Edible vegetable oil production whole process traceability data processing system

Abstract

The invention relates to the technical field of food safety monitoring and supply chain digital management, in particular to an edible vegetable oil production whole-process traceability data processing system, which comprises a traceability data management center used for obtaining a current quality state predicted value and an aging rate constant; the risk collision analysis unit is used for carrying out safety threshold comparison analysis on the collected current quality state prediction value and carrying out compliance judgment processing on the edible vegetable oil to be judged to obtain a safety retention signal or a risk interception signal; when the risk interception signal is generated, the dynamic decision execution unit is used for generating a freezing instruction and sending the freezing instruction to the warehouse management system; when the security retention signal is generated, the dynamic decision execution unit is used for obtaining dynamic shelf life data and mapping the dynamic shelf life data to a terminal display port; according to the invention, the technical problem that the traditional static traceability cannot sense the environment thermal history is solved, and the predictability and accuracy of the full-life-cycle quality control are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of food safety monitoring and supply chain digital management, in particular to a full-process traceability data processing system for edible vegetable oil production. BACKGROUND

[0002] In the digital supervision scene of modern food supply chain, the full life cycle of edible vegetable oil covers multiple discrete links such as raw material acquisition, processing production and warehouse logistics; the data generated by these links has the characteristics of multi-source heterogeneity, and the quality of oil products is significantly affected by dynamic physical field factors such as environmental thermal history and light, resulting in complex correlation between time and space dimensions of data; At present, the existing traceability data processing technology generally adopts a static recording mode based on a relational database, that is, by scanning barcodes to associate fixed production batch information and factory detection reports; although this way can realize path tracking of physical nodes, there is significant lag and logical discontinuity in data calculation level; the existing computing architecture lacks the modeling capability of the nonlinear evolution relationship between environmental flow data and physicochemical indicators, cannot deeply couple and dynamically deduce heterogeneous data, and leads to the system only being able to feedback historical state and being unable to perceive potential quality deterioration caused by logistics environment fluctuation, making it difficult to realize risk warning and dynamic decision-making based on data calculation before physicochemical indicators exceed the standard; therefore, how to perform quality evolution dynamics calculation based on deep fusion of multi-source heterogeneous data to realize accurate state prediction and active risk interception in the whole process has become a technical problem to be solved in the field. SUMMARY

[0003] To solve the above technical problems, the present application provides a full-process traceability data processing system for edible vegetable oil production, in particular, the technical scheme of the present application comprises: a traceability data management center, a multi-source data acquisition unit, a dynamics evolution modeling unit, a risk collision analysis unit and a dynamic decision-making execution unit; The traceability data management center is used to call the flow information of the full life cycle of edible vegetable oil, and send the flow information to the multi-source data acquisition unit for heterogeneous data extraction analysis to obtain raw material fingerprint data, process stress data and environmental flow data, and send the raw material fingerprint data, process stress data and environmental flow data to the dynamics evolution modeling unit for quality evolution coupling calculation to obtain current quality state prediction value and aging rate constant; The risk collision analysis unit is used for safety threshold comparison analysis on the collected current quality state prediction value, and compliance discrimination processing on the edible vegetable oil to be judged to obtain a safe retention signal or a risk interception signal; When the risk interception signal is generated, the dynamic decision execution unit is configured to perform supply chain blocking processing on the target batch of products, generate a freezing instruction, and send the freezing instruction to a warehouse management system. When the safety retention signal is generated, the dynamic decision execution unit is configured to quantitatively analyze the remaining safety window period by using an aging rate constant, obtain dynamic shelf life data, determine whether the dynamic shelf life data meet minimum sales cycle requirements, generate a freezing instruction if the dynamic shelf life data do not meet the minimum sales cycle requirements, and map the dynamic shelf life data to a terminal display port if the dynamic shelf life data meet the minimum sales cycle requirements.

[0004] Optionally, the heterogeneous data extraction and analysis process is as follows: The production batch code of the edible vegetable oil is obtained, and the initial acid value and moisture content are set as raw material fingerprint data based on the production batch code tracing to the raw material origin and the initial physicochemical test report. Key node parameters in the production link are obtained, and the pressing leaching temperature, refining deacidification alkali liquor concentration, and time-temperature integral of the deodorization section are set as process stress data. Continuous time series data in the warehouse and transportation process are obtained, and the temperature fluctuation curve, light intensity exposure time, and oxygen permeability of the packaging material in the logistics process are set as environmental flow data.

[0005] Optionally, the kinetic evolution coupling calculation process is as follows: The temperature fluctuation curve in the environmental flow data is obtained, thermodynamic equivalent conversion analysis of the temperature fluctuation curve is performed by using the Arrhenius equation, and the duration of different temperature sections is converted into equivalent thermal aging time at a standard reference temperature. Based on the time-temperature integral of the deodorization section in the process stress data, the consumption proportion of the oxidation induction period is calculated by calling a preset oxidation induction period loss model, the lipid oxidation reaction rate constant under the current environmental condition is solved by using a reaction rate equation in combination with the equivalent thermal aging time, and the lipid oxidation reaction rate constant is set as the aging rate constant.

[0006] Optionally, the current quality state prediction value acquisition process is as follows: The initial peroxide value in the raw material fingerprint data is obtained as the starting point of the reaction kinetics model. Based on the aging rate constant and the equivalent thermal aging time, non-linear fitting calculation is performed by using a free radical chain reaction growth model to simulate the cumulative growth trend of the peroxide value with time, and the calculated peroxide value simulation data at the current time is set as the current quality state prediction value.

[0007] Optionally, the safety threshold comparison and analysis process is as follows: The upper threshold value of the peroxide value is called from a preset safety standard database, and the upper threshold value is set as a safety critical value. The current quality state prediction value is compared with the safety threshold value in numerical size: If the current quality state prediction value is less than the safety threshold value, a safety retention signal is generated; If the current quality state prediction value is greater than or equal to the safety threshold value, a risk interception signal is generated.

[0008] Optionally, the quantitative measurement and analysis process is as follows: In response to the safety retention signal, the safety threshold value is obtained; The current quality state prediction value is used as the initial state, the reaction kinetics model in the kinetic evolution modeling unit is used, the real-time aging rate constant under the current storage environment is used as the evolution parameter, the time span required for the peroxide value to reach the safety threshold value is deduced, and the time span is set as the remaining shelf life; The remaining shelf life is superimposed with the current date to obtain the recommended best consumption date, and the recommended best consumption date is set as the dynamic shelf life data.

[0009] Optionally, the system further comprises an abnormal pattern recognition unit, which is configured to: Obtain a reference aging rate model of similar products, calculate the deviation degree of the aging rate constant of the current batch of products from the reference aging rate model; The deviation degree is compared and analyzed with a preset abnormal fluctuation threshold value, if the deviation degree exceeds the preset abnormal fluctuation threshold value, it is determined that there is a process defect or a risk of adulteration, a process warning signal is generated and sent to the production control center.

[0010] Optionally, the process stress data further comprises trans fatty acid generating factors; The kinetic evolution modeling unit is further configured to establish a correlation model of temperature-time-trans fatty acid generation amount, calculate the cumulative generation amount of trans fatty acid based on the thermal history data of the refining link, and include the cumulative generation amount of trans fatty acid as an auxiliary judgment index into the discrimination logic of the risk collision analysis unit.

[0011] Compared with the prior art, the present application has the following beneficial effects: 1、The present application solves the technical problem that the traditional static traceability cannot perceive the environmental thermal history by constructing a kinetic evolution modeling unit; the system converts the temperature fluctuation and light data in the logistics process into equivalent thermal aging time based on the Arrhenius equation and the oxidation induction period loss model, realizes real-time dynamic deduction of the physicochemical quality of edible vegetable oil. This mechanism can accurately capture potential quality deterioration caused by changes in environmental physical fields, significantly improving the predictability and accuracy of whole life cycle quality control; 2、The application constructs a deep fusion and decoupling mechanism of multi-source heterogeneous data, breaks through the logical fault caused by data dispersion in the prior art; by abstracting complex full life cycle information into three core elements of raw material fingerprint, process stress and environmental flow, the system successfully establishes accurate mapping of the physical world and the digital model; such data structuring not only ensures the physical and chemical meaning of the input parameters, but also realizes unified calculation of data of different dimensions, providing a solid data foundation for subsequent quality evolution analysis; 3、The application realizes dynamic decision-making and risk interception based on the current quality state, changing the rigid mode of traditional fixed shelf life; the system uses the aging rate constant to quantitatively calculate the remaining safety window period and generate dynamic shelf life data; under the premise of ensuring food safety, interception instructions are automatically generated for high-risk products that have undergone high-temperature transportation, and the best consumption date is established for well-stored products, effectively balancing the contradiction between supply chain risk blocking and reducing food waste; 4、The application introduces an abnormal pattern recognition and benchmark model comparison mechanism to improve the identification ability of process defects and potential adulteration behavior; by calculating the deviation degree of the current batch aging rate and the historical benchmark model, the system can deeply mine problem products that are temporarily qualified but have poor internal stability; this monitoring method based on statistical distribution overcomes the limitations of single-point detection, realizing deep inspection of production process consistency and reverse verification of raw material authenticity. BRIEF DESCRIPTION OF DRAWINGS

[0012] The application will be further explained in conjunction with the accompanying drawings and embodiments: Figure 1 is a structural diagram of the system of the application. DETAILED DESCRIPTION

[0013] In order to make the purpose, technical scheme and advantages of the application clearer, the application will be further described in detail in conjunction with specific embodiments.

[0014] Embodiment 1: Please refer to Figure 1 , the edible vegetable oil production whole-process traceability data processing system includes a traceability data management center, a multi-source data acquisition unit, a dynamics evolution modeling unit, a risk collision analysis unit and a dynamic decision-making execution unit; The traceability data management center is used to retrieve the flow information of the whole life cycle of edible vegetable oil, and send the flow information to the multi-source data acquisition unit for heterogeneous data extraction and analysis to obtain raw material fingerprint data, process stress data and environmental flow data; the raw material fingerprint data, process stress data and environmental flow data are sent to the dynamics evolution modeling unit for quality evolution coupling calculation to obtain the current quality state prediction value and the aging rate constant; The risk collision analysis unit is used for safety threshold comparison analysis on the collected current quality state prediction value, compliance discrimination processing on the edible vegetable oil to be judged, and obtaining a safe retention signal or a risk interception signal; When the risk interception signal is generated, the dynamic decision execution unit is used for supply chain blocking processing on the target batch product, generates a freezing instruction and sends it to the warehouse management system; when the safe retention signal is generated, the dynamic decision execution unit is used for quantitative measurement analysis on the remaining safe window period by using the aging rate constant, obtains dynamic shelf life data, judges whether the dynamic shelf life data meets the minimum sales cycle requirement, if not, generates a freezing instruction; if yes, maps the dynamic shelf life data to the terminal display port.

[0015] The embodiment details the specific architecture and operation logic of the edible vegetable oil production whole-process traceability data processing system, aiming to solve the problem that the existing traceability system cannot perceive the environmental heat history leading to internal physicochemical quality deterioration; the traceability data management center serves as a data hub, and retrieves whole-life cycle circulation information including production batch records and logistics tracks through an API interface; the multi-source data acquisition unit performs decoupling operation on the above circulation information, extracts three types of core elements determining oil quality: raw material fingerprint data, process stress data and environmental circulation data; the kinetic evolution modeling unit constructs a virtual simulation mapping, based on the principles of physical and chemical reaction kinetics, uses the above three types of data for quality evolution coupling calculation, and calculates the current quality state prediction value representing the current time edible oil internal accumulated peroxide value and the aging rate constant representing the speed of lipid oxidation reaction ; On this basis, the risk collision analysis unit performs compliance discrimination, compares with the preset safety standard, and generates a safe retention signal if the value is lower than the threshold, otherwise generates a risk interception signal; the dynamic decision execution unit executes control according to the signal: in response to the risk interception signal, generates a freezing instruction to lock the warehouse out-of-warehouse permission; in response to the safe retention signal, calculates the remaining safe time window by using , generates dynamic shelf life data and pushes it to the consumer terminal; The embodiment introduces the material change process of edible oil into the data traceability system by constructing a closed-loop processing system including a kinetic evolution modeling unit; to verify the technical effect of the system, two batches, i.e. batch A and batch B of the same brand of first-grade soybean oil, are selected for comparison test; batch A is connected to the system for whole-process data processing, and batch B adopts the traditional traceability method; In the simulation transportation experiment, the same environmental heat history, i.e. 40℃ for 48 hours, is applied to the two batches of products; the system calculates the aging rate constant from the baseline / Rising to / , the current quality state prediction value Although not exceeding the standard but close to the critical value, the system determines that the remaining shelf life is not enough to support terminal sales, and generates a risk interception signal; Batch B is temporarily qualified and is normally stored; Subsequent tracking detection shows that the peroxide value of batch B breaks through the upper limit of the national standard 10 / , causing a food safety accident during the shelf life, while batch A is successfully intercepted, proving the significant advantage of the system in risk early warning.

[0016] Example 2: The heterogeneous data extraction and analysis process is as follows: obtaining the production batch code of edible vegetable oil, tracing to the raw material origin and the initial physicochemical test report based on the production batch code, setting the initial acid value and moisture content as the raw material fingerprint data; Obtaining key node parameters in the production process, setting the pressing leaching temperature, refining deacidification alkali concentration and time-temperature integral in the deodorization section as process stress data; Obtaining continuous time series data in the storage and transportation process, setting the temperature fluctuation curve, light intensity exposure time and oxygen permeability of packaging materials in the logistics process as environmental flow data.

[0017] This embodiment specifically limits the heterogeneous data extraction and analysis process, aiming to ensure the accuracy of the parameters input into the model in terms of physical and chemical significance; The multi-source data acquisition unit obtains the production batch code, and traces to the raw material origin and the initial physicochemical test report, extracts the initial acid value , moisture content and initial peroxide value , and sets them as raw material fingerprint data, wherein and determine the potential of lipid hydrolysis, and serves as the starting point for subsequent kinetic evolution; The system captures key node parameters from the production control system, sets the pressing leaching temperature which affects the phospholipid dissolution, the refining deacidification alkali concentration which affects the soapstock residue and the time-temperature integral in the deodorization section which is related to the retention rate of natural antioxidants as process stress data; The system accesses vehicle and warehouse Internet of Things devices to obtain continuous time series data, sets the temperature fluctuation curve in the logistics process, light intensity exposure time and oxygen permeability of packaging materials as environmental flow data; The embodiment abstracts the complex production logistics process into a set of parameter sets that can be input into a mathematical model by accurately extracting key physicochemical variables; in particular, in the refining process control scene, the time-temperature integral and the packaging oxygen permeability are introduced, so that the model can comprehensively consider the cumulative contribution of heat history and oxygen permeation to the oxidation reaction, and significantly improve the accuracy of subsequent quality prediction.

[0018] Embodiment 3: The kinetic evolution coupling calculation process is as follows: Obtain the temperature fluctuation curve in the environmental flow data, and use the Arrhenius equation to perform thermodynamic equivalent conversion analysis on the temperature fluctuation curve, convert the duration of different temperature sections into equivalent heat aging time at a standard reference temperature; Based on the time-temperature integral of the deodorization section in the process stress data, call the preset oxidation induction period loss model to calculate the consumption proportion of the oxidation induction period, combine the equivalent heat aging time, and use the reaction rate equation to solve the lipid oxidation reaction rate constant under the current environmental condition, and set the lipid oxidation reaction rate constant as the aging rate constant.

[0019] The embodiment describes in detail the kinetic evolution coupling calculation process, which is the core step of converting environmental data into chemical reaction rate; the kinetic evolution modeling unit obtains the temperature fluctuation curve , calculates the equivalent heat aging time using the integral form of the Arrhenius equation, and the formula is as follows: wherein, is the equivalent heat aging time , is the total duration of temperature records , is the lipid oxidation activation energy, which is 85000 / , is the ideal gas constant 8.314 / ( ), is the real-time temperature fluctuation curve , is the standard reference temperature 298 , is the natural base; For the discrete data processing needs of the computer system, the above integral formula is converted into Riemann sum form for numerical calculation: wherein, is the total number of temperature monitoring points, is the sampling interval, The real-time temperature value of the first sampling point; The real-time temperature value of the first sampling point; The system calls a preset oxidative induction period loss model to calculate the consumption ratio of the oxidative induction period based on the deodorization time-temperature integral in the process stress data The specific structure of the model is established based on the thermal degradation kinetics of antioxidants, and the calculation formula is as follows: Among them, is the consumption ratio representation value, dimensionless, , is the sensitivity coefficient 0.8, is the deodorization thermal integral ( ), is the critical thermal integral threshold (5000 ), is the nonlinear index 1.5; Combined with the light intensity exposure time in the environmental flow data and the oxygen permeability of the packaging material , the standard reference state aging rate constant corrected by the environmental stress is calculated ; Since the subsequent calculation has adopted the equivalent thermal aging time , the temperature fluctuation has been normalized, so the rate constant here only needs to be corrected for non-thermal factors and no longer contains the Arrhenius index term of temperature, and the calculation formula is as follows: Among them, is the corrected standard reference state aging rate constant ( ), is the basic reaction rate constant at the standard reference temperature ( ) ; The average light intensity representing the logistics process ; is the oxygen permeability of the packaging material, with the unit of / ( ² ) is the photo-oxygen catalytic coefficient, with the value of , which is used to map the light and oxygen permeability into the reaction rate increment; ​This embodiment utilizes integral transformation to normalize complex non-constant temperature logistics processes into equivalent processes under standard conditions. In scenarios where cold chain interruptions or drastic fluctuations in ambient temperature transportation occur, this method solves the problem of difficulty in quantifying oxidation rates under varying temperature conditions. At the same time, it clearly provides an oxidation-induced period loss model and a mathematical structure for photo-oxidation coupling correction, quantifying the weakening effect of processing technology on antioxidant capacity and the catalytic effect of environmental photo-oxidation, making the calculated aging rate constant more consistent with the true chemical activity of oil products.

[0020] Example 4: The process for obtaining the current quality status prediction value is as follows: The initial peroxide value in the raw material fingerprint data is obtained as the starting point for the reaction kinetics model; Based on the aging rate constant and equivalent thermal aging time, a nonlinear fitting calculation was performed using a free radical chain reaction growth model to simulate the cumulative growth trend of peroxide value over time. The calculated peroxide value simulation data at the current moment was set as the predicted value of the current quality status.

[0021] This embodiment illustrates the process of obtaining the current quality status prediction value, namely, using a kinetic model to simulate the growth of the lipid oxidation chain reaction; the system obtains the initial peroxide value from the raw material fingerprint data. This is used as the starting point for the reaction kinetics model; the standard reference state aging rate constant calculated based on the preceding steps is used. and equivalent thermal aging time Nonlinear fitting calculations were performed using a free radical chain reaction growth model; the model structure includes linear terms characterizing single-molecule decomposition, i.e. And the autocatalytic nonlinear term characterizing branched chain reactions, i.e. The time square term reflects the accelerating effect of oxidation product accumulation on the reaction rate, and the specific calculation formula is as follows: in, The current quality status prediction value ( / ), Initial peroxide value ( / ), The corrected standard reference state aging rate constant is the one calculated in Example 3. This is the equivalent thermal aging time. The base is the natural number; Among them, the autocatalytic coefficient It is not a fixed constant, but rather based on the initial acid value in the raw material fingerprint data. and moisture content Real-time corrections are made to reflect the promoting effect of free fatty acids and water on the oxidation reaction. The calculation formula is as follows: in, This is the corrected autocatalytic coefficient. The base coefficient, in this embodiment, is taken as a value of , The hydrolysis promoting factor has a value of [value missing]. , Initial acid value, in units of , The value represents the moisture content, expressed as a percentage (%). It should be noted that when substituting values ​​into the above formula for calculation... and All values ​​are taken and dimensionless calculations are performed. For example, when the moisture content is 0.5%, Take 0.5; acid value is 2 / hour, Take 2; The system will calculate Output to the risk collision analysis unit; This embodiment establishes a nonlinear growth model that includes an autocatalytic term and introduces raw material fingerprint data to dynamically correct the model parameters. It not only considers the primary oxidation reaction but also simulates the explosive growth in the later stage of oxidation caused by the deterioration of the raw material background. In long-term storage scenarios, the system can accurately predict high-risk oil products that are currently qualified but are on the eve of an oxidation outbreak due to the high acid value of the raw materials, thus achieving more forward-looking quality control.

[0022] Example 5: The safety threshold comparison and analysis process is as follows: Retrieve the upper limit threshold for peroxide value from the preset safety standard database, and set the upper limit threshold as the safety critical value; Compare the current quality status prediction value with the safety threshold value: If the current quality status prediction value is less than the safety threshold, a safety retention signal is generated. If the current quality status prediction value is greater than or equal to the safety threshold, a risk interception signal is generated.

[0023] This embodiment details the safety threshold comparison and analysis process, which is the logical judgment step for the system to make automated decisions. Based on the product type, the system retrieves the upper limit threshold for peroxide value from the corresponding national standard in the preset safety standard database and sets it as the safety critical value. The calculated current quality status prediction value and Perform numerical comparison; respond to This indicates that the current physicochemical indicators of the oil are not exceeding the standards, and the system generates a safe retention signal, allowing the product to continue circulating in the supply chain; in response to This indicates that the oil has undergone severe oxidation and deterioration, and the system generates a risk interception signal, triggering subsequent blocking mechanisms; This embodiment transforms complex chemical prediction results into binary control signals; in large-scale logistics distribution scenarios, it realizes automated identification of food safety risks, avoids the subjectivity and lag of manual judgment, and ensures that unqualified products cannot enter the end market.

[0024] Example 6: The quantitative calculation and analysis process is as follows: In response to a security retention signal, the security threshold value is obtained; Using the current quality status prediction value as the starting state, and using the reaction kinetic model in the kinetic evolution modeling unit, with the real-time aging rate constant under the current storage environment as the evolution parameter, the time span required for the peroxide value to reach the safety critical value is deduced, and this time span is set as the remaining shelf life. The remaining shelf life is added to the current date to calculate the recommended best-before date, and this recommended best-before date is set as dynamic shelf life data.

[0025] This embodiment provides a detailed explanation of the quantitative calculation and analysis process, specifically how to calculate the dynamic shelf life; and how the system acquires the safety threshold in response to the safety retention signal. and current quality status prediction value Using the inverse operation of the reaction kinetics model, the real-time aging rate constant under the current storage environment is calculated. As an evolution parameter, estimate the remaining shelf life required for the peroxide value to increase from its current value to the critical value. It should be noted that although a complex model containing quadratic terms was used in the preceding quality traceability, a linearized approximation method based on the current instantaneous reaction rate was adopted here to meet the real-time requirements of the terminal calculation when predicting the remaining shelf life. The calculation formula is as follows: The denominator term represents the value based on the current time. The instantaneous oxidation reaction rate, This is the safety threshold value. This is a predicted value for the current quality status. Remaining shelf life The coefficient 48 is the product of the derivative coefficient 2 of the time square term and the conversion coefficient 24 of the hour to day, and is used to unify the units of time. The aging rate constant at the current storage temperature. , The base is the natural number; The method of obtaining it is as follows: inheriting the standard reference state aging rate constant, which includes antioxidant loss and photo-oxidation history effects, calculated in Example 3. Combined with real-time temperature of the storage environment Thermodynamic extrapolation is performed, and the calculation formula is as follows: This calculation ensures that the shelf life forecast accurately reflects the quality damage accumulated in the preceding processing and transportation stages of the oil product; the calculated... With current system date By performing cumulative calculations, a recommended optimal consumption date can be obtained. and will Set up dynamic shelf-life data to be pushed to the terminal for display; This embodiment completely changes the rigid model of fixed shelf life in the traditional sense. In the scenario of supermarket shelf management, the system can give a shorter recommended consumption period for oil products that have undergone high-temperature transportation, or confirm the safety of well-stored oil products. While protecting the health of consumers, it can minimize food waste caused by blindly discarding food.

[0026] Example 7: The system also includes an anomaly pattern recognition unit, which is used for: Obtain the benchmark aging rate model for similar products, and calculate the deviation between the aging rate constant of the current batch of products and the benchmark aging rate model. The deviation is compared with the preset abnormal fluctuation threshold. If the deviation exceeds the preset abnormal fluctuation threshold, it is determined that there is a process defect or adulteration risk, and a process early warning signal is generated and sent to the production control center.

[0027] This embodiment introduces an anomaly pattern recognition unit to detect potential process defects or adulteration. The system acquires a baseline aging rate model of similar products under standard storage conditions. This model is not a single static value, but a statistical distribution model built based on historical data. The construction method is as follows: retrieve aging rate constant data of all qualified batches from the production line in the past 12 months with a sample size greater than 1000, remove outliers, and retain values ​​falling within the specified range. Data within the specified range is used, and data points outside this range are removed. For the initial calculation of the mean, To initially calculate the standard deviation, a normal distribution is then fitted to obtain the corrected mean. and standard deviation , the mean Set as baseline value ; Calculate the aging rate constant of the current batch of products. Mean parameters of the baseline aging rate model deviation The formula is ;Will Abnormal fluctuation thresholds set based on model distribution characteristics Comparative analysis was conducted; among them, The logic of the setting makes full use of the statistical distribution information of the model, and calculates as For example, in this embodiment, the value calculated based on historical data is 0.15, which is 15%; in response to ,For example Significantly higher than The system detects an anomaly, which may mean that new oil with high polar components has been added to the raw materials or that the deodorization temperature during the refining process is too high, causing the natural antioxidants to be completely lost. The system generates a process warning signal and sends it to the production control center, prompting a process check. This embodiment utilizes dynamic parameters to conduct in-depth monitoring of product consistency and discloses the statistical construction method of the benchmark model and a clear outlier removal algorithm. In the supply chain quality audit scenario, monitoring based on the rate constant can better reflect the inherent stability of the product, effectively identify problematic products that are temporarily qualified but have extremely poor stability, and play an auxiliary role in process feedback and anti-adulteration.

[0028] Example 8: Process stress data also includes trans fatty acid formation factors; The kinetic evolution modeling unit is also used to establish a correlation model of temperature-time-trans fatty acid production. Based on the thermal history data of the refining process, it calculates the cumulative production of trans fatty acids and incorporates the cumulative production of trans fatty acids as an auxiliary judgment indicator into the discrimination logic of the risk collision analysis unit.

[0029] This embodiment further considers trans fatty acids as a key safety indicator, enhancing the system's risk assessment dimensions; the kinetic evolution modeling unit establishes a correlation model between temperature, time, and trans fatty acid production; and it utilizes historical thermal data from the refining process, particularly deodorization temperature. and time Calculate the cumulative amount of trans fatty acids produced. The calculation formula is as follows: in, The cumulative amount of trans fatty acids produced ( / 100g), It is the frequency factor of the isomerization reaction. , Deodorization process holding time , The activation energy for the cis-trans isomerization reaction is 125000. / , The ideal gas constant is 8.314. / ( ), The processing temperature for the deodorization process , The system will use the natural base. As an auxiliary judgment indicator, it is included in the risk collision analysis unit. If the set threshold is exceeded, the system will issue a risk warning even if the peroxide value is within acceptable limits. This embodiment solves the double-edged sword problem in the high-temperature deodorization process; in the scenario of refining process optimization, by quantitatively calculating the cumulative amount of trans fatty acids, the system can comprehensively evaluate the overall impact of thermal processing on oil safety and prevent the hidden danger of excessive trans fatty acids due to the one-sided pursuit of low peroxide value.

[0030] 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 data processing system for the entire process of edible vegetable oil production traceability, characterized in that, It includes a traceability data management center, a multi-source data acquisition unit, a dynamic evolution modeling unit, a risk collision analysis unit, and a dynamic decision execution unit; The traceability data management center is used to retrieve the circulation information of edible vegetable oil throughout its entire life cycle, and send the circulation information to the multi-source data acquisition unit for heterogeneous data extraction and analysis to obtain raw material fingerprint data, process stress data and environmental circulation data. The raw material fingerprint data, process stress data and environmental circulation data are then sent to the dynamic evolution modeling unit for quality evolution coupling calculation to obtain the current quality status prediction value and aging rate constant. The risk collision analysis unit is used to perform safety threshold comparison analysis on the collected current quality status prediction value, and to perform compliance judgment processing on the edible vegetable oil to be judged, so as to obtain a safety retention signal or a risk interception signal. When a risk interception signal is generated, the dynamic decision execution unit is used to block the supply chain of the target batch of products, generate a freeze command and send it to the warehouse management system. When a safe retention signal is generated, the dynamic decision execution unit uses the aging rate constant to quantitatively calculate and analyze the remaining safe window period to obtain dynamic shelf life data. It then determines whether the dynamic shelf life data meets the minimum sales cycle requirement. If it does not meet the requirement, a freeze command is generated; if it does meet the requirement, the dynamic shelf life data is mapped to the terminal display port.

2. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The heterogeneous data extraction and analysis process is as follows: Obtain the production batch code of edible vegetable oil, trace the origin of raw materials and the initial physicochemical test report based on the production batch code, and set the initial acid value and moisture content as the raw material fingerprint data; Key node parameters in the production process are obtained, and the pressing and leaching temperature, the concentration of refining and deacidifying alkali solution, and the time-temperature integral of the deodorization section are set as process stress data. Acquire continuous time-series data during warehousing and transportation, and set temperature fluctuation curves, light intensity exposure duration, and oxygen permeability of packaging materials as environmental flow data during the logistics process.

3. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The dynamic evolution coupling calculation process is as follows: The temperature fluctuation curves in the environmental flow data are obtained, and the thermodynamic equivalent transformation analysis of the temperature fluctuation curves is performed using the Arrhenius equation to convert the duration of different temperature ranges into the equivalent thermal aging time at the standard reference temperature. Based on the time-temperature integral of the deodorization section in the process stress data, the consumption ratio of the oxidation induction period is calculated by calling the preset oxidation induction period loss model. Combined with the equivalent thermal aging time, the lipid oxidation reaction rate constant under the current environmental conditions is calculated using the reaction rate equation, and the lipid oxidation reaction rate constant is set as the aging rate constant.

4. The edible vegetable oil production process traceability data processing system according to claim 3, characterized in that, The process for obtaining the current quality status prediction value is as follows: The initial peroxide value in the raw material fingerprint data is obtained as the starting point for the reaction kinetics model; Based on the aging rate constant and equivalent thermal aging time, a nonlinear fitting calculation was performed using a free radical chain reaction growth model to simulate the cumulative growth trend of peroxide value over time. The calculated peroxide value simulation data at the current moment was set as the predicted value of the current quality status.

5. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The security threshold comparison and analysis process is as follows: Retrieve the upper limit threshold for peroxide value from the preset safety standard database, and set the upper limit threshold as the safety critical value; Compare the current quality status prediction value with the safety threshold value: If the current quality status prediction value is less than the safety threshold, a safety retention signal is generated. If the current quality status prediction value is greater than or equal to the safety threshold, a risk interception signal is generated.

6. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The quantitative measurement and analysis process is as follows: In response to a security retention signal, the security threshold value is obtained; Using the current quality status prediction value as the starting state, and using the reaction kinetic model in the kinetic evolution modeling unit, with the real-time aging rate constant under the current storage environment as the evolution parameter, the time span required for the peroxide value to reach the safety critical value is deduced, and this time span is set as the remaining shelf life. The remaining shelf life is added to the current date to calculate the recommended best-before date, and this recommended best-before date is set as dynamic shelf life data.

7. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The system further includes an anomaly pattern recognition unit, which is used for: Obtain the benchmark aging rate model for similar products, and calculate the deviation between the aging rate constant of the current batch of products and the benchmark aging rate model. The deviation is compared with the preset abnormal fluctuation threshold. If the deviation exceeds the preset abnormal fluctuation threshold, it is determined that there is a process defect or adulteration risk, and a process early warning signal is generated and sent to the production control center.

8. The edible vegetable oil production process traceability data processing system according to claim 1, characterized in that, The process stress data also includes trans fatty acid formation factors; The dynamic evolution modeling unit is also used to establish a correlation model of temperature-time-trans fatty acid production. Based on the thermal history data of the refining process, it calculates the cumulative production of trans fatty acids and incorporates the cumulative production of trans fatty acids as an auxiliary judgment indicator into the discrimination logic of the risk collision analysis unit.

Images