Method for analyzing flavor dynamic change of sun vinegar aging period based on flavoromics

Abstract

The application discloses a kind of based on flavoromics's sun vinegar aging period flavor dynamic change analysis method, it is related to flavoromics technical field, to solve the current concentration rainstorm causes the temperature and humidity gradient of inner and outer wall of pottery jar to increase dramatically, causes the technical problem that pottery jar astringency is generated, including the following steps: S100, real-time monitoring and risk assessment of heat and humidity stress;S200, rainstorm event fusion early warning;S300, dynamic deduction and strategy optimization based on digital twinning;S400, dynamic microenvironment blocking control execution;S500, closed-loop feedback based on proxy index;S600, post-evaluation and model optimization.The application realizes early quantitative perception to astringency risk by real-time monitoring and calculating heat and humidity stress index representing ion precipitation driving force, combined with the research and judgment model of early warning confidence of fusion meteorological forecast and real-time data, starts the blocking control system, carries out physical intervention to pottery jar, guarantees the quality stability of sun vinegar in rainy season.

Description

Technical Field

[0001] This invention relates to the field of flavor omics technology, and more specifically, to a method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics. Background Technology

[0002] Traditional sun-dried vinegar production relies on earthenware jars for natural aging. These jars are made from local clay rich in minerals and have walls covered with natural micropores, giving them a unique "breathing effect." This promotes the exchange of substances and flavor transformation between the vinegar mash and the external environment, making them one of the core carriers for the mellow taste and unique flavor of sun-dried vinegar.

[0003] However, under subtropical monsoon climate conditions, the characteristics of high temperature and humidity, concentrated rainy season, and significant diurnal temperature differences, especially the frequent alternation of short-term heavy rainfall and persistent high temperature and humidity in summer, pose unique environmental challenges to the aging process of vinegar in earthenware jars. During and after rainstorms, the air humidity rapidly saturates, surface moisture evaporation intensifies, and the diurnal temperature difference can reach 10-12℃, resulting in a strong temperature and humidity gradient between the inner wall of the earthenware jar in contact with the vinegar mash and the outer wall in contact with the environment. Driven by this gradient, the microporous structure of the earthenware jar undergoes a violent cycle of moisture absorption, heat release, expansion, and contraction, significantly enhancing the dissolution of metal ions such as calcium, magnesium, and iron from the jar wall. These dissolved metal ions react in complex ways with flavor precursors such as phenolic substances, organic acids, and proteins in the vinegar mash, forming insoluble complexes or directly producing an astringent taste, ultimately leading to the "earthenware jar precipitate type of astringency". Once this astringent substance is formed, it is not only difficult to remove later, but it also masks the inherent mellow aroma and mild acidity of sun-dried vinegar, leading to a significant decline in product quality or even the scrapping of the entire batch. This has become a long-standing technical bottleneck restricting the large-scale and stable production of high-quality sun-dried vinegar in the production area. In view of this, we propose a method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology, adapt to practical needs, and provide a method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, so as to solve the technical problem that concentrated rainstorms cause a sharp increase in the temperature and humidity gradient between the inner and outer walls of ceramic jars, leading to the generation of astringent flavor in the ceramic jars.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, comprising the following steps:

[0006] S100, Real-time monitoring and risk assessment of thermal and humidity stress: Temperature and humidity sensors on the inner and outer walls of the aged ceramic jar are used to simultaneously measure the temperature and humidity of the inner and outer walls in real time and calculate the thermal and humidity stress index. Based on historical data, early warning thresholds and action thresholds are set to conduct risk assessment.

[0007] S200, Rainstorm Event Fusion Early Warning: Access meteorological forecast data, use the moving average method to preprocess rainfall forecast data and perform multi-source information fusion analysis, and generate a corresponding level of early warning signal when the confidence level exceeds the set threshold;

[0008] S300, Dynamic simulation and strategy optimization based on digital twin: Construct a digital twin containing a multi-physics coupling model and a reaction dynamics model, input meteorological data to predict the evolution trajectory of the thermal and wet stress index, and select the optimal control strategy through multi-strategy parallel simulation and Pareto optimization;

[0009] S400, Dynamic Microenvironment Resistance Control Execution: Responds to the early warning signal and starts the resistance control system, executes physical resistance control measures and dynamically adjusts based on the real-time thermal and moisture stress index;

[0010] S500, closed-loop feedback based on proxy indicators: real-time acquisition of pH value, conductivity and spectral data proxy indicators, dynamic adjustment of resistance control parameters and verification of resistance control effect;

[0011] S600, Post-event assessment and model optimization: Archive event data and analyze the contributing factors of astringent taste risk, and use the gradient descent method to iteratively optimize the weight coefficient of the thermal and humid stress index and the early warning threshold.

[0012] Preferably, step S100 further includes the following steps:

[0013] S101. Deploy a multi-parameter monitoring network: Install temperature and humidity sensors on the inner and outer walls of representative aged ceramic jars to simultaneously measure the inner wall temperature in real time. Inner wall humidity and outer wall temperature External wall humidity ;

[0014] S102. Calculate the thermal and moisture stress index and conduct a risk assessment: Based on the data collected in S101, the data processing center calculates the thermal and moisture stress index in real time. Set early warning thresholds based on historical data. With action threshold Conduct a risk assessment;

[0015] In step S102, the thermal and damp stress index The calculation formula is:

[0016] ;

[0017] in, The temperature difference weighting coefficient is the thermal and damp stress index. The humidity difference weighting coefficient is used. This refers to the temperature of the inner wall. Inner wall humidity; For the outer wall temperature, External wall humidity;

[0018] Warning threshold The calculation formula is:

[0019] ;

[0020] in, Astringency as a sensory odor; This represents the upper limit of acceptable astringency intensity. The acceptable low-risk probability threshold;

[0021] Action threshold The calculation formula is:

[0022] ;

[0023] in, This refers to the calcium ion concentration. This is the critical ion concentration threshold; This is a high-risk probability threshold that requires action.

[0024] Preferably, step S200 further includes the following steps:

[0025] S201. Data Acquisition and Preprocessing: Accessing meteorological forecast data, rainfall probability... Expected rainfall Rainstorm warnings were issued, and the moving average method was used to preprocess the rainfall forecast data;

[0026] S202. Execute integrated early warning and judgment: Use a quantitative confidence formula to comprehensively judge multi-source information, and when heavy rain is predicted or occurs and the thermal and humidity stress index... Exceeding the warning threshold At that time, a Level 1 warning signal is generated.

[0027] Preferably, in step S201, the formula for the moving average method is:

[0028] ;

[0029] in, To smooth out the predicted rainfall; For the first Raw hourly forecast rainfall for each hour; To adjust the sliding window size;

[0030] In step S202, the quantification confidence formula is: ;

[0031] in, For the confidence level of the early warning; This is the weighting coefficient for the reliability of weather forecasts; This represents the weighting coefficient for the current level of risk. The weighting coefficient for the trend of risk change; This is the current real-time thermal and moisture stress index; Action threshold; for Value change trend.

[0032] Preferably, step S300 further includes the following steps:

[0033] S301. Construct a digital twin of vinegar aging: The digital twin includes a multiphysics coupling model, a reaction kinetics model, and a control strategy library;

[0034] S302. Future Risk Simulation: Input future meteorological data and predict the thermal and moisture stress index using the digital twin. evolutionary trajectory ;

[0035] S303, Strategy Simulation and Optimal Selection: When a risk is predicted, multi-strategy parallel simulation is initiated, and the optimal control strategy is determined based on the Pareto optimal solution. .

[0036] Preferably, in step S303, the Pareto optimal solution is calculated using the following formula:

[0037] ;

[0038] in, This is the optimal strategy; For the first Various control strategies; For the first A comprehensive evaluation index for this strategy The threshold for the evaluation index.

[0039] Preferably, step S400 further includes the following steps:

[0040] S401. Start the barrier control system: In response to the first-level warning signal in step S202, automatically activate the barrier control equipment corresponding to the high-risk area;

[0041] S402. Implement physical barrier measures, including deploying the intelligent rainproof and breathable awning and activating the low-wind-speed micro-circulation air system; wherein, the awning deployment angle... and micro-circulation wind speed According to the thermal and damp stress index Dynamic adjustment;

[0042] S403, Adaptive Closed-Loop Control: Employs a PID control algorithm based on real-time data... The deviation between the current value and the target value is dynamically adjusted to control parameters.

[0043] Preferably, in step S403, the formula for the PID control algorithm is:

[0044] ;

[0045] in, for Wind speed adjustment over time. for Timing deviation; This is the proportionality coefficient. The integral coefficient; is the differential coefficient.

[0046] Preferably, step S500 further includes the following steps:

[0047] S501, Real-time Monitoring of Agent Indicators: Continuously monitor the conductivity and pH value of the vinegar mash using an online analyzer; and establish a conductivity change rate... With calcium ion precipitation rate Association model;

[0048] S502. Feedback Verification: Dynamically adjust the resistance intensity based on the proxy indicators and conduct rapid laboratory verification after the rainstorm ends.

[0049] Preferably, step S600 further includes the following steps:

[0050] S601. In-depth data archiving and causal analysis: Store complete event chain data in the case library and run a reverse tracing model to analyze the causes of risks;

[0051] S602. Input the case data into the neural network model and output the quantitative contribution percentage of each factor to the risk of astringency.

[0052] S603. Iterative optimization of system parameters: Using accumulated case data, the weight coefficients are periodically optimized using the gradient descent method. and And iteratively update the warning threshold. With action threshold .

[0053] Compared with the prior art, the beneficial effects of the present invention are:

[0054] 1. This invention achieves early quantitative perception of astringent taste risk by real-time monitoring and calculation of the thermo-humid stress index, which characterizes the driving force of ion precipitation. Combined with a judgment model that integrates weather forecasts and real-time data to assess the confidence level of early warning, it achieves accurate early warning. Finally, by activating a dynamic microenvironmental resistance control system, it physically intervenes in the ceramic jars to suppress the excessive precipitation of key metal ions that cause astringent taste, ensuring the quality stability of sun-dried vinegar during the rainy season and solving the problem of astringent taste formation in ceramic jars caused by the dramatic increase in temperature and humidity gradients on the inner and outer walls due to concentrated rainstorms.

[0055] 2. This invention also constructs a digital twin of vinegar aging that includes a multi-physics coupling model and a reaction kinetics model. Before high-risk weather occurs, future meteorological data can be input to perform risk extrapolation and strategy simulation. The effects of different control strategies can be simulated through a multi-objective optimization decision-making algorithm, and the Pareto optimal solution can be selected as the execution strategy. This transforms traditional experience-based intervention into scientific decision-making based on virtual simulation, significantly improving the success rate and economy of intervention measures.

[0056] 3. This invention also establishes a dynamic correlation formula between the tarpaulin unfolding angle, micro-circulation wind speed, and real-time thermal and moisture stress index, and adopts a PID control algorithm. It can automatically and accurately adjust the actuator according to the deviation and changing trend of the real-time thermal and moisture stress index, forming an adaptive closed-loop control loop. This dynamic and smooth intervention achieves gentle and continuous fine-tuning of the microenvironment, effectively suppressing ion precipitation while minimizing human interference with the natural aging process. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0058] Figure 2 This is a flowchart of the real-time monitoring and risk assessment of thermal and humid stress according to the present invention;

[0059] Figure 3 This is a flowchart of the dynamic deduction and strategy optimization based on digital twins of the present invention;

[0060] Figure 4 This is a flowchart of the dynamic microenvironment resistance control execution and adaptive closed-loop control of the present invention. Detailed Implementation

[0061] Example: Figures 1 to 4 As shown, the present invention relates to a method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, comprising the following steps:

[0062] S100, Real-time Monitoring and Risk Assessment of Thermal and Humid Stress:

[0063] S101. Deploy a multi-parameter monitoring network: Select a representative cluster of aged pottery jars, with each cluster consisting of 50 jars as a monitoring unit. Select 3 representative jars to cover the edge, middle, and low-lying areas of the drying yard, and deploy the following sensors and data transmission and storage devices:

[0064] S1011. Monitoring of the ceramic jar body: Corrosion-resistant miniature temperature and humidity sensors are installed on both the inner wall of the vinegar mash immersed in each monitoring ceramic jar and the outer wall exposed to air, respectively, for synchronous and real-time measurement of the inner wall temperature. Inner wall humidity and outer wall temperature External wall humidity ;

[0065] S1012. Environmental and climate monitoring: Install an air temperature and humidity sensor (model SHT30) at a height of 1.5 meters above the ground in the drying yard, and bury a soil moisture sensor (model EC-5) in the soil at a depth of 10 cm to collect environmental baseline data.

[0066] S1013. Rainfall monitoring: Install tipping bucket rain gauges, model YL-63, in open areas to record rainfall intensity and cumulative amount in real time.

[0067] S1014. Data transmission and storage: All sensor data is transmitted to the local data gateway and synchronized to the cloud server via low-power IoT technologies such as LoRa.

[0068] S102. Calculate the thermal and moisture stress index and conduct a risk assessment:

[0069] The data processing center performs the following core calculations:

[0070] S1021. Calculation of Thermal and Moisture Stress Index: For each monitored ceramic jar, the thermal and moisture stress index is calculated in real time based on the data collected in step S101. This index is a key parameter characterizing the driving force of material migration caused by differences in the internal and external environments of a ceramic jar. The calculation formula is as follows:

[0071] ;

[0072] in, The temperature difference weighting coefficient is the thermal and damp stress index. The humidity difference weighting coefficient is used. and These are the weighting coefficients obtained through correlation analysis between historical disaster data and ion leaching concentration. This refers to the temperature of the inner wall. Inner wall humidity; For the outer wall temperature, External wall humidity;

[0073] Determining the weighting coefficients: Weighting coefficients and Multiple linear regression was used to fit the relationship between historical thermal and wet stress and measured ion precipitation rates. The specific steps are as follows:

[0074] collect A set of historical data, including thermo-hydro-stress data, for the first One sample ( Record its temperature difference and humidity difference ,Right now and and the corresponding calcium ion precipitation rate ;

[0075] Establish a linear regression model: ,in, , For the first The residuals of each sample; For the first The actual observed values ​​of each sample; For the first Model observations for each sample;

[0076] The least squares method is used for parameter estimation. The goal of this method is to find a set of parameters. To minimize the sum of squared residuals, the objective function is:

[0077] ;

[0078] in, Indicates about parameters and Minimize the objective function, For the first Calcium ion precipitation rate of each sample; For the first The absolute value of the temperature difference between the inner and outer walls of the ceramic jars in each sample; For the first The absolute value of the humidity difference between the inner and outer walls of the ceramic jars in each sample.

[0079] S1022, Risk Threshold Determination: The system will... The value is set with two levels of risk thresholds and an early warning threshold. With action threshold The threshold is determined by analyzing historical data on normal aging and raw / unripe flavor events. Values ​​were determined through statistical analysis;

[0080] Warning threshold Determination: Using a logistic regression model, a... , making when When the probability of the astringency intensity exceeding an acceptable level is lower than a preset low-risk probability threshold, the warning threshold is activated. The calculation formula is:

[0081] ;

[0082] in, Astringency as a sensory odor; This represents the upper limit of acceptable astringency intensity. The acceptable low-risk probability threshold;

[0083] Action threshold Method for determining: Similarly, determine a Value, so that when history When the value exceeds this threshold, the probability of the critical ion concentration exceeding the standard is higher than a preset high-risk probability threshold, triggering an action threshold. The expression is:

[0084] ;

[0085] in, This refers to the calcium ion concentration. This is the critical ion concentration threshold; A high-risk probability threshold that requires action;

[0086] when Exceeding the warning threshold When the threshold for action is exceeded, the system generates an initial warning; if the threshold is exceeded... If this occurs, a Level 1 alarm will be triggered and the blocking and control system will be activated.

[0087] S200 and Rainstorm Event Integrated Early Warning:

[0088] S201. The system accesses short-term and nowcast data from the meteorological department in real time to obtain the probability of rainfall within a certain period of time in the future. Expected rainfall To reduce forecast noise, and in accordance with rainstorm warning levels, a moving average method is used to preprocess the rainfall forecast data. The formula for the moving average method is:

[0089] ;

[0090] in, To smooth out the predicted rainfall; For the first Raw hourly forecast rainfall for each hour; This is the size of the sliding window.

[0091] S202. Perform integrated early warning and judgment: The system performs comprehensive judgment based on multi-source information, replacing qualitative analysis with a quantitative confidence formula:

[0092] ;

[0093] in, This represents the confidence level of the early warning; the higher the value, the stronger the confidence level of the early warning. This is the weighting coefficient for the reliability of weather forecasts; This represents the weighting coefficient for the current level of risk. The risk change trend weighting coefficient, This is the current real-time thermal and moisture stress index; Action threshold; for Value change trend , Indicates a rapid rise. Indicates stability;

[0094] S2021. Upon receiving a weather heavy rain warning, the system immediately checks the real-time data of all monitoring points. The value and its changing trend, if the pottery jars in a certain area The value has exceeded the warning threshold. If the number of cases continues to rise, the system determines that the area faces a high risk and generates a Level 1 warning.

[0095] If the pottery jars in a certain area The value did not exceed the warning threshold. Or related to the warning threshold If they are equal, the aging process of the sun-dried vinegar is considered to be normal, and the current monitoring frequency should be maintained.

[0096] S2022, the early warning information clearly identifies the number, location, and current status of high-risk ceramic jars. The value is recorded and notified to management personnel through the control center screen, mobile application push, and audible and visual alarms.

[0097] S300, Dynamic simulation and strategy optimization based on digital twins:

[0098] S301. Constructing a digital twin of the vinegar aging process: A digital twin is a high-fidelity virtual model that uses multiphysics coupling simulation technology to accurately map the actual aging process. Its core components include:

[0099] a. Multiphysics coupling model:

[0100] Heat conduction model: Describes the heat transfer process inside and outside the ceramic jar, using an unsteady-state heat conduction equation:

[0101] ;

[0102] in, The density of the pottery material; The specific heat capacity of the ceramic jar material; The thermal conductivity coefficient of the ceramic jar material; For internal heat source items; For temperature; For time.

[0103] Moisture migration model: describes the moisture diffusion process within the micropores of the jar wall.

[0104] ;

[0105] in, Humidity; The diffusion coefficient of water molecules; This is a water source term.

[0106] b. Reaction kinetic model: Establishing the thermal and wet stress index Quantitative relationship with ion precipitation rate:

[0107] ;

[0108] in, This represents the rate of calcium ion precipitation. This is the stress-dependent reaction rate constant; This represents the temperature-dependent reaction rate constant. It is the activation energy; It is the gas constant; For temperature.

[0109] c. Control strategy library: Pre-stores various resistance control strategies and their parameterized descriptions.

[0110] ;

[0111] in, For the first Various control strategies; Wind speed; The angle at which the tarpaulin is unfolded; Duration of action; Energy consumption indicators.

[0112] S302, Future Risk Simulation: The system receives high-resolution forecast data for the next 72 hours provided by the meteorological department.

[0113] By numerically solving the multiphysics coupling model, prediction The evolution trajectory of the value:

[0114] ;

[0115] in, for Prediction of time value; Air temperature; Relative humidity; Rainfall; For ambient wind speed; Solar radiation intensity; The current moment; It is the integral variable.

[0116] S303, Strategy Simulation and Optimization: When predictions indicate risk, initiate multi-strategy parallel simulation:

[0117] S3031, Strategy Performance Evaluation: For each candidate strategy... Simulate its effect during the prediction period:

[0118] ;

[0119] in, For the first A comprehensive evaluation index for various strategies; For simulation value; , and These are the weighting coefficients; satisfying... ; To predict the termination time; Energy consumption indicators.

[0120] S3032. Multi-objective optimization decision-making: Using the Pareto optimal solution selection strategy.

[0121] ;

[0122] in, This is the optimal strategy; The threshold for the evaluation index must be met; and the following constraints must also be satisfied: , This is the safety threshold.

[0123] S400, Dynamic Microenvironment Restriction Control Execution:

[0124] S401. Activate the barrier control system: Upon receiving a Level 1 warning, the system automatically activates the barrier control equipment corresponding to the high-risk area.

[0125] S402. Implement physical containment measures, the main measures include:

[0126] S4021, Deployable Intelligent Rainproof and Breathable Canopy: This system controls the rapid deployment of a deployable tarpaulin system covering the ceramic jars. The tarpaulin uses a highly breathable and waterproof material, such as ePTFE membrane, which completely blocks direct rainwater flow while allowing water vapor molecules to slowly pass through, preventing the accumulation of hot, humid steam inside the jars and thus avoiding increased internal pressure. The tarpaulin's deployment angle is... Value correlation, dynamic adaptation to risk level, and tarpaulin unfolding angle. and The value correlation formula is:

[0127] ;

[0128] in, The angle at which the tarpaulin unfolds; the larger the angle, the wider the rainproof coverage. Minimum unfolding angle; The scaling factor controls the unfolding angle as... Sensitivity to changes in value; This represents the current thermal and moisture stress index; This is the warning threshold; This is the action threshold.

[0129] S4022. Activate the low-speed micro-circulation air system: Open the annular air supply ducts arranged between the ceramic jar arrays to generate laminar micro-winds, causing the airflow to flow parallel to the outer surface of the ceramic jars. This aims to gently and continuously remove the water film and heat from the outer wall, thereby reducing the temperature and humidity gradient between the inner and outer walls of the ceramic jars, and thus reducing... Value, wind speed and Value deviation is correlated, thereby enabling dynamic adjustment of wind speed. and The value deviation correlation formula is:

[0130] ;

[0131] in, For micro-circulation wind speed, To maintain the minimum wind speed; As a proportionality coefficient, the wind speed is controlled by... Sensitivity to changes in value deviation; This represents the current thermal and moisture stress index; The target control value is usually set below [a certain value]. .

[0132] S403, Adaptive Closed-Loop Control: The resistance control process is dynamically adjusted based on real-time feedback.

[0133] S4031, with the following Value controlled at action threshold The following measures are taken, and the rate of change is minimized as much as possible to control the rate of change.

[0134] S4032, adopts PID control algorithm, based on real-time... The deviation between the target value and the actual value is used to dynamically adjust the wind speed of the micro-circulation air system. The PID control algorithm formula is as follows:

[0135] ;

[0136] in, for Wind speed adjustment over time. for Timing deviation; ; This is a proportionality coefficient that determines the strength of the response to the current deviation; These are integral coefficients to eliminate steady-state errors; These are differential coefficients, used to predict error trends and improve system stability.

[0137] S4033, forming a monitoring system The value is used to calculate the control quantity, and then the actuator is adjusted to affect the microenvironment. The real-time feedback control closed loop, which detects changes in values, works in synergy with the rainproof tarpaulin and the circulation of a gentle breeze to smooth out temperature and humidity gradients, creating a synergistic resistance effect that effectively suppresses ion precipitation.

[0138] S500, Closed-loop feedback and effect verification based on proxy indicators:

[0139] S501. During the blocking operation, the system continuously monitors two key proxy indicators through an online analyzer to reflect the ion precipitation and astringent substance generation status in real time.

[0140] S5011. Using an online pH meter and a conductivity meter deployed at the sampling port of the ceramic jar, the pH value and conductivity of the vinegar mash inside the jar are monitored in real time. Ion dissolution directly affects the rate of change of these parameters. A conductivity change rate is established. With calcium ion precipitation rate Association model:

[0141] ;

[0142] in, This represents the rate of calcium ion precipitation. The rate of change of electrical conductivity; This is the proportionality coefficient; This is the intercept term.

[0143] S5012, Online Near-Infrared Spectrometer: Continuously scans the absorbance changes of vinegar mash at specific characteristic wavenumbers, such as those associated with phenolic substances.

[0144] S502, Execution Feedback and Verification: Dynamically adjust the control strategy based on agent indicators and conduct rapid verification after the rainstorm ends.

[0145] S5021. During the resistance control process, if the rate of change of conductivity... If the proxy indicators do not stabilize quickly, the control system will automatically increase the resistance strength.

[0146] S5022. Within 24 hours of the rainstorm ending, samples should be taken for rapid laboratory verification.

[0147] This includes: using Raman spectroscopy for semi-quantitative detection of the intensity of tannin characteristic peaks;

[0148] Potassium ion concentration was quantitatively determined using ion chromatography.

[0149] An electronic tongue was used to detect the astringent taste response value.

[0150] S5023, Verification Standard: If the key flavor defect indicators (including tannin content, potassium ion concentration and sensory astringency value) of the vinegar sample after the control treatment are significantly lower than the average value of historical risk event samples without control treatment, the emergency control can be considered to have met the standard.

[0151] S600, Post-event Evaluation and Model Optimization:

[0152] S601, In-depth data archiving and causal analysis, stores the data of each complete event chain, meteorological data, value curves, control parameters, verification results, and precise laboratory analysis data, such as the porosity of ceramic jars measured by mercury porosimetry and the content of various metal ions measured by ICP-MS, into the case library.

[0153] S602. Run the reverse tracing model by inputting the case data into a trained analytical model, such as a neural network model incorporating AHP weights. Model inputs include: tannin content, potassium / calcium / magnesium ion content, ceramic jar porosity, and 4-ethylguaiacol content. The model outputs the quantitative contribution percentage of each factor to the risk of astringency.

[0154] S603. Iteratively optimize system parameters and periodically retrain the model using accumulated case data:

[0155] S6031. The gradient descent method is used to optimize the weighting coefficients in the formula for the thermal and damp stress index. and Define the loss function:

[0156] ;

[0157] in, The loss function measures the model's prediction error. This represents the number of samples in the current case library. For the first The actual calcium ion precipitation rate of each sample; For the first Temperature difference of each sample; For the first Humidity difference between samples;

[0158] Update rules:

[0159] ;

[0160] ;

[0161] in, , These are the updated weighting coefficients; , The weighting coefficients are those obtained before the update, based on the correlation analysis between historical disaster data and ion dissolution concentration. The learning rate controls the step size for each parameter update; , This is the partial derivative of the loss function with respect to the weight coefficients.

[0162] S6032. Optimize the two-level early warning thresholds based on more accurate risk prediction results. and The goal is to continuously improve the system's accuracy in predicting the risk of astringency after heavy rain, thereby enabling the system to evolve itself.

[0163] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, characterized in that, Includes the following steps: S100, Real-time monitoring and risk assessment of thermal and humidity stress: Temperature and humidity sensors are installed on both the inner and outer walls of the aged ceramic jars to simultaneously measure the temperature and humidity of the inner and outer walls in real time and calculate the thermal and humidity stress index. Based on historical data, early warning thresholds and action thresholds are set to conduct risk assessment. thermal and damp stress index The calculation formula is: ； in, The temperature difference weighting coefficient is the thermal and damp stress index. The humidity difference weighting coefficient; This refers to the temperature of the inner wall. Inner wall humidity; For the outer wall temperature, External wall humidity; Determining the weighting coefficients: Weighting coefficients and The data were fitted using multiple linear regression based on the historical data of thermal and wet stress and the measured ion precipitation rate. S200, Rainstorm Event Fusion Early Warning: Access meteorological forecast data, use the moving average method to preprocess rainfall forecast data and perform multi-source information fusion analysis, and generate a corresponding level of early warning signal when the confidence level exceeds the set threshold; S300, Dynamic simulation and strategy optimization based on digital twin: Construct a digital twin containing a multi-physics coupling model and a reaction dynamics model, input meteorological data to predict the evolution trajectory of the thermal and wet stress index, and select the optimal control strategy through multi-strategy parallel simulation and Pareto optimization; S400, Dynamic Microenvironment Resistance Control Execution: Responds to the early warning signal and starts the resistance control system, executes physical resistance control measures and dynamically adjusts based on the real-time thermal and moisture stress index; The implementation of physical control measures mainly includes: The intelligent rainproof and breathable awning unfolds rapidly: a controllable awning system covering the pottery jars rapidly deploys. This awning is made of highly breathable and waterproof material, and its unfolding angle... and The value correlation formula is: ； in, The angle at which the tarpaulin unfolds; the larger the angle, the wider the rainproof coverage. Minimum unfolding angle; The scaling factor controls the unfolding angle as... Sensitivity to changes in value; This represents the current thermal and moisture stress index; This is the warning threshold; Action threshold; Activate the low-speed micro-circulation air system: Open the annular air supply ducts arranged between the ceramic jars to generate laminar micro-wind, causing the airflow to flow parallel to the outer surface of the ceramic jars, with a wind speed of... and The value deviation correlation formula is: ； in, For micro-circulation wind speed, To maintain the minimum wind speed; The proportionality coefficient controls the wind speed as follows Sensitivity to changes in value deviation; This represents the current thermal and moisture stress index; The target control value is usually set below [a certain value]. ; S500, closed-loop feedback based on proxy indicators: real-time acquisition of pH value, conductivity and spectral data proxy indicators, dynamic adjustment of resistance control parameters and verification of resistance control results; S600, Post-event assessment and model optimization: Archive event data and analyze the contributing factors of astringent taste risk, and use the gradient descent method to iteratively optimize the weight coefficient of the thermal and humid stress index and the early warning threshold.

2. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 1, is characterized in that... Step S100 further includes the following steps: S101. Deploy a multi-parameter monitoring network: Install temperature and humidity sensors on the inner and outer walls of representative aged ceramic jars to simultaneously measure the inner wall temperature in real time. Inner wall humidity and outer wall temperature External wall humidity ; S102. Calculate the thermo-moisture stress index and conduct a risk assessment: Based on the data collected in step S101, the data processing center calculates the thermo-moisture stress index in real time. Set early warning thresholds based on historical data. With action threshold Conduct a risk assessment; Warning threshold The calculation formula is: ； in, Astringency as a sensory odor; This represents the upper limit of acceptable astringency intensity. The acceptable low-risk probability threshold; Action threshold The calculation formula is: ； in, This refers to the calcium ion concentration. This is the critical ion concentration threshold; This is a high-risk probability threshold that requires action.

3. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 1, is characterized in that... Step S200 further includes the following steps: S201. Data Acquisition and Preprocessing: Accessing meteorological forecast data, rainfall probability... Expected rainfall Rainstorm warnings were issued, and the moving average method was used to preprocess the rainfall forecast data; S202. Execute integrated early warning and judgment: Use a quantitative confidence formula to comprehensively judge multi-source information, and when heavy rain is predicted or occurs and the thermal and humidity stress index... Exceeding the warning threshold At that time, a Level 1 warning signal is generated.

4. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 3, is characterized in that... In step S201, the formula for the moving average method is: ； in, To smooth out the predicted rainfall; For the first Raw hourly forecast rainfall for each hour; To adjust the sliding window size; In step S202, the quantification confidence formula is: ; in, For the confidence level of the early warning; This is the weighting coefficient for the reliability of weather forecasts; This represents the weighting coefficient for the current level of risk. The weighting coefficient for the trend of risk change; This is the current real-time thermal and moisture stress index; Action threshold; for Value change trend.

5. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 1, is characterized in that... Step S300 further includes the following steps: S301. Construct a digital twin of vinegar aging: The digital twin includes a multiphysics coupling model, a reaction kinetics model, and a control strategy library; S302. Future Risk Simulation: Input future meteorological data and predict the thermal and moisture stress index using the digital twin. evolutionary trajectory ; S303, Strategy Simulation and Optimal Selection: When a risk is predicted, multi-strategy parallel simulation is initiated, and the optimal control strategy is determined based on the Pareto optimal solution. .

6. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 5, is characterized in that... In step S303, the formula for calculating the Pareto optimal solution is: ； in, This is the optimal strategy; For the first Various control strategies; For the first A comprehensive evaluation index for this strategy The threshold for the evaluation index.

7. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 3, is characterized in that... Step S400 further includes the following steps: S401. Start the barrier control system: In response to the first-level warning signal in step S202, automatically activate the barrier control equipment corresponding to the high-risk area; S402. Implement physical barrier measures, including deploying the intelligent rainproof and breathable awning and activating the low-wind-speed micro-circulation air system; wherein, the awning deployment angle... and micro-circulation wind speed According to the thermal and damp stress index Dynamic adjustment; S403, Adaptive Closed-Loop Control: Employs a PID control algorithm based on real-time data... The deviation between the current value and the target value is dynamically adjusted to control parameters.

8. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 7, is characterized in that... In step S403, the formula for the PID control algorithm is: ； in, for Wind speed adjustment over time. for Timing deviation; This is the proportionality coefficient. The integral coefficient; is the differential coefficient.

9. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 1, is characterized in that... Step S500 further includes the following steps: S501, Real-time Monitoring of Agent Indicators: Continuously monitor the conductivity and pH value of the vinegar mash using an online analyzer; and establish a conductivity change rate... With calcium ion precipitation rate The association model; S502. Feedback Verification: Dynamically adjust the resistance intensity based on the proxy indicators and conduct rapid laboratory verification after the rainstorm ends.

10. The method for analyzing the dynamic changes in flavor during the aging period of sun-dried vinegar based on flavor omics, as described in claim 1, is characterized in that... Step S600 further includes the following steps: S601. In-depth data archiving and causal analysis: Store complete event chain data in the case library and run a reverse tracing model to analyze the causes of risks; S602. Input the case data into the neural network model and output the quantitative contribution percentage of each factor to the risk of astringency. S603. Iterative optimization of system parameters: Using accumulated case data, the weight coefficients are periodically optimized using the gradient descent method. and And iteratively update the warning threshold. With action threshold .

Images