have been processed since the occurrence timestamp of the process change event to the numerical index, and deduct the original contribution value from the batch cumulative compliance matrix. Based on the second judgment rule corresponding to the second process stage indicated by the process change event, the second compliance risk increment of the processed environmental parameters is recalculated and added to the batch cumulative compliance matrix. Furthermore, the method also includes a dual-track judgment step for process transition period: when the context flow indicates that the process stage is switching from the first process stage to the second process stage, a logical transition time window covering the switching time is determined based on preset process physical inertia parameters; for environmental parameters whose physical timestamps are located within the logical transition time window, a first judgment rule and a second judgment rule are loaded; the compliance status of the environmental parameters is calculated based on the first judgment rule and the second judgment rule respectively; only when the environmental parameters are determined to generate compliance risk increments under both the first and second judgment rules is the compliance risk increment updated to the batch cumulative compliance matrix; The context stream is generated and sent by the production execution system. Process change events include cleaning start event, enzymatic hydrolysis start event, inactivation start event, and filtration start event. The environmental data stream is collected and sent by an IoT sensor network distributed on the production site. Environmental parameters include reactor temperature, solution pH value, and stirring motor current.

2. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The retrospective recalculation step also includes: after completing the deduction of the original contribution value and the accumulation of the second compliance risk increment, generating a status correction log, which records the original contribution value, the second compliance risk increment, and the system time of executing the retrospective recalculation step; associating the status correction log with the electronic batch record of the production batch; unlocking the batch cumulative compliance matrix and restoring the processing of the real-time environmental data stream.

3. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The steps for calculating the first compliance risk increment caused by environmental parameters follow the following calculation logic: ,in This represents the first increase in compliance risk. These are the measured values ​​of environmental parameters. The compliance benchmark value specified in the first judgment rule. The time difference between the environmental parameters and the previous sampling point. This is the risk weighting coefficient associated with the first process stage.

4. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The dual-track judgment steps for the process transition period also include: if the environmental parameter is determined to have not generated an incremental compliance risk under either the first or second judgment rule, an exemption mark is generated; the exemption mark is associated with the environmental parameter; the risk value generated by the environmental parameter is prohibited from being accumulated to the numerical indicator; and the transition period exemption count is added to the batch cumulative compliance matrix.

5. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The method also includes a parameter logic topology verification step: preset parameter logic association rules in the rule base, which define the numerical linkage relationship or state mutual exclusion relationship that must be satisfied between the first type of environmental parameter and the second type of environmental parameter; before performing the step of calculating the first compliance risk increment, obtain the associated environmental parameter with the same physical timestamp as the currently processed environmental parameter; determine whether the combination of the environmental parameter and the associated environmental parameter satisfies the parameter logic association rules; if not, intercept the environmental parameter, prohibit the update of the batch cumulative compliance matrix, and synchronously accumulate the data credibility anomaly count in the batch cumulative compliance matrix.

6. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The numerical indicators in the batch cumulative compliance matrix include the cumulative heat dose indicator and the cumulative pH stress indicator; The cumulative heat dose index is used to characterize the total risk of thermal degradation borne by the active peptides of the giant salamander throughout the entire production cycle; The cumulative pH stress index is used to characterize the total integral value of the acid-base environment change pressure that the active peptides of giant salamander are subjected to during the enzymatic hydrolysis and inactivation process. The step of calculating the first compliance risk increment generated by environmental parameters includes calculating the corresponding compliance risk increment according to the type of environmental parameter and adding it to the cumulative heat dose index or the cumulative pH stress index.

7. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 1, characterized in that, The method also includes a compliance status invalidation step: after each update of the batch cumulative compliance matrix, the numerical indicators are compared with a preset management termination threshold; If the numerical indicator exceeds the management termination threshold, a batch compliance failure instruction is generated; the status identifier of the production batch is changed to failure status, and the compliance calculation process for subsequent data of that production batch is terminated.

8. The method for full-cycle traceability and management of giant salamander active peptide production based on the Internet of Things as described in claim 5, characterized in that, In the parameter logic topology verification step, the first type of environmental parameter is the stirring motor current, and the second type of environmental parameter is the dissolved oxygen concentration. The parameter logic association rule is limited to the fact that when the stirring motor current indicator is in the off state, the rate of change of dissolved oxygen concentration must not exceed the preset natural diffusion fluctuation threshold. If the stirring motor current is zero and the rate of change of dissolved oxygen concentration exceeds the natural diffusion fluctuation threshold, it is determined that the combination of environmental parameter and associated environmental parameter does not meet the parameter logic association rule.

Images