Automatic Liquid Nitrogen Flow Rate Control System and Its Application in Abalone Preservation

Abstract

This invention relates to the field of automatic control technology for the quick-freezing process of aquatic products, specifically an automatic liquid nitrogen flow rate adjustment system and its application in abalone preservation. The system includes a data acquisition module for real-time acquisition of the physical parameters of the target aquatic product; a thermal property calculation module for determining the instantaneous thermal conductivity of the aquatic product; a state prediction module for predicting the system state at the next moment, including the core temperature and surface temperature at the next moment; a risk prediction module for determining the predicted heat stress risk index at the next moment; and a control decision module for combining the deviation between the core temperature predicted by the state prediction module and the preset target temperature, and the predicted heat stress risk index determined by the risk prediction module, to solve for the minimum value of a preset cost function to determine the optimal liquid nitrogen injection intensity. This invention eliminates the reliance on fixed cooling procedures, ensuring the stability and high quality of the quick-freezing preservation effect.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for the quick-freezing process of aquatic products, specifically to an automatic liquid nitrogen flow rate adjustment system and its application in abalone preservation. Background Technology

[0002] In the field of low-temperature preservation of aquatic products, especially in precision temperature control applications such as liquid nitrogen quick-freezing, the accuracy of the control strategy directly determines the preservation quality and economic benefits. Existing quick-freezing control technologies generally adopt open-loop or semi-closed-loop control schemes based on solidification thermophysical parameters. These schemes set a universal cooling program based on the empirical model of the sample, which cannot be adjusted in real time for individual differences of the processed objects and dynamic changes during the freezing process.

[0003] This technical approach faces significant bottlenecks. The inherent individual heterogeneity of aquatic products, such as abalone, with its variations in size, water content, and tissue composition, means that fixed thermophysical parameters cannot accurately reflect the true physical characteristics of any single sample, leading to control inaccuracies. During phase transitions, key parameters such as thermal conductivity of the sample undergo drastic nonlinear changes, which static control models cannot capture and adapt to. This imprecise and non-dynamic control method makes it difficult to achieve fine-tuning of the cooling rate, easily causing excessive thermal stress and ice crystal recrystallization, resulting in irreversible damage to cell structure and affecting preservation quality and subsequent application value. It also leads to unnecessary consumption of liquid nitrogen, increasing operating costs. Therefore, how to overcome the limitations of static models and achieve precise, dynamic, and personalized closed-loop control that can adapt to individual heterogeneity and process dynamics to synergistically improve preservation quality and energy efficiency is a key technical problem that urgently needs to be solved. Summary of the Invention

[0004] The purpose of this invention is to provide an automatic liquid nitrogen flow rate regulation system and its application in abalone preservation. This system can adapt to the individual heterogeneity of aquatic products and the dynamic nonlinear characteristics of phase change processes, achieving precise, dynamic, and personalized control of the cooling process, thereby synergistically improving preservation quality and energy efficiency. Specifically, the technical solution of this invention is as follows:

[0005] An automatic liquid nitrogen flow rate control system includes:

[0006] The data acquisition module is used to acquire the physical parameters of the target aquatic products in real time. The physical parameters include: core temperature, surface temperature, tissue impedance modulus and sound wave propagation speed.

[0007] The thermal property calculation module is used to determine the instantaneous thermal conductivity of aquatic products based on the tissue impedance modulus and sound wave propagation velocity obtained by the data acquisition module.

[0008] The state prediction module is used to combine the core temperature and surface temperature obtained by the data acquisition module and the instantaneous thermal conductivity determined by the thermal property calculation module, and use the extended Kalman filter model to predict the system state at the next moment, which includes the core temperature and surface temperature at the next moment.

[0009] The risk prediction module is used to determine the predicted heat stress risk index for the next moment based on the system state at the next moment obtained by the state prediction module.

[0010] The control decision module is used to combine the deviation between the core temperature predicted by the state prediction module and the preset target temperature at the next moment, and the predicted heat stress risk index determined by the risk prediction module, to solve for the minimum value of the preset cost function to determine the optimal liquid nitrogen injection intensity.

[0011] Optionally, the thermal property calculation module is specifically used for:

[0012] The instantaneous ice crystal integral is determined based on the tissue impedance modulus, the preset unfrozen state impedance modulus, and the fully frozen state impedance modulus.

[0013] Instantaneous thermal conductivity is calculated based on a parallel thermal resistance model by combining the instantaneous ice crystal integral number, the sound wave propagation velocity, and the preset acoustic-thermal coupling coefficient.

[0014] Optionally, the state prediction module is specifically used for:

[0015] Construct the current system state vector, which includes the core temperature and the surface temperature;

[0016] The current system state vector is predicted by using a nonlinear state transition equation incorporating instantaneous thermal conductivity.

[0017] The prediction results are corrected by using an observation equation that incorporates real-time sensor measurements to output the system state at the next moment.

[0018] Optionally, the risk prediction module is specifically used for:

[0019] The dynamic thermal gradient is calculated and predicted based on the difference between the core temperature and the surface temperature at the next time step.

[0020] Based on the predicted dynamic thermal gradient, the predicted proportional risk term is determined.

[0021] Based on the predicted rate of change of the dynamic thermal gradient, the predicted differential risk term is determined.

[0022] The predicted proportional risk term and the predicted differential risk term are weighted and summed to obtain the predicted heat stress risk index.

[0023] Optionally, the predicted proportional risk term is the ratio of the predicted dynamic thermal gradient to the preset peak dynamic thermal gradient threshold.

[0024] Optionally, the predicted differential risk term is the ratio of the predicted rate of change of the dynamic thermal gradient to the preset peak dynamic thermal gradient threshold, multiplied by the preset system thermal response time constant.

[0025] Optionally, the cost function configured in the control decision module includes an operating cost term and a risk penalty term; the operating cost term is proportional to the liquid nitrogen injection intensity; the risk penalty term is determined in the following way:

[0026] The predicted heat stress risk index is compared with the preset risk reference threshold.

[0027] If the predicted heat stress risk index is greater than the risk reference threshold, the risk penalty item is determined based on the difference between the two.

[0028] If the predicted heat stress risk index is not greater than the risk reference threshold, the risk penalty term is set to zero. Optionally, the control decision module is specifically used for:

[0029] Liquid nitrogen injection intensity was used as the optimization variable.

[0030] The liquid nitrogen injection intensity that minimizes the cost function is determined by a numerical optimization algorithm, and this intensity is output as the optimal liquid nitrogen injection intensity.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. This invention dynamically calculates the instantaneous thermal conductivity of aquatic products during the freezing process by real-time monitoring of physical quantities such as tissue impedance and sound velocity. This method overcomes the shortcomings of traditional technologies that rely on solidification and universal thermophysical parameters, and can accurately reflect the real physical characteristics of different individuals and different freezing stages. It provides a reliable dynamic model basis for achieving personalized and precise control, and significantly improves the accuracy of control.

[0033] 2. Based on the extended Kalman filter model, this invention can accurately predict the internal temperature distribution of the sample at the next moment according to the current state. The system innovatively constructs a risk assessment model that comprehensively considers the magnitude and rate of change of the thermal gradient, realizing the forward-looking quantification of the risk of future heat stress damage. This proactive risk avoidance control method effectively avoids cell damage caused by excessive temperature difference and significantly improves the preservation quality of the sample.

[0034] 3. This invention unifies preservation quality and operating costs within an optimal control framework. By constructing a cost function that includes risk penalty terms and energy consumption cost terms, the control system can automatically find and execute the most energy-efficient liquid nitrogen injection strategy while ensuring sample safety and not triggering risk thresholds. This refined flow regulation avoids liquid nitrogen waste caused by excessive cooling and significantly reduces operating costs.

[0035] 4. This invention constructs a complete closed-loop control system from real-time perception, dynamic modeling, future prediction to optimization decision-making. This system can automatically adapt to the individual differences of aquatic products and the nonlinear dynamic changes in thermophysical properties during freezing, providing each sample with a personalized optimal cooling path throughout the process. This high degree of adaptability and automation eliminates the dependence on fixed cooling programs, ensuring the stability and high quality of quick-freezing preservation. Attached Figure Description

[0036] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0037] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0039] Example 1:

[0040] Please see Figure 1 An automatic liquid nitrogen flow rate control system includes:

[0041] The data acquisition module is used to acquire the physical parameters of the target aquatic products in real time. The physical parameters include: core temperature, surface temperature, tissue impedance modulus and sound wave propagation speed.

[0042] The thermal property calculation module is used to determine the instantaneous thermal conductivity of aquatic products based on the tissue impedance modulus and sound wave propagation velocity obtained by the data acquisition module.

[0043] The state prediction module is used to combine the core temperature and surface temperature obtained by the data acquisition module and the instantaneous thermal conductivity determined by the thermal property calculation module, and use the extended Kalman filter model to predict the system state at the next moment, which includes the core temperature and surface temperature at the next moment.

[0044] The risk prediction module is used to determine the predicted heat stress risk index for the next moment based on the system state at the next moment obtained by the state prediction module.

[0045] The control decision module is used to combine the deviation between the core temperature predicted by the state prediction module and the preset target temperature at the next moment, and the predicted heat stress risk index determined by the risk prediction module, to solve for the minimum value of the preset cost function to determine the optimal liquid nitrogen injection intensity.

[0046] This invention provides an automatic liquid nitrogen flow rate regulation system to address the technical problem in existing technologies where the thermophysical parameters are fixed due to the individual heterogeneity of aquatic products, making it impossible to achieve precise, dynamic, and personalized control of the cooling process. This system is applied to the liquid nitrogen quick-freezing and preservation process of abalone. Through a combination of multimodal sensing, dynamic modeling, and predictive control, it achieves optimized control of liquid nitrogen consumption while ensuring preservation quality. The system constitutes a closed-loop technology, including a data acquisition module, a thermophysical property calculation module, a state prediction module, a risk prediction module, and a control decision module.

[0047] The data acquisition module provides real-time, multi-dimensional physical state input for the analysis, prediction, and control of the entire system. This module utilizes high-frequency acoustic transducers deployed on the inner wall of the preservation chamber, thermocouples implanted on the surface of the abalone samples, thermocouples implanted in the sample core, and micro-impedance electrodes implanted in the tissue. This module can synchronously and in real-time acquire four physical parameters of the target aquatic product: core temperature... Surface temperature Tissue impedance modulus and the speed of sound wave propagation These four parameters form the basis for a comprehensive characterization of the internal thermophysical state of the sample.

[0048] The thermophysical property calculation module overcomes the uncertainty of thermodynamic parameters caused by the heterogeneity of biological individuals, dynamically calculating instantaneous thermophysical parameters reflecting the current state of the sample; this module receives tissue impedance modulus values ​​provided by the data acquisition module. With the speed of sound wave propagation The instantaneous thermal conductivity of the sample is calculated in real time using a built-in physical model. Instantaneous thermal conductivity The effective thermal conductivity of an organism at a specific time t is a core parameter that provides accurate and dynamically changing information for predicting subsequent heat transfer processes. It is based on real-time sensor data. and Calculated online;

[0049] The state prediction module is used to accurately predict the future evolution trend of the system based on the current state, providing a decision basis for implementing feedforward optimization control. The core of this module is an extended Kalman filter model; it integrates the core temperature data acquired by the data acquisition module. With surface temperature The real-time measured values, and the dynamically changing instantaneous thermal conductivity provided by the thermal property calculation module. This model can accurately predict the core temperature at the next moment. With the surface temperature at the next moment The system state at the next moment ;

[0050] The risk prediction module quantifies complex cell damage risks into a metric that can be used for optimization calculations; this module is based on the system state at the next time step obtained by the state prediction module. By using a risk model that comprehensively considers the magnitude and rate of change of the thermal gradient, the predicted heat stress risk index for the next moment is determined. Predicting the heat stress risk index It refers to the quantitative assessment of the risk of irreversible cell damage to a sample due to excessive internal and external temperature differences or rapid changes within a future control period. It directly incorporates process safety into the scope of control decision-making and is based on the calculation results of predicted temperature conditions.

[0051] The control decision module, as the final execution decision-making link of the entire closed-loop system, dynamically optimizes to determine the best control output. This module weighs two core objectives within a unified cost function: reducing the core temperature predicted by the state prediction module for the next time step. With preset target temperature The deviation between these parameters ensures cooling efficiency; the predicted thermal stress risk index is determined by the risk prediction module. As a constraint, the preservation quality is ensured; ultimately, the optimal liquid nitrogen injection intensity is determined by solving for the minimum value of a preset cost function. ;

[0052] The system may further include a fault diagnosis module, which can monitor the rationality of the data of each sensor in real time through preset thresholds or historical data models. When an abnormal sensor data or communication interruption is detected, the system can automatically switch to a degraded control mode based on the remaining effective sensors, or start a preset safety shutdown procedure to ensure the safety of aquatic products and equipment, thereby improving the overall robustness of the system.

[0053] This embodiment constructs a closed loop from real-time perception, dynamic modeling, future prediction to optimized control through the collaborative work of the aforementioned modules. It no longer relies on static, universal thermophysical parameters, but instead performs personalized parameter calculations and state predictions for each aquatic product at every moment of the cooling process. This design enables the system to accurately anticipate and avoid the risk of cell damage caused by excessive thermal gradients, while precisely controlling the liquid nitrogen flow rate according to actual needs. This resolves the contradiction between preservation quality and energy efficiency, achieving high-quality, high-efficiency, and personalized automatic control of the quick-freezing process. It should be noted that this embodiment uses a core-surface two-point model to characterize the overall thermal state of the sample and aggregates spatially varying physical parameters into a single instantaneous effective value. This is an effective simplification of complex three-dimensional physical processes while ensuring the feasibility of real-time control calculations, suitable for application scenarios where the overall cooling rate and macroscopic temperature gradient are the primary concerns.

[0054] Example 2:

[0055] The thermal property calculation module is specifically used for:

[0056] The instantaneous ice crystal integral is determined based on the tissue impedance modulus, the preset unfrozen state impedance modulus, and the fully frozen state impedance modulus.

[0057] Instantaneous thermal conductivity is calculated based on a parallel thermal resistance model by combining the instantaneous ice crystal integral number, the sound wave propagation velocity, and the preset acoustic-thermal coupling coefficient.

[0058] To provide a detailed explanation of the thermal property calculation module described in Example 1, the implementation of this module aims to improve instantaneous thermal conductivity. The accuracy of the calculation;

[0059] This module uses a semi-empirical model based on the effective medium theory to calculate the instantaneous ice crystal integral number. Instantaneous ice crystal integral Used to characterize the volume percentage of liquid water that transforms into solid ice within the tissue at time t, it serves as a key intermediate variable connecting measurable tissue impedance and indirect thermal conductivity. Its calculation method is as follows:

[0060]

[0061] in: The value of tissue impedance modulus measured at the current time t is in ohms and is obtained from real-time input by the miniature impedance electrode of the data acquisition module.

[0062] The impedance modulus of the tissue in a completely unfrozen state, in ohms, serves as the initial state reference and is obtained from measurements taken before cooling begins.

[0063] The impedance modulus of the sample in a fully frozen state, in ohms, serves as a terminal state reference and is obtained by measurement after the sample is fully frozen or extracted from a reference database.

[0064] The geometric correction factor is a dimensionless parameter determined through offline calibration experiments. To clarify the calibration process, a calibration dataset is defined. ,in The impedance modulus of a representative sample k is measured experimentally under a certain frozen condition. The integral number of ice crystals measured by differential scanning calorimetry under the same conditions; the least squares method is used to... By performing a fitting, the parameter γ can be calibrated; this calibration process ensures that the variables used for calibration are accurate. and Variables used as input during model runtime and the variables as output Clear distinctions in symbols ensure the feasibility of the technical solution;

[0065] To ensure the instantaneous ice crystal integral number The physical meaning of , its value range should be between 0 and 1, and its calculation result should be clamped, that is, if the calculated value is less than 0, take 0, if it is greater than 1, take 1;

[0066] Obtaining the instantaneous ice crystal integral. Subsequently, to more comprehensively reflect the impact of tissue solidification degree on heat conduction, this embodiment further integrates acoustic information and calculates the final instantaneous thermal conductivity based on a modified parallel thermal resistance model. The underlying logic is that freezing not only alters the ice-water ratio but also significantly increases hardness. Changes in sound velocity can characterize changes in hardness, thus providing a second dimension of physical information for calculating thermal conductivity. The calculation method is as follows:

[0067]

[0068] in: The effective thermal conductivity of the organization at time t, in units of W / (m⋅K), is calculated from the current step and output to the state prediction module;

[0069] and These are the reference thermal conductivity values ​​for water and ice, respectively, in W / (m⋅K). They are physical constants and are obtained from standard physical property data handbooks.

[0070] This is the instantaneous ice crystal integral, which is dimensionless and is derived from the calculation in the previous step.

[0071] The sound velocity is measured in real time, in m / s, and is obtained from the real-time input of the high-frequency acoustic transducer of the data acquisition module.

[0072] The initial sound velocity of the unfrozen tissue is expressed in m / s and is obtained from measurements taken before cooling begins.

[0073] The acoustic-thermal coupling coefficient, measured in W / (m⋅K) / (m / s), represents the weight of the contribution of sound velocity changes to thermal conductivity. This parameter is obtained through multiphysics simulations or joint measurement experiments, calibrated using the least squares method. To clarify the calibration process, a calibration dataset is defined. ,in , and These are the thermal conductivity, sound velocity, and ice crystal fraction experimentally measured for a representative sample k under specific conditions; the least squares method is used to... By performing a fitting, the parameters can be calibrated. This calibration process also ensures the independence of the calibration variables from the variables used during model runtime in terms of sign.

[0074] Through this implementation method, the thermal property calculation module creatively integrates sensor data from two different modes: electrical and acoustic. The calculation precisely correlates impedance, an easily measurable macroscopic quantity, with the microscopic state of ice crystal formation; further, by introducing the speed of sound... As a characterization of tissue stiffness, the classic hybrid law model was modified; this multimodal data fusion method improves instantaneous thermal conductivity. The robustness and accuracy of the solution provide a more reliable physical model basis for subsequent state prediction and control decisions.

[0075] Example 3:

[0076] The state prediction module is specifically used for:

[0077] Construct the current system state vector, which includes the core temperature and the surface temperature;

[0078] The current system state vector is predicted by using a nonlinear state transition equation incorporating instantaneous thermal conductivity.

[0079] The prediction results are corrected by using an observation equation that incorporates real-time sensor measurements to output the system state at the next moment.

[0080] To illustrate the state prediction module described in Example 1, this module employs an extended Kalman filter model to achieve accurate prediction of the system state.

[0081] This module constructs the current system state vector, which includes the core temperature and surface temperature; system state vector This is a mathematical expression that fully describes the thermodynamic state of the system at time t. It serves as the starting point for the iterative calculations of the Kalman filter algorithm and originates from real-time temperature measurements integrated with the data acquisition module. This vector is defined as... ,in and These are the core and surface temperatures at the current moment;

[0082] This module employs a nonlinear state transition equation incorporating instantaneous thermal conductivity to predict the current system state vector; the technical solution of this invention utilizes the instantaneous thermal conductivity calculated by the thermophysical property solution module. It is dynamically incorporated as a time-varying parameter into the state transition function discretized based on Fourier's law of heat conduction. In the middle; due to It is essentially temperature, which is its state. This is a function of , which makes the state transition equation nonlinear, and its mathematical expression is:

[0083]

[0084] in It is the predicted state at the next moment, which is derived from the calculation results of this equation; It is the current state, which is derived from the filtered output of the previous moment; It is the control input variable of the current control cycle. When the control decision module performs optimization, it is substituted into the equation as a candidate value to predict the future state of the system under different control strategies. Its source is the output of the control decision module. It is the instantaneous thermal conductivity provided by the thermal property calculation module; It is the process noise vector, and its covariance matrix is... These are adjustable parameters that reflect the degree of confidence in the physical model and are determined through system identification or empirical tuning.

[0085] This module uses an observation equation incorporating real-time sensor measurements to correct the prediction results and output the system state at the next moment. To correct the deviation between the model prediction and the actual physical process, this embodiment uses an observation equation to establish the relationship between the internal state and the sensor measurements, the mathematical expression of which is:

[0086]

[0087] in It is the actual sensor measurement vector at time t. Its source is the data acquisition module; This is the observation function, here a linear identity matrix, representing the direct measurement of core and surface temperatures; It is a measurement noise vector, and its covariance matrix is... The source is the technical manual of the sensor device; the extended Kalman filter algorithm calculates the Kalman gain between the predicted and measured values, and uses this gain to weight and correct the initial prediction results, finally outputting an optimal estimate of the system state at the next time step. ;

[0088] By dynamically calculating the instantaneous thermal conductivity By incorporating the state transition equation of the extended Kalman filter as a core parameter, the state prediction module in this embodiment can reflect the drastic changes in internal heat transfer characteristics caused by sample freezing in real time and adaptively, thereby enabling the prediction of the core temperature at the next moment. The prediction accuracy has been significantly improved; this high-precision prediction capability is a key prerequisite for realizing feedforward, preventive risk control and refined flow regulation.

[0089] Example 4:

[0090] The risk prediction module is specifically used for:

[0091] The dynamic thermal gradient is calculated and predicted based on the difference between the core temperature and the surface temperature at the next time step.

[0092] Based on the predicted dynamic thermal gradient, the predicted proportional risk term is determined.

[0093] Based on the predicted rate of change of the dynamic thermal gradient, the predicted differential risk term is determined.

[0094] The predicted proportional risk term and the predicted differential risk term are weighted and summed to obtain the predicted heat stress risk index.

[0095] The predicted proportional risk term is the ratio of the predicted dynamic thermal gradient to the preset peak dynamic thermal gradient threshold.

[0096] The predicted differential risk term is the ratio of the predicted rate of change of the dynamic thermal gradient to the preset peak dynamic thermal gradient threshold, multiplied by the preset system thermal response time constant. .

[0097] To elaborate on the risk prediction module described in Example 1, this module constructs a dynamic assessment system capable of predicting and quantifying future heat stress risks;

[0098] This module calculates and predicts the dynamic thermal gradient based on the difference between the core temperature and the surface temperature at the next time step; predicting the dynamic thermal gradient... This refers to the prediction of the temperature difference between the inside and outside of the sample at the next control moment. It serves as the basis for assessing the magnitude of future thermal stress and is derived from the system state at the next moment output by the state prediction module. Perform the calculation; the specific calculation is as follows: ;

[0099] This module constructs a comprehensive predictive heat stress risk index. The calculation model; an effective risk assessment should not only focus on the magnitude of the current deviation, but also predict the rate of change of the deviation, so as to achieve early warning of risk; therefore, this embodiment performs a weighted summation of the predicted proportional risk term and the predicted differential risk term to obtain the predicted heat stress risk index, which is calculated as follows:

[0100]

[0101] The first term on the right side of the formula is the predicted proportional risk term, and the second term is the predicted differential risk term.

[0102] The predicted proportional risk term is the ratio of the predicted dynamic thermal gradient to the preset peak dynamic thermal gradient threshold, i.e. Peak dynamic thermal gradient threshold This refers to a pre-set safe temperature difference upper limit based on the physiological characteristics of biological tissues. It serves as a clear and quantifiable reference benchmark for risk assessment. It is derived from offline experiments to find the gradient inflection point where cell damage increases sharply. This ratio directly reflects how close the predicted temperature difference is to the safe upper limit.

[0103] The predicted differential risk term is the ratio of the predicted rate of change of the dynamic thermal gradient to the preset peak dynamic thermal gradient threshold, multiplied by the preset system thermal response time constant, i.e. This differential term is used to warn of the risk of abrupt changes in the thermal gradient; the system thermal response time constant. This refers to the inherent property of how fast a system responds to thermal disturbances. Its function is to unify the physical dimensions on both sides of the formula into dimensionless quantities. It is obtained by conducting step response tests on the system.

[0104] in: and , a dimensionless weighting coefficient, is used to adjust the relative importance of amplitude and trend in risk assessment. Its value is determined based on the emphasis on preservation quality and efficiency, and is selected through simulation optimization, while satisfying , , and , . ;

[0105] In this way, the risk prediction module constructs a forward-looking, two-dimensional risk assessment model; it not only assesses the magnitude of the thermal gradient at the next moment, but also creatively assesses the changing trend of the thermal gradient; this composite risk assessment enables the system to anticipate and quantify potential risks caused by control behavior in advance, providing a foundation for seamlessly integrating process safety into subsequent optimization control decisions.

[0106] Example 5:

[0107] The cost function configured in the control decision module includes an operating cost item and a risk penalty item; the operating cost item is proportional to the liquid nitrogen injection intensity; the risk penalty item is determined in the following way:

[0108] The predicted heat stress risk index is compared with the preset risk reference threshold.

[0109] If the predicted heat stress risk index is greater than the risk reference threshold, the risk penalty item is determined based on the difference between the two.

[0110] If the predicted heat stress risk index is not greater than the risk reference threshold, then the risk penalty item will be set to zero.

[0111] The decision-making module is specifically used for:

[0112] Liquid nitrogen injection intensity was used as the optimization variable.

[0113] The liquid nitrogen injection intensity that minimizes the cost function is determined by a numerical optimization algorithm, and this intensity is output as the optimal liquid nitrogen injection intensity.

[0114] To elaborate on the control decision module, this module uses optimal control theory to maximize energy efficiency while ensuring safety. Its core is to construct and solve a cost function.

[0115] The cost function configured in this control decision module Includes operating cost items and risk penalty items; cost function This refers to controlling the behavior, specifically the intensity of liquid nitrogen injection. The merits and demerits are quantified into a single numerical mathematical function, which serves as a clear optimization objective for finding the optimal control strategy. It is constructed based on the penalty function method of optimal control theory; its specific form is:

[0116]

[0117] The physical dimension of this cost function is unified as cost per unit time;

[0118] Operating cost item is This item is related to the liquid nitrogen injection intensity. Proportional, of which The unit is watt, and its source is the optimization variable of this module; It is the cost coefficient per unit of energy consumption, expressed in Cost / J, and is a preset economic parameter. The purpose of this design is that, under all constraints, the system will tend to select the control scheme with the lowest energy consumption.

[0119] The risk penalty is determined by the following method: the predicted heat stress risk index. Compared with the preset risk reference threshold Comparison; Risk reference threshold This is a preset safety boundary, typically set to 1.0 based on technical requirements or experimental statistics, representing that the risk index has reached an acceptable upper limit; if the predicted heat stress risk index exceeds the risk reference threshold, then... The risk penalty term is then determined based on the difference between the two, and this term is... ;in This is a high-weighted penalty coefficient, measured in Cost / s, and is a preset safety parameter. Its value must be set to ensure that it is significantly greater than the cost factor upon activation, thereby forcing the system to prioritize risk avoidance. If the predicted heat stress risk index is not greater than the risk reference threshold... If the result of the max function is zero, then the risk penalty term will be set to zero.

[0120] The specific working method of this control decision module is as follows: Injecting liquid nitrogen intensity... As optimization variables; the cost function is solved using a numerical optimization algorithm. Minimum liquid nitrogen injection intensity This intensity is then output as the optimal liquid nitrogen injection intensity; because It is about The function, therefore the entire cost function It is about The system can quickly find the function that minimizes the total cost in each control cycle using numerical optimization algorithms such as gradient descent. smallest And send it as an instruction to the liquid nitrogen supply unit;

[0121] By constructing this cost function, the control decision module unifies the two objectives of energy saving and quality assurance into a single optimization problem. The design of the risk penalty term defines a safety red line for the system; only when the predicted risk is below a threshold will the system search for the most energy-efficient control point within the safe zone. Once any control action is predicted to potentially touch or cross this red line, the significant penalty cost immediately forces the optimization algorithm to abandon that action and choose a safer control strategy. This prediction and optimization-based decision-making mechanism achieves an intelligent upgrade from passive response to proactive avoidance, ensuring that an optimal balance between safety and efficiency is always achieved during dynamic changes. The optimal liquid nitrogen injection intensity... It can be converted into the opening signal or pulse width modulation (PWM) signal of the liquid nitrogen flow control valve through a preset conversion model or lookup table, so as to achieve physical control of the liquid nitrogen injection power.

[0122] Example 6:

[0123] This invention provides an application of an automatic liquid nitrogen flow rate regulation system in abalone preservation;

[0124] In specific application scenarios, live abalone samples or abalone samples that have just undergone pretreatment are placed in a preservation chamber equipped with liquid nitrogen spraying function; as described in the aforementioned embodiments, micro thermocouples are implanted on the surface and core of the abalone sample, micro impedance electrodes are implanted in the tissue, and high-frequency acoustic transducers are deployed on the inner wall of the preservation chamber. In actual deployment, in order to reduce the impact of invasive measurement on the characteristics of the sample itself, micro sensors with extremely small diameters should be selected, and the implantation position and depth should be standardized to ensure consistency.

[0125] After the system starts up, the data acquisition module measures and records the impedance magnitude in the completely unfrozen state. and initial speed of sound As a baseline parameter, the liquid nitrogen injection system begins operation, and the entire automatic control system enters a closed-loop control process: the data acquisition module continuously collects the real-time core temperature of the abalone. Surface temperature Tissue impedance and speed of sound Thermophysical property calculation module receives and The instantaneous ice crystal integral of the abalone tissue is calculated in real time. and instantaneous thermal conductivity The state prediction module integrates actual measurements. , and solution By using an extended Kalman filter model, the core temperature of the abalone can be accurately predicted at the next moment. and surface temperature The risk prediction module uses the predicted temperature state to calculate the predicted dynamic thermal gradient. Furthermore, by combining proportional and differential risk models, a predicted heat stress risk index is obtained. The control decision module will With safety threshold Comparison, and consideration of both The deviation from the target cooling curve is minimized by the cost function. The optimal liquid nitrogen injection intensity was determined. ;Should The command is sent to the liquid nitrogen flow control valve to precisely adjust the injection intensity and pulse of liquid nitrogen; this process is repeated at a high frequency until the abalone core temperature reaches the preset endpoint temperature.

[0126] Applying the system of this invention to abalone preservation yields significant technological advancements. Due to variations in abalone size, water content, and other factors, their thermophysical parameters also differ. This system provides a personalized cooling control scheme for each abalone, dynamically adjusting the liquid nitrogen flow rate to ensure its core temperature closely follows the optimal cooling curve. This ensures rapid passage through the maximum ice crystal formation zone, effectively inhibiting ice crystal growth and minimizing damage to the abalone's cell structure, thus preserving its thawed texture, flavor, and nutritional value. Furthermore, it avoids energy waste and the risk of surface cracking caused by excessive liquid nitrogen spraying in the later stages of cooling, ultimately achieving a dual improvement in abalone preservation quality and processing economics.

[0127] 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. An automatic liquid nitrogen flow rate regulation system, characterized in that... ,include: The data acquisition module is used to acquire the physical parameters of each individual aquatic product to be processed in real time. The physical parameters include: core temperature, surface temperature, tissue impedance modulus, and sound wave propagation speed. The thermal property calculation module is used to determine the instantaneous thermal conductivity of aquatic products based on the tissue impedance modulus and sound wave propagation velocity obtained by the data acquisition module. The state prediction module combines the core temperature and surface temperature obtained by the data acquisition module with the instantaneous thermal conductivity determined by the thermal property calculation module, and uses an extended Kalman filter model to predict the system state at the next moment, including the core temperature and surface temperature at the next moment. The risk prediction module is used to determine the predicted heat stress risk index for the next moment based on the system state obtained by the state prediction module. The control decision module is used to combine the deviation between the core temperature predicted by the state prediction module and the preset target temperature at the next moment, and the predicted heat stress risk index determined by the risk prediction module, to solve for the minimum value of the preset cost function to determine the optimal liquid nitrogen injection intensity. The risk prediction module is specifically used for: The dynamic thermal gradient is calculated and predicted based on the difference between the core temperature and the surface temperature at the next time step. Based on the predicted dynamic thermal gradient, the predicted proportional risk term is determined; Based on the predicted rate of change of the dynamic thermal gradient, the predicted differential risk term is determined; The predicted proportional risk term and the predicted differential risk term are weighted and summed to obtain the predicted heat stress risk index. The predicted proportional risk term is the ratio of the predicted dynamic thermal gradient to the preset peak dynamic thermal gradient threshold. The predicted differential risk term is the ratio of the predicted rate of change of the dynamic thermal gradient to the preset peak dynamic thermal gradient threshold, multiplied by the preset system thermal response time constant.

2. The automatic liquid nitrogen flow rate adjustment system according to claim 1, characterized in that... The thermal property calculation module is specifically used for: The instantaneous ice crystal integral is determined based on the tissue impedance modulus, the preset impedance modulus of the unfrozen state, and the impedance modulus of the fully frozen state. Instantaneous thermal conductivity is calculated based on a parallel thermal resistance model by combining the instantaneous ice crystal integral number, the sound wave propagation velocity, and the preset acoustic-thermal coupling coefficient.

3. The automatic liquid nitrogen flow rate adjustment system according to claim 1, characterized in that... The state prediction module is specifically used for: Construct the current system state vector, which includes the core temperature and the surface temperature; The current system state vector is predicted by employing a nonlinear state transition equation incorporating instantaneous thermal conductivity. The prediction results are corrected by using an observation equation that incorporates real-time sensor measurements to output the system state at the next moment.

4. The automatic liquid nitrogen flow rate adjustment system according to claim 1, characterized in that... The cost function configured in the control decision module includes an operating cost item and a risk penalty item; the operating cost item is proportional to the liquid nitrogen injection intensity; the risk penalty item is determined in the following way: The predicted heat stress risk index is compared with the preset risk reference threshold. If the predicted heat stress risk index is greater than the risk reference threshold, the risk penalty item is determined based on the difference between the two. If the predicted heat stress risk index is not greater than the risk reference threshold, then the risk penalty term will be set to zero.

5. The automatic liquid nitrogen flow rate adjustment system according to claim 4, characterized in that... The control decision module is specifically used for: Liquid nitrogen injection intensity was used as the optimization variable. The liquid nitrogen injection intensity that minimizes the cost function is determined by a numerical optimization algorithm, and this intensity is output as the optimal liquid nitrogen injection intensity.

6. The application of the liquid nitrogen flow rate automatic adjustment system according to any one of claims 1-5 in abalone preservation.

Images