An all-process traceability system for livestock products

Abstract

The present application relates to the technical field of livestock product traceability, and discloses a full-process traceability system for livestock products, which comprises a data aggregation interface module, a logic reconstruction verification engine and a traceability permission control module, accesses the breeding unit environment and individual data of the system, generates group cumulative energy and individual biomass data sets, calls a biological energy conversion model to calculate a theoretical biomass increment threshold value, and establishes individual relative position topology; a double interlocking arbitration logic is executed to verify in parallel whether the group actual increment conforms to the theoretical threshold value and whether the individual position change conforms to the biological inertia; and the asset transfer permission is controlled according to the double verification results; the present application uses group energy conservation and individual distribution inertia to construct a logic verification mechanism, identifies and blocks false traceability data chains that conform to the formal requirements but violate the biological evolution logic without increasing the cost of front-end sensing hardware, and improves the penetration and security of supply chain financial supervision.

Description

Technical Field

[0001] This invention relates to a full-process traceability system for livestock products, belonging to the field of livestock product traceability technology. Background Technology

[0002] Current livestock supply chain management and supervision technologies generally adopt a discrete node identification and recording mechanism. When biological assets are transferred to physical nodes such as immunization, weighing, or slaughter, automatic identification equipment is used to read the identity carrier and generate a structured log containing time, location, and identity code. This log is then uploaded to a central database or distributed ledger. This data processing model meets the basic inventory counting and path tracking needs in the logistics management of industrial standard parts or non-biological assets.

[0003] In supply chain finance supervision or high-value traceability scenarios for live biological assets, existing technologies face a fundamental logical contradiction between physical evolution and static identification. Biological assets have long-cycle continuous growth characteristics, and conventional data collection can only obtain sparse and discrete snapshots. Within the data blind spots between two recording nodes, existing data processing systems default to maintaining the assumption that holding the identification means holding the asset. This assumption ignores the risk of the asset being replaced, removed, or improperly disposed of, resulting in the inability of formal compliance at the data level to reflect the physical reality. Relying on improving physical anti-tampering technology or adding high-frequency sensors in actual agricultural production faces exponential increases in equipment costs, high maintenance difficulty, and limited battery life, which are engineering constraints that violate the cost-effectiveness principle of supply chain management. Even if information technology is used to aggregate data from each link, without the inherent logical verification of the data content, the data entry problem still cannot be solved. This leads to the paradox of true trust. For example, Chinese invention patent CN112070620A discloses a full-process traceability system for livestock products. By integrating positioning sensors, environmental monitoring sensors, and camera hardware, it constructs a comprehensive platform that includes quality monitoring, livestock traceability, and data processing. This platform enables multi-dimensional data collection and visualization of feeding, disease prevention, and transportation processes. However, the essence of this technical solution remains at the level of passively recording and mechanically storing external input data. Although it establishes a data transmission link, it lacks a logical closed-loop verification mechanism for the data content itself. It cannot reverse-engineer the authenticity of the data based on the conservation of biological energy or the laws of growth and metabolism. Once the data entered at the source is falsified, the weight gain is artificially inflated, or the physical entity is replaced by a "bathed cow" in the blind spot, the traceability system, which relies solely on formal compliance and lacks biological logical interlocking, cannot identify anomalies, leading to regulatory failure.

[0004] Therefore, the technical problem to be solved by this invention is how to utilize low-frequency discrete business data to construct a logical verification mechanism that does not rely on high-frequency hardware monitoring and can identify false data that violates the laws of biological evolution. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: a full-process traceability system for livestock products, the system comprising:

[0006] Data aggregation interface module, logic reconstruction verification engine, and source tracing permission control module;

[0007] The data aggregation interface module connects to the environmental control system and individual metering terminal of the physically enclosed aquaculture unit, performs time-series data cleaning operations, and generates a first-dimensional population dataset containing cumulative energy input parameters and a second-dimensional individual dataset containing discrete time-point biomass characterization parameters.

[0008] The logic reconstruction verification engine includes an energy conversion constraint unit and a topology consistency analysis unit. The energy conversion constraint unit calls a pre-set bioenergy conversion model to map the first-dimensional population dataset to the theoretical total biomass increment threshold of the physically enclosed aquaculture unit within the target time window. The topology consistency analysis unit performs statistical sorting operations on the second-dimensional individual dataset to establish the relative position topology index of each unique identifier in the population distribution. The logic reconstruction verification engine executes a double interlocking arbitration logic: in the first thread, it verifies whether the sum of the actual biomass increments of all unique identifiers is within the tolerance range of the theoretical total biomass increment threshold, and in the parallel processing second thread, it verifies whether the variation of the relative position topology index of the target unique identifier within the continuous time window is within the biological inertia constraint range.

[0009] The traceability and access control module responds to the double verification pass signal output by the logic reconstruction verification engine, activates the asset transfer permission instruction of the target's unique identifier in the distributed ledger; and responds to the verification failure signal of any thread, outputs asset freeze and audit warning instructions.

[0010] Preferably, the data aggregation interface module further includes a source ontology authenticity verification unit, which performs trend stripping operations on the first-dimensional group dataset to extract high-frequency residual sequences and calculates the multi-scale permutation entropy value of the high-frequency residual sequences; the source ontology authenticity verification unit performs the following logical judgment based on physical noise fingerprints, and its judgment condition is: ,in, The multi-scale permutation entropy value of the obtained high-frequency residual sequence is calculated. and These are the lower and upper limits of the entropy range, respectively, pre-calibrated based on preset physical sensor background thermal noise and mechanical vibration characteristic data; when When the above judgment conditions are met, the source ontology authenticity identification unit determines that the first dimension group dataset is synthetic data that was not physically collected, and sends a source blocking instruction to the logic reconstruction verification engine to prohibit the execution of double interlock arbitration logic based on the dataset.

[0011] Preferably, the logic reconstruction verification engine also includes a benchmark dynamic correction unit, which calculates the statistical distribution skewness index of the biomass characterization parameters of all unique identifiers in the physically enclosed aquaculture unit in real time, and monitors the temporal change characteristics of the statistical distribution skewness index; the benchmark dynamic correction unit generates a dynamic tolerance compensation coefficient in response to the detection of a unidirectional drift of the statistical distribution skewness index that conforms to the preset biological stress response model; when calculating the total threshold of theoretical biomass increment, the energy conversion constraint unit applies the dynamic tolerance compensation coefficient to nonlinearly correct the standard threshold output based on the bioenergy conversion model, thereby expanding the tolerance range in the first thread.

[0012] Preferably, the system also includes a cross-node transfer status bridging module, which, in response to an asset transfer request from a physically enclosed aquaculture unit, acquires the spatiotemporal environmental parameters of the transfer path and calculates the expected physiological negative gain range based on a biological stress loss model; the cross-node transfer status bridging module acquires the biomass characterization parameters received by the destination node and calculates its actual gain value relative to the starting node; before executing the double interlock arbitration logic, the logic reconstruction verification engine prioritizes comparing the actual gain value with the physiological negative gain range; when the actual gain value falls outside the physiological negative gain range, it determines that the asset transfer chain is broken and triggers the traceability permission control module to perform a locking operation.

[0013] Preferably, the bioenergy conversion model includes a nonlinear coupling relationship between the environmental temperature integral parameter and the feed conversion rate parameter. The energy conversion constraint unit obtains the environmental temperature and humidity integral data of the physically enclosed breeding unit input by the data aggregation interface module, dynamically adjusts the feed conversion rate parameter, and then corrects the theoretical total biomass increment threshold.

[0014] Preferably, when establishing the relative position topology index, the topology consistency analysis unit maps the biomass characterization parameters of all unique identifiers in the second-dimensional individual dataset to a standardized normal distribution curve, and determines the percentile value of each unique identifier in the standardized normal distribution curve; the biological inertia constraint interval is defined as the maximum allowable drift of the percentile value of an individual in a non-pathological state between two adjacent sampling periods, and the maximum drift is determined based on the compensatory growth limit rate model of the species.

[0015] Preferably, the data aggregation interface module includes a multi-protocol heterogeneous data acquisition adapter, which reads the sensor register values ​​of the environmental control system through the industrial bus protocol and reads the identity and weight data of individual metering terminals through the radio frequency identification protocol; the adapter performs timestamp alignment and data format standardization processing on the read data, and uniformly converts the original signals of different sampling frequencies into a first-dimensional group dataset and a second-dimensional individual dataset.

[0016] Preferably, the traceability access control module includes a smart contract execution unit that maintains a state machine for each unique identifier on the distributed ledger. The state machine only performs a transition operation from the breeding state to the circulation state when it receives a double verification pass signal, and generates an immutable digital certificate containing the current verification hash value. The digital certificate serves as the sole legal basis for downstream nodes to receive the asset.

[0017] Preferably, the logic reconstruction verification engine also includes an abnormal pattern classification unit, which analyzes the feature vector of abnormal data when the double interlocked arbitration logic verification fails; when the abnormal feature is that the total weight gain of the population is normal but the relative position topology index of the individual has random perturbations that exceed the confidence level, it is determined to be physical identifier replacement behavior; when the abnormal feature is that the total weight gain of the population exceeds the theoretical total biomass increment threshold and meets the linear amplification feature, it is determined to be data inflation behavior; and based on this, it outputs a clearly categorized audit warning instruction.

[0018] Preferably, the system is deployed in a layered architecture that includes edge computing nodes and cloud servers; the data aggregation interface module and the source ontology authenticity identification unit reside on the local edge computing node of the physically enclosed breeding unit to perform real-time data acquisition and preliminary authenticity screening; the logic reconstruction verification engine and the traceability permission control module reside on the cloud server to perform global logic deduction and permission management based on the aggregated first-dimensional group dataset and second-dimensional individual dataset.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] 1. In the whole-process traceability of livestock products, a dual data verification mechanism of energy conservation in closed space and topological inertia of population distribution is constructed. The theoretical biomass increment is deduced by using cumulative energy input parameters and pre-set conversion models. The relative position index of individuals in the population statistical distribution is calculated, and the micro-identity topological continuity constraint is established. The interlocking logic of macro total amount and micro position makes the logic anomaly triggered by the disruption of the population thermodynamic balance or statistical inertia for single-dimensional tampering or entity replacement behavior. This makes it possible for data fraudsters to forge complex data chains that conform to the laws of bioenergy conversion and fit the dynamic distribution characteristics of the population. Without relying on high-frequency continuous physical monitoring, through logical mutual verification at the data processing level, false traceability data that conforms to the format specifications but violates the physical evolution logic can be identified and eliminated.

[0021] 2. An ontological verification logic based on source residual entropy spectrum is introduced. The data aggregation interface extracts high-frequency residual sequences from the trend of the time-series data of the receiving group, calculates multi-scale permutation entropy values ​​to characterize the physical disorder of the signal, and uses the background thermal noise and mechanical vibration features inevitably introduced during the acquisition process of physical sensors as unforgeable fingerprints. The dimension of data authenticity identification is shifted from the numerical content logic layer to the signal waveform physical layer. Based on this, it distinguishes between discrete signals from real physical acquisition and smoothed signals from mathematical model fitting, ensuring that the basic data participating in subsequent energy conservation calculations has physical authenticity and blocking the fraudulent path of generating false benchmark data through algorithm collusion.

[0022] 3. Establish an adaptive benchmark dynamic correction mechanism for distribution skewness drift. The logic reconstruction verification engine monitors the skewness and kurtosis of population biomass characterization parameters in real time. When the distribution morphology is distorted in accordance with the preset biological stress response model, it automatically generates a dynamic tolerance coefficient to nonlinearly correct the total threshold of theoretical biomass increment. It uses the asymmetric evolution characteristics of population statistical morphology under stress to invert the physical environment state, improves the data processing system's ability to distinguish between malicious fraud and environmental damage in the absence of dedicated environmental sensors, ensures the sensitivity of abnormal data detection, and improves the system's adaptability to complex application scenarios by dynamically adapting to environmental changes. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the system logic reconstruction verification architecture and data flow of the present invention;

[0024] Figure 2 This is a graph showing the nonlinear correction relationship between the temperature and humidity index and the biomass threshold in this invention.

[0025] Figure 3 This is a fishbone diagram illustrating the hierarchical division and composition logic of the system functional modules of this invention. Detailed Implementation

[0026] The following embodiments are used to illustrate the present invention, but not to limit the scope of protection of the present invention.

[0027] This specific embodiment provides a full-process traceability system for livestock products, including a data aggregation interface module, a logic reconstruction and verification engine, and a traceability access control module. The core workflow is as follows: the data aggregation interface module accesses the physical sensing data of the breeding unit and generates a standardized dataset; the logic reconstruction and verification engine performs dual interlock verification on the dataset based on energy conservation and distributed inertia logic; the traceability access control module manages asset transfer permissions on a distributed ledger based on the verification results; the data aggregation interface module is deployed on the local computing node of the physically enclosed breeding unit to build a reliable data input foundation. This module is equipped with a multi-protocol heterogeneous data acquisition adapter, which reads the sensor register values ​​of the environmental control system through the industrial bus protocol, including the cumulative feed amount of the feed line controller and the integral data of environmental temperature and humidity; at the same time, it reads the unique identification of the individual metering terminal and the weight data of the electronic weighbridge through the radio frequency identification protocol. The adapter performs a timestamp alignment operation on the above raw signals to generate a first-dimensional population dataset containing cumulative energy input parameters and a second-dimensional individual dataset containing discrete time-point biomass characterization parameters.

[0028] To prevent the injection of false data generated by the algorithm, the source ontology authenticity identification unit within the data aggregation interface module executes a verification procedure based on the physical texture of the signal. This procedure utilizes the background thermal noise and mechanical vibration characteristics introduced by physical sensors during analog signal acquisition as verification criteria. Specifically, the unit receives high-frequency time-series data from the first-dimensional group dataset, i.e., the instantaneous weighing or current reading sequence of the automated material line. The unit performs a trend stripping operation on this sequence, i.e., uses differential processing to filter out low-frequency trend terms and extract the high-frequency residual sequence. Next, the unit calculates the multi-scale permutation entropy value of this high-frequency residual sequence. Finally, the logical judgment is executed, and... With the preset entropy range Comparison was performed; among them The numerical range is pre-calibrated based on the inherent thermal noise and mechanical vibration characteristics of the connected physical sensors. This indicates that the data sequence lacks the chaotic characteristics of physical sampling. The source ontology authenticity identification unit determines that it is synthetic data and sends a source blocking instruction to the logical reconstruction verification engine to prohibit the execution of subsequent verification logic based on this dataset. The logical reconstruction verification engine is the core processing unit of the system, which includes an energy conversion constraint unit and a topology consistency analysis unit, and is used to perform logical deductions for the total population and individual position in parallel.

[0029] The energy conversion constraint unit is used to verify the physical conservation of energy input and output in the population. This unit calls the bioenergy conversion model to establish a nonlinear mapping relationship between the cumulative energy input parameters and the theoretical biomass increment. The process is as follows: This unit acquires the environmental temperature and humidity integral data of the physically enclosed aquaculture unit and dynamically adjusts the feed conversion ratio parameter in the model accordingly to reflect the impact of environmental heat stress on metabolic efficiency. Using the adjusted feed conversion ratio parameter and the cumulative energy input parameter in the first-dimensional population dataset, the unit calculates the total theoretical biomass increment threshold of the aquaculture unit within the target time window. In the arbitration logic of the first thread, the logic reconstruction and verification engine calculates the sum of the actual biomass increments of all unique identifiers within the aquaculture unit and verifies whether this sum is within the theoretical biomass increment threshold and its allowable tolerance. Within the range of variation; the topological consistency analysis unit is used to verify the biological inertia of individuals in the population distribution. This unit establishes the relative position topological index of each unique identifier in the population. The processing procedure is that this unit performs statistical sorting operations on all biomass characterization parameters in the second dimension individual dataset, maps the value of each individual to a standardized normal distribution curve, and determines its corresponding percentile value, i.e., the relative position topological index. In the arbitration logic of the second thread, the system verifies the variation range of the relative position topological index of the target unique identifier within a continuous time window. The system compares this variation range with the biological inertia constraint interval. The biological inertia constraint interval is determined based on the compensatory growth limit rate model of the species and is used to define the maximum positional drift that an individual is allowed to undergo under non-pathological conditions.

[0030] The logic reconstruction verification engine also includes a benchmark dynamic correction unit, used to dynamically adjust the verification benchmark in non-steady-state environments. This unit calculates the statistical distribution skewness index of biomass characterization parameters for all unique identifiers within the physically enclosed aquaculture unit in real time and monitors its temporal changes. When a unidirectional drift of the statistical distribution skewness index that conforms to a preset biological stress response model is detected, it indicates that the population is suffering from systemic environmental stress. This unit generates a dynamic tolerance compensation coefficient, which the energy conversion constraint unit applies to nonlinearly correct the theoretical total biomass increment threshold, thereby expanding the tolerance range in the first thread and avoiding false alarms caused by environmental factors. The system also includes a cross-node flow status bridging module, used to verify the continuity of assets during physical space transfer. This module responds to asset transfer requests, obtains the spatiotemporal environmental parameters of the transfer path, and calculates the expected physiological negative gain interval based on the biological stress loss model. When the asset arrives at the destination node and is remeasured, this module calculates its actual gain value relative to the transfer starting node. Before executing the double-lock arbitration logic, the logic reconstruction verification engine first compares the actual gain value with the physiological negative gain range. If the actual gain value falls outside this range, indicating that reasonable physiological loss is not reflected, the system determines that the asset circulation chain is broken and triggers a locking operation. The traceability and access control module manages asset permissions based on the output of the logic reconstruction verification engine. The smart contract execution unit within this module maintains a state machine for each unique identifier on the distributed ledger. When it receives a double verification pass signal, i.e., the actual increase of the population meets the theoretical threshold and the change in individual position conforms to biological inertia, the state machine executes the transition from the breeding state to the circulation state and generates a digital certificate containing the current verification hash value. If any thread fails verification, the system outputs an asset freeze command. In addition, if the total weight gain of the population is normal but the individual relative position topology index has random perturbations, it is determined to be a physical identifier replacement behavior. If the total weight gain of the population exceeds the theoretical threshold and conforms to the linear amplification characteristic, it is determined to be a data inflation behavior, and a corresponding audit warning command is output.

[0031] Example 1: In a physically enclosed breeding unit scenario of a Black Angus beef cattle fattening farm, the system operates within a supply chain finance regulatory environment for high-value biological assets. This unit houses 50 beef cattle in their critical fattening period, and there is a data blind spot lasting several weeks between two discrete physical weighing nodes. Within this blind spot, existing technologies rely solely on the holding status of static identity markers to infer the asset's sustainability, posing a risk of non-compliant replacement of the physical entity—that is, retaining the compliant identifier but replacing the underlying asset. The present invention's full-process traceability system for livestock products connects to the automated feed line controller and environmental monitoring terminal of this breeding unit, generating a first-dimensional population dataset in real time, including daily cumulative feed intake and environmental temperature and humidity integrals. This is combined with a second-dimensional individual dataset collected periodically by individual metering terminals, constructing a data verification closed loop based on biological logic, and performing logical reconstruction verification. The energy conversion constraint unit in the engine calls the bioenergy conversion model specifically for Angus cattle, calculates the heat stress correction factor by combining the real-time environmental temperature and humidity integral, and maps the cumulative energy input parameters in the first-dimensional population dataset to the theoretical total biomass increment threshold of the breeding unit within the target time window. When the system monitors that the total feed input of the breeding unit within a 30-day time window is 18,000 kg, and the environmental temperature and humidity integral shows that it is in the suitable growth range, the theoretical total biomass increment threshold derived by the model is 2,400 kg. At this time, the logic reconstruction and verification engine summarizes the actual biomass increment of all unique identifiers in the breeding unit in the first thread. If the actual total reaches 3,500 kg, exceeding the theoretical total biomass increment threshold and its preset tolerance range, it is determined that there is illegal mixing of external adult cattle or data falsification.

[0032] The topology consistency analysis unit performs statistical sorting operations on the second-dimensional individual dataset to establish the relative position topology index of each unique identifier in the population distribution. Taking the unique identifier CN-2025-X as an example, the system records that the relative position topology index in the previous monitoring period was at the 15th percentile of the population distribution, which is the bottom 20% of the low weight range. However, in the current monitoring period, the biomass characterization parameter corresponding to this identifier jumps to the 90th percentile of the population distribution. The logic reconstruction and verification engine calculates the magnitude of this position jump in the second thread and compares it with the biological inertia constraint range determined based on the Angus cattle compensated growth limit rate model. Since the magnitude of this jump exceeds the biologically allowed natural growth limit, the system determines that the physical entity corresponding to this unique identifier has undergone irregular permutation, thus constraining and preventing data inflation at the population level.

[0033] Example 2: This example presents a systematic experiment conducted in a controlled industrial environment. The experimental platform was built in a standardized closed-loop cattle breeding simulation chamber, equipped with a programmable automated feeding system, an environmental control system, and individual weighing terminals. This realistically simulates the material input and biological output processes in a real-world cattle breeding environment. The first dimension of the population dataset, including daily cumulative feed intake and integrated environmental temperature and humidity, is collected in real-time by the automated feed line controller and environmental monitoring terminal. The second dimension, the individual dataset, i.e., discrete individual weight data, is automatically acquired at preset time points by an RFID channel and an electronic weighbridge. To ensure the authenticity and representativeness of the experimental data, all sensors underwent industrial-grade calibration, and random noise and measurement errors consistent with actual working conditions were introduced during the experiment, such as dynamic fluctuations in feed weighing and non-uniform distribution of environmental temperature and humidity. To comprehensively evaluate the system's technical effectiveness, this experimental design includes... The invention includes a multi-dimensional comparison system with a sample group and a control group. The control group uses traditional traceability logic based on static identity identifiers, which only verifies the holding status and circulation compliance of identity identifiers, without performing logical verification of energy conservation and topological consistency. The sample group of this invention, on the other hand, fully deploys a logical reconstruction verification engine including an energy conversion constraint unit and a topological consistency analysis unit. The experiment simulates a 60-day fattening cycle, during which three typical attack scenarios are systematically introduced: first, the data inflation scenario, which artificially increases the total weight gain of the herd by 20% by modifying database records while keeping the feed amount constant; second, the individual replacement scenario, which replaces two low-weight individuals with high-weight individuals in the middle of fattening to simulate bathing behavior; and third, the environmental stress scenario, which simulates a continuous high-temperature heat stress environment from day 30 to day 40, leading to a decrease in feed intake and stagnation of weight gain in the herd.

[0034] The experimental process and key data records are as follows: In the data inflation scenario, when the nominal total weight gain of the population exceeds the theoretical biomass increment threshold calculated based on the actual feed input by 15%, the energy conversion constraint unit of the present invention successfully triggers the first thread verification failure signal. In contrast, the control group failed to identify this logical contradiction and still determined that the data was compliant. In the individual replacement scenario, although the weight data of the replaced individual numerically conforms to the adult standard of the breed, the topology consistency analysis unit of the present invention detected that the relative position topology index corresponding to the identifier has a jump exceeding the biological inertia constraint range within the continuous monitoring period, thereby triggering the verification failure signal in the second thread. In the environmental stress scenario, the benchmark dynamic correction unit detected that the population weight distribution skewness has a unidirectional drift that conforms to the heat stress model, automatically generates a dynamic tolerance compensation coefficient, corrects the theoretical threshold, thereby avoiding false alarms caused by environmental factors and ensuring the transfer rights of compliant assets. See Table 1, which lists the key monitoring data under the data inflation scenario.

[0035] Table 1: Comparison of Key Verification Parameters in Data Injection Scenarios

[0036]

[0037] The above experimental results show that by constructing a dual interlocking mechanism of energy conservation and distributed inertia, the present invention can effectively identify and block false traceability data that meets formal requirements but violates biological logic. In particular, when faced with fraudulent behaviors such as data inflation and individual substitution that cannot be detected by traditional technologies, the system demonstrates deterministic logical penetration. At the same time, the introduction of the benchmark dynamic correction mechanism improves the system's adaptability and stability in non-steady-state environments.

[0038] Example 3: This example combines Figures 1 to 3 A description of a full-process traceability system for a livestock product, such as... Figure 1 As shown, the system's logical architecture begins with a physically enclosed aquaculture unit. This unit houses an environmental control system and individual metering terminals. Data aggregation interfaces are used to access and generate population and individual-level datasets, while simultaneously performing time-series data cleaning. At the front end of the data link, a source authenticity verification unit screens the data based on a physical noise fingerprint entropy value determination mechanism, aiming to block synthetic data. If the verification passes, high-frequency residual sequences are allowed to enter the logical reconstruction verification engine; otherwise, a source blocking command is triggered, leading to asset freezing and audit warning processes. The logical reconstruction verification engine integrates a benchmark dynamic correction unit, an energy conversion constraint unit, and a topology consistency unit. The system comprises a performance analysis unit, a benchmark dynamic correction unit responsible for monitoring statistical skewness and correcting thresholds, an energy conversion constraint unit responsible for calculating the total threshold energy conservation logic of theoretical biomass increment at the subject level, and a topological consistency analysis unit responsible for establishing the biological inertia logic of individual relative position topological indicators at the micro level. These logic units converge to a dual-interlocked arbitration logic module to perform parallel verification and logical deduction. If the output verification failure signal is received, the system executes asset freezing, audit warning, or prompts a logical anomaly. If the output verification passes signal, the smart contract execution unit in the traceability permission control module is triggered, ultimately activating asset transfer permissions and generating an immutable digital certificate.

[0039] like Figure 2 As shown, the left vertical axis represents the biomass threshold in kg, the right vertical axis represents the temperature and humidity index (THI), and the horizontal axis represents the time scale from 0 to 60. The chart plots three curves that change over time: the standard theoretical threshold (kg) represented by the dashed line, the dynamically corrected threshold (kg) represented by the solid line, and the temperature and humidity index (THI) represented by the dotted line. Observing the curve trends, it can be seen that as the temperature and humidity index rises and reaches its peak between day 30 and day 40, the dynamically corrected threshold shows a significant downward deviation compared to the linearly increasing standard theoretical threshold. This non-linear correction relationship intuitively reflects the system's dynamic compensation to the theoretical weight gain expectation based on environmental pressure. Figure 3As shown, the first skeleton above the main system core target corresponds to the data aggregation and authenticity verification module, with three sub-items below it: source ontology authenticity verification, physical noise fingerprint verification, and multi-scale permutation entropy calculation. The second skeleton above corresponds to the logic reconstruction verification engine, which includes three core components: energy conversion constraint unit, topology consistency analysis unit, and double interlock arbitration logic. The first skeleton below belongs to the environment adaptation and flow bridging module, which is specifically expanded into a benchmark dynamic correction unit, cross-node flow state bridging, and physiological negative gain model. The second skeleton below corresponds to the traceability permission and trust control module, which covers a smart contract execution unit, anomaly mode classification unit, and asset freezing and audit early warning functions.

[0040] Example 4: This example, by clearly defining the difference order, entropy range, attenuation logic of the correction factor, and position drift threshold, transforms abstract biological logic into a deterministic engineering procedure that can be rigorously executed by a computer program. During system initialization, the data aggregation interface module performs source ontology authentication on the received first-dimensional group dataset. For the time-series load signal transmitted via the automated material line, the system employs a second-order difference algorithm to perform trend stripping operations. To eliminate low-frequency fluctuations in the feeding cycle and extract high-frequency residual sequences, the system is configured with an embedding dimension of 5 and a time delay of 1. The system calculates the multi-scale permutation entropy value of the residual sequence. Based on the pre-calibration of the background noise characteristics of the physical sensors, the system sets 0.6 to 0.9 as the legal physical chaos interval. Only when the calculated entropy value falls within this interval is the data stream considered to have physical authenticity and allowed to enter the subsequent energy conversion constraint unit.

[0041] In the energy conservation calculation stage, the energy conversion constraint unit executes dynamic threshold generation logic based on environmental heat stress. The system pre-sets the feed conversion ratio benchmark for Angus cattle at a standard ambient temperature (20°C). ,in The empirical coefficient is set to 0.02, representing the corrected feed conversion ratio. The system divides the accumulated energy input parameter by The theoretical total biomass increment threshold is obtained. In the topology consistency verification stage, the topology consistency analysis unit executes quantile tracking logic based on normal distribution. The system standardizes the biomass characterization parameters (body weight) collected by all unique identifiers within the physically enclosed aquaculture unit in the current time window into Z-scores and obtains the corresponding cumulative probability density values ​​by looking up a table. This value is the relative position topological index, and the biological inertial constraint interval is defined as the maximum allowable rate of change of this index between two adjacent sampling periods, such as 24 hours. Based on the compensated growth limit model of Angus cattle, the system sets When the rate of change of the position index of the target's unique identifier The logic refactoring verification engine will only generate a verification pass signal when the total actual increment of the group does not exceed the theoretical threshold after environmental correction.

[0042] Example 5: In this example, under non-standard environmental conditions, a pre-deployment calibration and parameter adaptation procedure is performed to ensure the accurate execution of the system's core algorithm logic. Before the system is actually deployed in any new physically enclosed aquaculture unit, an environmental baseline calibration procedure is executed. A portable high-precision environmental monitoring instrument deployed at key points in the aquaculture unit continuously collects temperature and humidity data for at least 72 hours and compares it synchronously with the integrated temperature and humidity data output by the automated environmental control system. The system uses a linear regression algorithm to calculate the deviation coefficient between the two and adjusts the register calibration parameters of the environmental monitoring terminal to ensure that the error range of the environmental data input to the energy conversion constraint unit is controlled within ±0.5℃ (temperature) and ±2% (humidity). This calibration step eliminates environmental data drift caused by sensor aging or improper installation location.

[0043] In addition, to address individual differences among different batches of biological assets, the system executes a fine-tuning adaptation procedure for biological model parameters. During the first 7-day observation period after each batch of biological assets is put into the pen, the system puts the energy conversion constraint unit in learning mode. In this mode, the system does not trigger a verification failure signal, but instead uses the actual collected cumulative energy input parameters and the actual biomass increment data of the population to backfit the actual feed conversion rate benchmark value of the current batch using the least squares method. When the fitting result deviates from the system's preset standard benchmark value by more than 5%, the system automatically updates the model parameters of the current batch and generates an adaptation log. At the same time, it uses the individual position data during the observation period to calibrate the upper and lower limit thresholds of the biological inertia constraint interval to adapt to the growth characteristics of specific varieties or batches.

[0044] Example 6: This example ensures the universality and accuracy of the logic reconstruction verification engine in diverse application scenarios by executing standardized parameter calibration and model initialization procedures. Before the system is officially deployed, a physical environment baseline calibration procedure is executed, requiring at least three high-precision temperature and humidity reference sensors to be evenly distributed within the physically enclosed aquaculture unit, continuously collecting environmental data for no less than 72 hours. The system uses a linear regression algorithm to establish a mapping function between the reference sensor data and the output data of the on-site environmental control system. The system automatically calculates the gain coefficient a and the bias coefficient b and writes them into the preprocessing configuration of the data aggregation interface module. This step aims to eliminate systematic errors caused by aging, drift, or improper installation of field sensors, and ensure that the environmental temperature and humidity integral data input to the energy conversion constraint unit has a unified physical reference, with the error range controlled within ±0.5℃.

[0045] In addition, considering the differences in growth characteristics among different biological species, an adaptive initialization procedure for biological model parameters is executed. During the first 10 days of the first fattening cycle, the system is in parameter self-learning mode. During this period, the system does not trigger abnormal alarms, but records the cumulative energy input parameters in real time. Compared with the actual biomass increment of the population Using the least squares method, the system fits the benchmark value of the actual feed conversion rate of the current batch of biological assets. , so that the objective function Minimize, if the fitted If the deviation from the system's preset standard variety parameters exceeds 10%, the system will automatically update the model baseline value, mark the batch as a specific growth population, and simultaneously calculate the variance of the fluctuation of the topological index of the relative position of individuals within the observation period. And set the threshold of the biological inertial constraint range to .

[0046] 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.

[0047] 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 full-process traceability system for livestock products, characterized in that the system... include: Data aggregation interface module, logic reconstruction verification engine, and source tracing permission control module; The data aggregation interface module connects to the environmental control system and individual metering terminal of the physically enclosed aquaculture unit, performs time-series data cleaning operations, and generates a first-dimensional population dataset containing cumulative energy input parameters and a second-dimensional individual dataset containing biomass characterization parameters of the species at discrete time points. The logic reconstruction verification engine includes an energy conversion constraint unit and a topology consistency analysis unit. The energy conversion constraint unit calls a pre-set bioenergy conversion model to map the first-dimensional population dataset to the theoretical total biomass increment threshold of the physically enclosed aquaculture unit within the target time window. The topology consistency analysis unit performs statistical sorting operations on the second-dimensional individual dataset to establish the relative position topology index of each unique identifier in the population distribution. The logic reconstruction verification engine executes a double interlocking arbitration logic: in the first thread, it verifies whether the sum of the actual biomass increments of all unique identifiers is within the tolerance range of the theoretical total biomass increment threshold, and in the parallel processing second thread, it verifies whether the variation of the relative position topology index of the target unique identifier within the continuous time window is within the biological inertia constraint range. The traceability and access control module responds to the double verification pass signal output by the logic reconstruction verification engine, activates the asset transfer permission instruction of the target's unique identifier in the distributed ledger; and responds to the verification failure signal of any thread, outputs asset freeze and audit warning instructions.

2. The full-process traceability system for livestock products according to claim 1, characterized in that, The data aggregation interface module also includes a source ontology authenticity verification unit, which performs trend stripping operations on the first-dimensional group dataset to extract high-frequency residual sequences and calculates the multi-scale permutation entropy value of the high-frequency residual sequences. The source ontology authenticity verification unit performs the following logical judgment based on physical noise fingerprints, and its judgment condition is as follows: ,in, The multi-scale permutation entropy value of the obtained high-frequency residual sequence is calculated. and These are the lower and upper limits of the entropy range, respectively, pre-calibrated based on preset physical sensor background thermal noise and mechanical vibration characteristic data; when When the above judgment conditions are met, the source ontology authenticity identification unit determines that the first dimension group dataset is synthetic data that was not physically collected, and sends a source blocking instruction to the logic reconstruction verification engine to prohibit the execution of double interlock arbitration logic based on the dataset.

3. The full-process traceability system for livestock products according to claim 1, characterized in that, The logic reconstruction verification engine also includes a benchmark dynamic correction unit, which calculates the statistical distribution skewness index of the biomass characterization parameters of all unique identifiers in the physically enclosed aquaculture unit in real time, and monitors the temporal change characteristics of the statistical distribution skewness index. The baseline dynamic correction unit generates a dynamic tolerance compensation coefficient in response to the detection of a unidirectional drift in the statistical distribution skewness index that conforms to the preset biological stress response model. When calculating the total threshold of theoretical biomass increment, the energy conversion constraint unit applies the dynamic tolerance compensation coefficient to nonlinearly correct the standard threshold output based on the bioenergy conversion model, thereby expanding the tolerance range in the first thread.

4. The full-process traceability system for livestock products according to claim 1, characterized in that, The system also includes a cross-node flow status bridging module, which responds to asset transfer requests from physically enclosed aquaculture units, obtains spatiotemporal environmental parameters of the transfer path, and calculates the expected physiological negative gain range based on a biological stress loss model. The cross-node transfer status bridging module obtains the biomass characterization parameters received by the destination node and calculates its actual gain value relative to the starting node. Before executing the double interlock arbitration logic, the logic reconstruction verification engine first compares the actual gain value with the physiological negative gain range. When the actual gain value falls outside the physiological negative gain range, it determines that the asset transfer chain is broken and triggers the traceability permission control module to perform a locking operation.

5. The full-process traceability system for livestock products according to claim 1, characterized in that, The bioenergy conversion model includes a nonlinear coupling relationship between the environmental temperature integral parameter and the feed conversion rate parameter. The energy conversion constraint unit obtains the environmental temperature and humidity integral data of the physically enclosed breeding unit input by the data aggregation interface module, dynamically adjusts the feed conversion rate parameter, and then corrects the theoretical total biomass increment threshold.

6. The full-process traceability system for livestock products according to claim 1, characterized in that, When establishing relative position topology indicators, the topology consistency analysis unit maps the biomass characterization parameters of all unique identifiers in the second-dimensional individual dataset to a standardized normal distribution curve and determines the percentile value of each unique identifier in the standardized normal distribution curve. The biological inertia constraint interval is defined as the maximum allowable drift of the percentile value between two adjacent sampling periods under non-pathological conditions. The maximum drift is determined based on the compensatory growth limit rate model of the species.

7. The full-process traceability system for livestock products according to claim 1, characterized in that, The data aggregation interface module includes a multi-protocol heterogeneous data acquisition adapter, which reads sensor register values ​​from the environmental control system via the industrial bus protocol and reads the identity and weight data of individual metering terminals via the radio frequency identification protocol. The adapter performs timestamp alignment and data format standardization processing on the read data, and uniformly converts the original signals of different sampling frequencies into a first-dimensional group dataset and a second-dimensional individual dataset.

8. The full-process traceability system for livestock products according to claim 1, characterized in that, The traceability and access control module includes a smart contract execution unit that maintains a state machine for each unique identifier on the distributed ledger. The state machine only performs the transition operation from the breeding state to the circulation state when it receives a double verification pass signal, and generates an immutable digital certificate containing the current verification hash value. The digital certificate serves as the sole legal basis for downstream nodes to receive the asset.

9. The full-process traceability system for livestock products according to claim 1, characterized in that, The logic reconstruction verification engine also includes an anomaly pattern classification unit, which analyzes the feature vector of the abnormal data when the double interlocked arbitration logic verification fails. When the abnormal feature is that the total weight gain of the population is normal but the relative position topology index of the individual has random perturbations that exceed the confidence level, it is judged as physical identifier replacement behavior. When the abnormal feature is that the total weight gain of the population exceeds the theoretical total biomass increment threshold and meets the linear amplification feature, it is judged as data inflation behavior. Based on this, it outputs clearly categorized audit warning instructions.

10. The full-process traceability system for livestock products according to claim 1, characterized in that, The system is deployed in a layered architecture that includes edge computing nodes and cloud servers; The data aggregation interface module and the source ontology authenticity identification unit reside on the local edge computing node of the physically enclosed breeding unit, performing real-time data acquisition and preliminary authenticity screening; the logic reconstruction verification engine and the traceability permission control module reside on the cloud server, performing global logic deduction and permission management based on the aggregated first-dimensional group dataset and second-dimensional individual dataset.

Images