Method for controlling pickling of poultry eggs, control terminal, control system, and storage medium

Abstract

The application provides an egg pickling control method, a control terminal, a control system and a storage medium, and belongs to the technical field of food processing; the control method is applied to the control terminal and comprises the following steps: receiving pickling data sent by a pickling device, wherein the pickling data comprises measured alkali concentration, measured temperature and measured gel characteristic data in the pickling device; determining state estimation data changing with time in the pickling device according to the pickling data, wherein the state estimation data comprises predicted alkali concentration in the albumen of each egg, predicted gel degree distribution and predicted gel front position of the radial space in the albumen; determining a target alkali supplement flow rate and a target temperature set value of the pickling device according to the state estimation data; and sending the target alkali supplement flow rate and the target temperature set value to the pickling device, so that the pickling device performs pickling control according to the target alkali supplement flow rate and the target temperature set value; and the control method can reduce the pickling quality fluctuation generated in the egg pickling process.

Description

Technical Field

[0001] This application relates to the field of food processing technology, specifically to a control method, control terminal, control system, and storage medium for pickling poultry eggs. Background Technology

[0002] The core of poultry egg curing is soaking the eggs in a solution containing sodium hydroxide (NaOH), salt, and metal ions, specifically alkali ions (OH-). - The liquid proteins in poultry eggs diffuse inward through the eggshell, gradually denaturing and cross-linking to form a gel. Currently, large-scale production of pickled poultry eggs mainly adopts the following two control methods: 1) Fixed parameter open-loop control: Based on experience, the initial alkali concentration (NaOH mass fraction 4%~6%), temperature (15~30℃), and total soaking time (20~40 days) are set without any active intervention; 2) Single-variable closed-loop control: pH or alkali concentration electrodes and temperature sensors are installed in a single tank. The measured alkali concentration is compared with the preset "time-concentration" empirical curve. The proportional-integral-derivative PID algorithm is used to drive the alkali replenishment pump and heater to maintain the alkali concentration on the predetermined trajectory.

[0003] However, all of the above-mentioned large-scale production solutions have the same problem: the quality of poultry eggs in the pickling equipment is prone to fluctuations, leading to an increase in the rate of substandard pickled eggs. Summary of the Invention

[0004] The main technical problem addressed in this application is how to reduce the fluctuations in the pickling quality of poultry eggs during the pickling process.

[0005] According to a first aspect, an embodiment provides a control method for pickling poultry eggs, applied to a control terminal, comprising: receiving pickling data sent by a pickling device; the pickling device containing poultry eggs, the pickling data including measured alkali concentration, measured temperature, and measured gel characteristic data within the pickling device; the measured gel characteristic data including at least the first wave velocity and characteristic frequency band attenuation coefficient of a non-invasive transmission signal, the non-invasive transmission signal being acquired by signal excitation of the pickling device; determining time-varying state estimation data within the pickling device based on the pickling data, the state estimation data including the predicted alkali concentration within the albumen of each poultry egg, the predicted gelation distribution in the radial space within the albumen, and the predicted gel front position; the predicted gel front position being the radial position corresponding to the first time the predicted gelation is lower than a preset threshold in the direction from the inner surface of the eggshell to the yolk in the predicted gelation distribution; determining a target alkali replenishment flow rate and a target temperature setpoint for the pickling device based on the state estimation data; and sending the target alkali replenishment flow rate and target temperature setpoint to the pickling device so that the pickling device performs pickling control according to the target alkali replenishment flow rate and target temperature setpoint.

[0006] In one embodiment, determining the state estimation data of the pickling device changing over time based on the pickling data includes: using the pickling data as an observation value and inputting it into a particle filter state estimator with a digital twin model as the state transition operator to obtain the state estimation data.

[0007] In one embodiment, the digital twin model includes a mapping relationship between the effective diffusion coefficient of the base and the gelation degree of the protein, as well as a differential equation for the evolution of gelation degree.

[0008] The mapping relationship satisfies: ;

[0009] The differential equation is: ;

[0010] Among them, D eff Where is the effective diffusion coefficient of the alkali, D0 is the intrinsic diffusion coefficient of the alkali ions in the uncoagulated protein, G is the gel strength of the protein, β is the gel resistance index and β > 1; C is the alkali concentration, in mol / m³. 3 t is time, k g (T) is the rate constant dependent on temperature T, m is the reaction order of the alkali concentration, and p is the gelation reaction order.

[0011] In one embodiment, the particle filter state estimator is used to: employ a particle filter algorithm, using a digital twin model as a state transition model and measured gel feature data as observations; map the state represented by each particle to predicted gel feature data using the observation model; calculate the deviation between the predicted gel feature data and the measured gel feature data, and update the weight of each particle according to the deviation to correct the state estimate represented by each particle.

[0012] In one embodiment, an observation model is used to map the state represented by each particle into predicted gel feature data, including: determining the average gel degree of the protein in each egg based on the predicted gel degree distribution; determining the gel layer thickness of the protein in each egg based on the predicted gel front position and the size of the egg; inputting the average gel degree and gel layer thickness of all eggs into the observation model, and outputting predicted gel feature data corresponding to the measured gel feature data.

[0013] In one embodiment, determining the target alkali replenishment flow rate and target temperature setpoint of the pickling device based on state estimation data includes: using the state estimation data as input, performing rolling optimization solution through a model predictive control method to obtain the requested alkali replenishment flow rate and temperature setpoint; and determining the target alkali replenishment flow rate and target temperature setpoint based on the requested alkali replenishment flow rate and temperature setpoint.

[0014] In one embodiment, there are multiple pickling devices, and the rolling optimization solution is performed by a model predictive control method, including: for the k-th pickling device, in the prediction time domain, using the first model predictive controller corresponding to the k-th pickling device, minimizing the first objective function as the objective, and solving based on the first constraint condition;

[0015] First objective function J k for:

[0016] ;

[0017] The first constraint includes:

[0018] C bulk,min ≤ C bulk (t) ≤ C bulk,max ;

[0019] T min ≤ T sp (j) ≤ T max ;

[0020] | T sp (j+1) - T sp (j) | ≤ ΔT rate ;

[0021] 0 ≤ v alk (j) ≤ v max ;

[0022] Where ω1, ω2, and ω3 are weighting coefficients. For in t fin Average gelation degree at time t fin To predict the first time the gelation degree reaches the target gelation degree G target The estimated time, Hp is the number of periods corresponding to the prediction time domain, v alk (j) represents the alkali replenishment flow rate in the j-th period of the prediction time domain, T sp (j) represents the temperature setpoint for the j-th period in the prediction time domain, T opt The preset reference marinating temperature, C bulk (t) represents the alkali concentration at time t, C bulk,min C is the preset minimum alkali concentration. bulk,max T is the preset maximum alkali concentration. min T is the preset minimum temperature. max T is the preset maximum temperature. sp (j+1) is the temperature setpoint for the (j+1)th period in the prediction time domain, ΔT rate v is the preset maximum temperature change. maxThe above values ​​represent the preset maximum alkali replenishment flow rate. All alkali replenishment flow rates are in L / min, and all temperatures are in °C.

[0023] In one embodiment, determining the target alkali replenishment flow rate and target temperature setting value based on the requested alkali replenishment flow rate and the requested temperature setting value includes: determining the total required heating power of all pickling devices based on the temperature setting request value of each pickling device, and determining the total required alkali replenishment flow rate of all pickling devices based on the requested alkali replenishment flow rate of each pickling device; when it is detected that the total required heating power does not exceed the preset upper limit of total heating power and the total required alkali replenishment flow rate does not exceed the preset upper limit of total alkali replenishment flow rate, the requested alkali replenishment flow rate value is used as the target alkali replenishment flow rate and the requested temperature setting value is used as the target temperature setting value; when it is detected that the total required heating power exceeds the preset upper limit of total heating power or the total required alkali replenishment flow rate exceeds the preset upper limit of total alkali replenishment flow rate, the target alkali replenishment flow rate and target temperature setting value of each pickling device are obtained by rolling optimization solution using a model predictive control method based on the total required heating power and the total required alkali replenishment flow rate.

[0024] In one embodiment, rolling optimization is performed using a model predictive control method based on the total demand heating power and the total demand alkali replenishment flow rate. This includes: inputting the total demand heating power and the total demand alkali replenishment flow rate into a second model predictive controller, and solving the problem based on a second constraint condition with the goal of minimizing a second objective function.

[0025] Second objective function J global for:

[0026]

[0027] The second constraint includes the first constraint, and:

[0028] ;

[0029] ;

[0030] Where N is the number of pickling devices, N≥2 and is an integer, t fin,k The predicted gelation degree of the k-th pickling device first reaches the target gelation degree G. target The estimated time, T ref The ideal curing time for poultry eggs is given by ρ, where ρ is a weighting factor. The total heating power required. To meet the demand for alkali replenishment, P total To preset the maximum total heating power, F total This is the preset upper limit for the total flow rate of alkali replenishment.

[0031] In one embodiment, after the pickling device controls the pickling process according to the target alkali replenishment flow rate and target temperature setpoint, the control method further includes: after the pickling device completes the pickling process, obtaining the measured gel distribution of the pickled poultry eggs; and adjusting the model parameters of the digital twin model according to the deviation between the measured gel distribution and the predicted gel distribution.

[0032] According to a second aspect, one embodiment provides a control terminal for pickling poultry eggs, comprising:

[0033] A memory for storing computer programs; a processor for executing the computer programs to implement the control methods of any of the first aspect embodiments.

[0034] According to a third aspect, one embodiment provides a control system for pickling poultry eggs, including a control terminal and at least one pickling device. The pickling device is used to load poultry eggs; the pickling device is used to acquire pickling data and send the pickling data to the control terminal; receive target alkali replenishment flow rate and target temperature setpoints sent by the control terminal, and perform pickling control according to the target alkali replenishment flow rate and target temperature setpoints; the control terminal is used to receive pickling data; determine the corresponding state estimation data of the pickling device over time based on the pickling data; and determine the target alkali replenishment flow rate and target temperature of the corresponding pickling device based on the state estimation data. The set values ​​are sent to the corresponding pickling device, including the target alkali replenishment flow rate and target temperature. The pickling data includes the measured alkali concentration, measured temperature, and measured gel characteristic data within the pickling device. The measured gel characteristic data is determined based on a non-invasive transmission signal, which is obtained by exciting the pickling device. The state estimation data includes the predicted alkali concentration, predicted gel distribution, and predicted gel front position of the protein of each egg over time. The predicted gel front position is the innermost radial position corresponding to the first time the predicted gel degree reaches the preset threshold in the predicted gel distribution.

[0035] In one embodiment, the pickling apparatus includes a pickling container for holding poultry eggs and a sensor assembly disposed on the pickling container for acquiring pickling data; a first communication unit for transmitting pickling data and receiving a target alkali replenishment flow rate and a target temperature setpoint; and an execution unit for performing pickling control according to the target alkali replenishment flow rate and the target temperature setpoint.

[0036] In one embodiment, the control terminal includes: a second communication unit for receiving pickling data and sending a target alkali replenishment flow rate and a target temperature setpoint; and a processor for determining, based on the pickling data, the corresponding state estimation data of the pickling device over time; and determining the target alkali replenishment flow rate and the target temperature setpoint based on the state estimation data.

[0037] In one embodiment, the sensor assembly includes an ultrasonic sensor disposed on the outer wall of the pickling container; the ultrasonic sensor includes a transmitting probe, a receiving probe, and a processing unit.

[0038] The transmitting probe is used to transmit ultrasonic excitation signals, which excite the outer wall to form guided waves. The guided waves propagate in the pickling container to generate ultrasonic waves, which include direct waves from the outer wall, multiple reflected waves, and transmitted waves that pass through the eggs. The receiving probe is used to receive the ultrasonic waves. The processing unit is used to remove the direct waves from the outer wall and multiple reflected waves, and select the transmitted waves within the first wave arrival time window for feature extraction to obtain measured gel feature data.

[0039] According to a fourth aspect, one embodiment provides a computer-readable storage medium including a computer program that can be executed by a processor to implement the control method of any of the embodiments of the first aspect.

[0040] According to the control method for pickling poultry eggs provided in the above embodiments, state estimation data is determined based on the measured alkali concentration, measured temperature, and measured gelation characteristic data in the pickling device. The predicted gelation distribution, predicted gel front position, and predicted alkali concentration in the state estimation data reflect the constraint relationship of "alkali-gel coupling" in the pickling process of poultry eggs, that is, the real-time, bidirectional, and nonlinear constraint relationship between alkali diffusion mass transfer and protein gelation reaction. Therefore, by determining the alkali replenishment flow rate and temperature setpoint of the pickling device based on the state estimation data reflecting the constraint relationship of alkali-gel coupling and performing pickling control, the pickling quality fluctuation caused by the nonlinearity of "alkali-gel coupling" can be eliminated, and the defect rate of pickled poultry eggs can be significantly reduced. Attached Figure Description

[0041] Figure 1 This is a schematic flowchart illustrating a method for controlling the pickling of poultry eggs according to one embodiment.

[0042] Figure 2 This is a schematic diagram illustrating the principle of an alkali-gel coupling model according to one embodiment;

[0043] Figure 3 This is a schematic diagram of the control logic for a two-layer MPC group control scheduling according to one embodiment;

[0044] Figure 4 This is a schematic diagram of a control terminal according to one embodiment;

[0045] Figure 5 This is a schematic diagram of a control system according to one embodiment;

[0046] Figure 6 A schematic diagram of the architecture of a control system for pickling preserved eggs according to one embodiment;

[0047] Figure 7This is a schematic diagram illustrating the working principle of an air-coupled ultrasonic guided wave sensor according to one embodiment.

[0048] Figure 8 This is a schematic diagram of the control flow of a control system according to one embodiment;

[0049] Figure 9 This is a schematic diagram of a computer-readable storage medium according to one embodiment. Detailed Implementation

[0050] The present application will now be described in further detail with reference to the accompanying drawings and specific embodiments. Similar elements in different embodiments are referred to by related similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the present application. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to the present application are not shown or described in the specification. This is to avoid obscuring the core parts of the present application with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0051] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0052] The serial numbers assigned to components in this document, such as "first" and "second," are used only to distinguish the described objects and have no sequential or technical meaning. The terms "connection" and "linkage" used in this application, unless otherwise specified, include both direct and indirect connections (linkages).

[0053] Studies have shown that the main reason why common poultry egg curing methods are prone to quality fluctuations is that they fail to consider the dynamic coupling relationship between alkali mass transfer and protein gelation during the curing process, thus exhibiting the following fundamental flaws:

[0054] 1) Unable to explain and resolve the concurrent quality issues of "alkali damage" and "alkali damage"; alkali ions (OH-) -The penetration of alkali ions causes proteins to form a dense gel layer. Once this gel layer is formed, it significantly increases the diffusion resistance of alkali ions, slowing down the subsequent penetration rate of alkali ions and creating a non-linear negative feedback loop of "the more solidified, the harder it is to penetrate." Therefore, simply maintaining a constant alkali concentration in the pickling device does not mean that the alkali supply inside the egg is appropriate. If the alkali concentration is slightly high in the early stage, the outer layer of the egg will quickly over-coagulate or even liquefy, which will hinder the entry of alkali ions into the inner layer, resulting in the phenomenon of "external liquefaction and internal fluidization," which is called "alkali damage." If the alkali concentration is too low in the early stage, the overall coagulation will be insufficient, and there will still be a large number of uncoagulated areas inside the egg at the end of the pickling cycle, which is called "alkali damage." The existing pickling control methods cannot cope with the above non-linear processes, resulting in a high rate of defective pickled eggs.

[0055] 2) The coagulation state inside the egg cannot be obtained; as a key indicator that determines the quality of pickled egg products, the spatial distribution of gelation degree inside the egg and the position of the gel front cannot be directly measured; operators can only judge the pickling process by periodically breaking the eggs during pickling and sampling them, which is very damaging to the eggs and seriously delays the progress.

[0056] 3) Individual differences in the pickling device group cannot be coordinated; when multiple pickling devices are in parallel production, the progress of the "alkali-gel coupling" process of each pickling device is different due to differences in the size of fresh eggs, eggshell thickness, initial temperature, etc.; if all pickling devices use the same pickling control parameters, the final coagulation degree of poultry eggs in each pickling device will inevitably be inconsistent, making it impossible to achieve synchronous maturation and standardized production of the whole flock.

[0057] Based on the above research findings, in order to solve the quality fluctuation problem in the pickling process of poultry eggs, this application provides a control method for pickling poultry eggs. The overall technical concept is to determine the state estimation data that can reflect the constraint relationship of "alkali-gel coupling" in the pickling process based on the pickling data, and then determine the alkali replenishment flow rate and temperature setpoint of the pickling device based on the state estimation data and carry out pickling control, thereby eliminating the quality fluctuation of pickling caused by the nonlinearity of "alkali-gel coupling".

[0058] In one implementation, a method for controlling the pickling of poultry eggs is provided and applied to a control terminal. Please refer to [link to relevant documentation]. Figure 1 The control method includes steps S101 to S104, as follows:

[0059] S101: Receive pickling data sent by the pickling device; the pickling device contains poultry eggs, and the pickling data includes the measured alkali concentration, measured temperature and measured gel characteristic data in the pickling device; the measured gel characteristic data includes at least the first wave velocity and characteristic frequency band attenuation coefficient of the non-invasive transmission signal, and the non-invasive transmission signal is obtained by signal excitation of the pickling device.

[0060] In the field of poultry and egg curing, curing equipment can be curing jars, such as the earthenware jars commonly used in curing preserved eggs.

[0061] Pickling data can be acquired through sensor components installed on the pickling device; the measured alkali concentration refers to the real-time alkali concentration of the pickling solution within the pickling device, which can be measured using hydroxide ions (OH-). - The concentration is expressed as 0, and the corresponding sensor is a strong alkali-resistant ion-selective electrode with a range of 0~10. 4 mol / m 3 NaOH, with an accuracy of ±0.5%, can measure the OH content of the pickling solution in real time. - Concentration C bulk .

[0062] The measured temperature refers to the temperature of the pickling liquid inside the pickling device. It can be measured by a temperature sensor, such as a platinum resistance thermometer, with an accuracy of ±0.1℃. The temperature T of the pickling liquid can be determined in real time. The measured temperature also serves as the feedback basis for temperature control.

[0063] The measured gel feature data are feature data or feature vectors obtained by feature extraction from non-invasive transmitted wave signals. They indirectly reflect the curing state inside the egg, such as the average gelation degree and gel layer thickness.

[0064] Non-invasive refers to detecting eggs without damaging them during pickling. Commonly used non-invasive transmission wave signals can be obtained through ultrasonic excitation. For example, a pair of air-coupled ultrasonic transducer probes with a center frequency of 100 kHz can be used, attached to the same horizontal axis on the outer wall of the pickling device. One probe is the transmitting probe, and the other is the receiving probe. The transmitting probe generates guided waves by exciting the cylinder wall with narrow pulses. Part of the energy in the guided waves is transmitted into the liquid and passes through the egg cluster to generate transmitted waves. The receiving probe collects the transmitted waves. By analyzing the first wave velocity v and the characteristic frequency band attenuation coefficient α of the transmitted waves, the feature vector is extracted by Fast Fourier Transform (FFT) and 80kHz~120kHz bandpass filtering to obtain the measured gel feature data. This feature data can indirectly reflect the average gelation degree and gel layer thickness inside the egg.

[0065] In addition to ultrasonic transmission waves, non-invasive methods such as microwave and elastic wave transmission detection can also be used to obtain measured gel characteristic data. This application does not specifically limit the form of non-invasive transmission signals.

[0066] S102: Determine the state estimation data of the pickling device changing over time based on the pickling data. The state estimation data includes the predicted alkali concentration inside the albumen of each egg, the predicted gelation distribution in the radial space inside the albumen, and the predicted gel front position. The predicted gel front position is the radial position corresponding to the first time the predicted gelation is lower than the preset threshold in the direction from the inner surface of the eggshell to the yolk in the predicted gelation distribution.

[0067] State estimation data reflects the state evolution of the lye solution and eggs within the curing apparatus, including data from three aspects:

[0068] Predicting gelation distribution The spatial distribution of predicted gelation degree obtained through simulation calculation, r j This represents the spatial grid nodes within the egg body partitioned in the simulation calculation, therefore This represents the dataset of predicted gelation degree corresponding to each spatial grid node in the egg body at the current moment; the gelation degree G at each node can be a normalized gelation degree index, i.e. G∈[0,1], where G represents the degree of protein transition from liquid to solid state, where 0 is fully liquid and 1 is completely solidified.

[0069] Predicting the location of the gel front R gel Based on predicted gelation distribution The defined parameters indicate the predicted gelation degree when it first falls below the preset threshold G during a scan from the inner surface of the eggshell towards the yolk. threshold The corresponding radial coordinates predict that the gel front position continuously advances towards the yolk as the curing time increases; G threshold To determine the gel strength threshold at the gel front, such as 0.5; for example, assuming an egg is considered a homogeneous sphere of radius R, with radial coordinate r=R at the eggshell, at time t1, based on the predicted gel strength distribution... Determine G=G threshold If the innermost radial coordinate is at 3R / 4, then it represents R. gel Located at a distance R / 4 from the eggshell; at time t2, G=G threshold If the innermost radial coordinate is at R / 2, then it means R gel Located at a distance of R / 2 from the eggshell; when R gel When the egg white reaches the surface of the yolk (r0), it indicates that the entire egg white has basically solidified.

[0070] Predicted alkali concentration : Represents the number of free OH groups at each grid node j in the egg white, obtained through simulation calculations. - Ion concentration, in units of mol / m³ 3 It can be used to predict future gelation trends.

[0071] S103: Based on the state estimation data, determine the target alkali replenishment flow rate and target temperature setpoint of the pickling device.

[0072] S104: Send the target alkali replenishment flow rate and target temperature setpoint to the pickling device so that the pickling device can perform pickling control according to the target alkali replenishment flow rate and target temperature setpoint.

[0073] The target alkali replenishment flow rate and target temperature setpoint are two important control variables in the poultry egg curing process; as the curing process continues, the alkali concentration in the curing device will continue to decrease, i.e., OH- - During the diffusion process into the egg, the free OH groups are continuously consumed due to the protein gelation reaction, leading to a decrease in the free OH groups within the pickling device. - The concentration continues to decrease; if no active alkali replenishment is performed, the alkali concentration in the tank will continue to decline, eventually leading to insufficient pickling momentum, incomplete egg coagulation, and a "alkali damage" defect. Therefore, the target alkali replenishment flow rate determined by state estimation data is an active control method to maintain the continuous progress of the pickling process. Thus, the target alkali replenishment flow rate corresponds to the replenishment rate of the high-concentration NaOH solution in the pickling device. The higher the value of the target alkali replenishment flow rate, the more alkali (OH) is added to the pickling device per unit time. - The more alkali added, the higher the alkali concentration in the pickling device; therefore, the target alkali replenishment flow rate is a direct control variable for controlling the gelation process inside the egg.

[0074] The target temperature setpoint corresponds to the pickling setpoint within the pickling device. It is the target temperature that the pickling solution in the pickling tank should maintain, as indicated by the control terminal. It is one of the active control variables that directly affects the gelation reaction rate. If the pickling temperature increases, the protein gelation rate accelerates, and the pickling process speeds up. If the pickling temperature decreases, the gelation rate slows down, and the pickling process decelerates. Therefore, by setting a reasonable target temperature setpoint, the pickling process and pickling quality can be adjusted.

[0075] It should be noted that the pickling temperature can regulate the gelation rate, which in turn affects the rate of increase of gelation degree G. The increase of gelation degree G, in turn, affects the diffusion rate of alkali. Therefore, the adjustment of the target temperature setting and the target alkali replenishment flow rate will produce an alkali-gel coupling cascade effect during the pickling process. The relationship between the two is synergistic optimization, rather than independent adjustment.

[0076] It should be noted that if there are multiple pickling devices, it is necessary to obtain the pickling data of each pickling device, then determine the state estimation data based on the pickling data, and then determine the corresponding target alkali replenishment flow rate and target temperature setpoint based on the state estimation data of each pickling device.

[0077] The poultry egg pickling control method provided in this application determines state estimation data based on measured alkali concentration, measured temperature, and measured gelation characteristic data within the pickling device. The predicted gelation distribution, predicted gel front position, and predicted alkali concentration in the state estimation data reflect the "alkali-gel coupling" constraint relationship during the poultry egg pickling process, i.e., the real-time, bidirectional, and nonlinear constraint relationship between alkali diffusion mass transfer and protein gelation reaction. Therefore, by determining the alkali replenishment flow rate and temperature setpoint of the pickling device based on the state estimation data reflecting the alkali-gel coupling constraint relationship and performing pickling control, the quality fluctuations caused by the nonlinearity of the "alkali-gel coupling" can be eliminated, significantly reducing the defect rate of pickled poultry eggs.

[0078] The above embodiments illustrate the overall concept of this application. In the following embodiments, the above solution will be further described.

[0079] In this application, "alkali-gel coupling" refers to the real-time, bidirectional, and nonlinear constraint relationship between the diffusion mass transfer of alkali and the protein gelation reaction during the pickling process of poultry eggs. This application defines and quantifies the quantitative characteristics of "alkali-gel coupling" as follows:

[0080] 1) The diffusion of alkali drives the gelation reaction, and the gelation reaction rate increases with the increase of local alkali concentration;

[0081] 2) An increase in the degree of gelation, i.e., an increase in the degree of gelation G, will lead to a significant decrease in the effective diffusion coefficient of alkali in the protein, which manifests as a strong hindrance to subsequent alkali penetration.

[0082] The aforementioned coupling effect, in which the diffusion coefficient decreases with increasing reaction products, is the fundamental reason for the failure of traditional control strategies.

[0083] Based on the quantitative characterization of alkali-gel coupling above, in some embodiments, the step S102 of determining the state estimation data of the pickling device changing over time based on the pickling data includes: taking the pickling data as the observation value and inputting it into a particle filter state estimator with a digital twin model as the state transition operator to obtain the state estimation data.

[0084] A digital twin model is a digital twin and a detection model created and run on a control terminal. It takes the alkali-gel coupling partial differential equation system as its core and the measured alkali concentration and measured temperature as boundary condition inputs to extrapolate the alkali concentration distribution and gelation distribution inside the egg over time. Since the internal state of the egg cannot be directly measured, the digital twin model plays the role of a "virtual sensor" and outputs predicted values ​​of the internal state, i.e., state estimation data.

[0085] The core of the digital twin model is the alkali-gel coupling nonlinear mechanism model. In some embodiments, for convenience, a single duck egg is simplified as a homogeneous sphere, with the albumen region extending from the yolk surface (r=r0) to the inner shell surface (r=R), where R is the radius of the homogeneous sphere corresponding to the egg. The following set of coupled partial differential equations is thus established:

[0086] 1) The mass conservation of base, i.e., the diffusion-reaction equation:

[0087] (1)

[0088] Where C(r,t) is the base concentration, which can be expressed as free OH-. - Concentration (alkali concentration, mol / m) 3 The alkali concentration C is a spatial distribution function of the egg body radially with respect to position r and time t; similarly, G(r,t) is the gelation distribution, which is a spatial distribution function of the gelation degree G in the egg body radially with respect to position r and time t, describing the degree of coagulation of each protein layer from the eggshell to the yolk; the first term on the right side of the equation (1) is a nonlinear diffusion term used to describe OH - Radial diffusion in spherical coordinates, D eff D is the effective diffusion coefficient of the alkali. eff (G) represents the nonlinear change of the diffusion coefficient with respect to gelation degree G; the second term is the pooled term for the consumption of alkali by protein gelation, used to describe OH- - The rate at which it is consumed by the protein gelation reaction, k c This is the consumption rate constant, with units of 1 / s.

[0089] 2) Effective diffusion coefficient, which reflects the key to "alkali-gel coupling":

[0090] D eff =D0×(1-G) β (2)

[0091] The above formula reflects the effective diffusion coefficient D of the alkali. eff The mapping relationship between D0 and the gelation degree G of protein; D0 is the intrinsic diffusion coefficient of base ions in uncoagulated protein, β is the gelation resistance index and β > 1; D0 is the OH- ions in uncoagulated protein. - The intrinsic diffusion coefficient (approximately 10) -9 m 2 / s on the order of magnitude), G∈[0,1] is the normalized gelation degree (G=0: fully liquid; G=1: completely solidified). As G increases, D eff The effect of drastic reduction perfectly captures the self-inhibitory effect of the gel layer on subsequent alkali penetration.

[0092] 3) Protein gelation kinetics, i.e., the differential equation for the evolution of gelation degree:

[0093] (3)

[0094] And when t ≤ t ind At that time, G=0.

[0095] Set the initial conditions as G(r,0)=0, and let the reaction rate term C(r,t) be... m It automatically approaches zero at low concentrations (because the concentration is low at this time), without the need for segmentation;

[0096] in, k is a temperature-dependent gelation rate constant. g0 is the pre-exponential factor or frequency factor of the gelation reaction, which is related to the collision frequency of the reactants and can be obtained by fitting experimental data; Ea is the activation energy of the gelation reaction, representing the energy barrier that the reaction needs to overcome; R is the universal gas constant, 8.314 J / (mol·K), T is the temperature; m is the reaction order with alkali concentration, reflecting the OH- . - The degree of nonlinear effect of concentration on gelation rate; p is the gelation reaction order, reflecting the driving effect of the proportion of uncoagulated protein (1-G) on the reaction rate; t ind This is the gelation induction time, which can be considered a constant, expressed in hours or days, representing OH. - The delay between the start of penetration and the appearance of a detectable gel occurs during which time G=0.

[0097] 4) Definite solution conditions:

[0098] Initial conditions: C(r,0)=0, G(r,0)=0.

[0099] Boundary conditions:

[0100] On the inner surface of the eggshell: at r=R, a third-type boundary condition (flux continuity type) is applied to account for the mass transfer resistance of the eggshell.

[0101] ;

[0102] Among them, h m The mass transfer coefficient of the eggshell is expressed in m / s and needs to be calibrated experimentally; when h m As the boundary condition approaches infinity, it degenerates into C(R,t)=C bulk (t);

[0103] Egg yolk interface: r=r0, There is no flux boundary.

[0104] Among them, C bulk (t) represents the measured alkali concentration (OH-) of the pickling solution over time t. - Concentration, unit: mol / m 3), which serves as the boundary condition for the inner surface of the eggshell.

[0105] The principle of the alkali-gel coupling model corresponding to the digital twin model can be found in [reference needed]. Figure 2 The curing solution located outside the eggshell is the external driving force for the curing mass transfer process, C bulk The measured alkali concentration of the pickling solution is represented by C, and the temperature of the pickling solution is represented by T, which serves as the external input to the alkali-gel coupling model. bulk T and C are the objects directly affected by the two core control variables (alkali replenishment flow rate and temperature setpoint) output by the control terminal, among which the alkali replenishment flow rate affects C. bulk The temperature setting affects the temperature T of the marinade.

[0106] alkali ions OH in the pickling solution - Under the influence of the concentration gradient, diffusion drives entry into the eggshell; r=R is the outer boundary of the alkali-gel coupling partial differential equation system, corresponding to the interface between the eggshell and the albumen; C(R,t) = C bulk (t) is the first type of boundary condition for the inner surface of the eggshell, indicating that the alkali concentration at the inner surface of the eggshell is equal to the measured alkali concentration of the external pickling solution, and t is the sampling time.

[0107] As the alkali diffuses into the protein region and is partially consumed by gelation, the diffusion-reaction equation corresponding to equation (1) applies. The gel layer corresponds to equation (3), which describes the growth dynamics of gelation degree G over time in this region, representing the region in the protein that has undergone gelation, which gradually progresses from the inner surface of the eggshell towards the yolk.

[0108] As the gel layer densifies, it hinders OH... - Further diffusion inward (corresponding to equation (2)), the digital twin model applies a flux-free boundary condition at r=r0, where r=r0 is the internal boundary of the alkaline-gel coupling partial differential equation system, corresponding to the interface between the protein and the yolk; the physical meaning of the flux-free boundary is that the yolk has a high flux-free effect on OH-. - Completely opaque, the alkali will not penetrate into the yolk, so the concentration normal gradient at the yolk interface is zero; the flux-free boundary condition and the first kind of boundary condition together constitute the complete boundary conditions of the partial differential equation system, making the equation system numerically unique.

[0109] In summary, by driving a pre-created digital twin model (i.e., a set of alkali-gel coupled partial differential equations) forward, prior predictions of the alkali concentration and gelation distributions at each radial node within the egg can be obtained. Therefore, the core function of the digital twin model is to map measurable macroscopic boundary conditions (i.e., measured alkali concentration and measured temperature, etc.) to the intangible microscopic state within the egg. It should be noted that relying solely on the forward deduction of the mechanistic model introduces model errors and parameter uncertainties. Therefore, a state estimation method is needed to fuse the model predictions with measured values, perform real-time correction on the digital twin model's output, and output a posterior estimate that more closely approximates the true state—the final state estimation data. Where k≥1 and is an integer, corresponding to the number of the pickling device, and i represents the sampling time; the state estimation data of a pickling device at the i-th sampling time is a complete estimated state vector, which is a composite array, specifically including: the predicted gel distribution. Predicting the position of the gel front R gel and predicting alkali concentration .

[0110] In some embodiments, the particle filter state estimator is used to: employ a particle filter algorithm, using a digital twin model as a state transition model and measured gel feature data as observations; map the state represented by each particle to predicted gel feature data using the observation model; calculate the deviation between the predicted gel feature data and the measured gel feature data, and update the weight of each particle according to the deviation to correct the state estimate represented by each particle.

[0111] The particle filtering algorithm uses a digital twin model as the state transition operator. It corrects the state variables (alkali concentration distribution and gelation distribution) of each particle, rather than the physical parameters of the digital twin model. The execution process of the particle filtering algorithm includes three sub-steps: state prediction, which recursively extrapolates Δt from the posterior state of the previous time step and the digital twin model to obtain the prior particle set; weight update, which calculates the likelihood of each particle using the observation model h(x), and the likelihood value is determined by the deviation between the predicted gelation feature data and the measured gelation feature data; resampling, which avoids particle degradation and ensures the effectiveness of the particle set; and finally, the weighted average of all resampled particles is used as the state estimate data for the current time step.

[0112] Therefore, the collaborative logic of particle filtering combined with digital twins is as follows: particle filtering requires a mathematical model describing how the system state evolves over time. This mathematical model is implemented through a digital twin model, where each particle represents a possible internal state (such as gelation distribution). The digital twin propagates the state of each particle forward by one time step. The gelation distribution cannot be directly measured by sensors, and when the digital twin model runs alone, its internal state will gradually deviate from the true value due to model errors and uncertainties in initial conditions, and it does not have the ability to correct itself. By fusing observable measured gelation feature data in real time through particle filtering, the internal state of the digital twin is corrected online based on the deviation between the measured gelation feature data and the predicted gelation feature data, thus achieving soft measurement.

[0113] Besides the particle filter algorithm, other filtering methods such as the linear Kalman filter can also be used. The reason why the particle filter algorithm is used in this embodiment is that it can handle the state estimation problem of nonlinear and non-Gaussian systems and is compatible with the strong nonlinear characteristics of the alkali-gel coupling mechanism model. The particle filter realizes the posterior estimation of the gelation degree distribution and the position of the gel front through the weighted statistics of the particle set, which can accurately describe the probability distribution of the nonlinear system and ensure that the state estimation results maintain high accuracy in each stage of the pickling process.

[0114] In some embodiments, the observation model is used to map the state represented by each particle to predicted gel feature data, including: determining the average gel degree of the protein of each egg based on the predicted gel degree distribution; determining the gel layer thickness of the protein of each egg based on the predicted gel front position and the size of the egg; inputting the average gel degree and gel layer thickness of all eggs into the observation model, and outputting the predicted gel feature data corresponding to the measured gel feature data.

[0115] Specifically, average gel strength It is the spatially weighted average of the predicted gelation distribution G(r,t) within the albumen region of the egg, reflecting the overall coagulation degree of the egg and used to judge the overall curing progress; the gel layer thickness L gel =RR gel The distance between the inner surface of the eggshell and the gel front reflects the thickness of the solidified area and is a direct indicator for assessing the pickling progress.

[0116] In some embodiments, the observation model employs a three-layer feedforward neural network pre-trained with experimental data to calculate the likelihood of each particle; the model input is the average gel strength. and gel layer thickness L gel Output predicted gel feature data Particle degradation is avoided through resampling; the experimental data used for model training is obtained by simultaneous destructive sampling to acquire the actual gelation distribution, and corresponding ultrasonic feature vectors are collected; after each pickling batch is completed, the neural network is incrementally learned or fine-tuned using new data.

[0117] In summary, the observation model effectively solves the complex nonlinear mapping problem from microscopic gel state to macroscopic ultrasonic features by extracting physical features (average gelation degree, gel layer thickness) and fitting with neural networks. This enables the particle filtering observation update step to be executed accurately, thereby establishing a precise mapping between internal unmeasurable states and external measurable ultrasonic signals. Furthermore, with the accumulation of pickling batches, the observation model can be continuously optimized using more experimental data, further improving the accuracy of ultrasonic feature prediction.

[0118] After obtaining the state estimation data, the next step is to determine the target alkali replenishment flow rate and target temperature setpoint based on the state estimation data. In some embodiments, for step S103, the state estimation data is used as input, and rolling optimization is performed through the model predictive control method to obtain the requested alkali replenishment flow rate and temperature setpoint. Based on the requested alkali replenishment flow rate and temperature setpoint, the target alkali replenishment flow rate and target temperature setpoint are determined.

[0119] Model Predictive Control (MPC) uses a rolling optimization method to solve control problems. Its approach is to use the current state estimate data as the initial state, and within a finite prediction time domain, use the output of a digital twin model to make multi-step predictions of the future pickling process and solve for an optimal control sequence that minimizes a predefined objective function.

[0120] After introducing MPC control, the pickling of poultry eggs can be transformed from a passive response to an active predictive control. Traditional proportional-integral-derivative (PID) control only passively adjusts based on the current deviation and cannot predict future nonlinear changes caused by "alkali-gel coupling". MPC control, on the other hand, simulates the future pickling process in the prediction time domain, identifies the densification trend of the gel layer in advance, and adjusts the alkali supplementation and temperature strategies in advance, thus achieving true feedforward + feedback composite control.

[0121] On the other hand, MPC control can simultaneously optimize two control variables: alkali replenishment flow rate and temperature setpoint. It also balances quality objectives and control energy consumption in the objective function, avoiding the "one-sided" problem in single-variable control, thus achieving multi-variable synergistic optimization.

[0122] It should be noted that the optimization results are determined as the "requested values" of the alkali replenishment flow rate and temperature settings through the MPC method, rather than directly determining the "target values". This is because other factors are taken into account, such as whether the total demand of multiple pickling devices exceeds the design limit or whether other limiting factors are met. On the other hand, determining the "target values" based on the "requested values" can effectively avoid the problem of resource constraints and better match the resource status of the entire pickling system.

[0123] Based on the above concept, for a pickling system with multiple pickling devices, a two-layer adaptive MPC control process can be designed. The first layer is the lower-level MPC control corresponding to the pickling device, which is used to determine the alkali replenishment flow rate request value and temperature setting request value of each pickling device. The second layer is the upper-level group control MPC control, which is used to determine the final target alkali replenishment flow rate and target temperature setting value of each pickling device according to the alkali replenishment flow rate request value and temperature setting request value of each device.

[0124] For the lower-level MPC, in some embodiments, there are multiple marinating devices. The rolling optimization solution is performed using a model predictive control method, including: for the k-th marinating device, in the prediction time domain, using a first model predictive controller corresponding to the k-th marinating device, with the goal of minimizing a first objective function, and solving based on a first constraint condition.

[0125] Specifically, each pickling device is configured with an independent MPC controller (i.e., the first model prediction controller), based on the current state estimation data output by the digital twin model corresponding to that pickling device. Starting from the first objective function and the first constraint condition, a rolling optimization solution is performed; the rolling optimization problem is solved in the prediction time domain Hp, and the first element of the optimal control sequence is output as the current control command.

[0126] First objective function J k for:

[0127] (4)

[0128] In the above formula: ω1 is the quality deviation weight, which acts on the deviation of gelation degree from the target value when leaving the vat, guiding the first model predictive controller to prioritize ensuring the final quality of the product;

[0129] ω2 is the alkali consumption weight, which acts on the square term of the alkali replenishment flow rate, guiding the first model predictive controller to save alkali solution as much as possible while meeting quality requirements;

[0130] ω3 is the energy consumption weight, which is applied to the square of the temperature deviation from the optimal value, guiding the first model to predict the controller to reduce unnecessary heating or cooling energy consumption;

[0131] For in tfin Average gelation degree at time t; fin To predict the first time the gelation degree reaches the target gelation degree G target The estimated time, t fin It is a key indicator for MPC optimization and a core variable for group control scheduling and coordination; It can be in t fin At time, the predicted gel strength distribution of the digital twin model output The average gelation degree obtained by spatial averaging corresponds to the average gelation degree of poultry eggs at the moment of completion of pickling; correspondingly, the target gelation degree G target This represents the preset average gelation degree that poultry eggs are required to achieve upon completion of the curing process, indicating that the protein has formed an ideal elastic gel.

[0132] Hp represents the number of periods corresponding to the prediction time domain, which is set according to requirements, for example, 48 periods = 8 hours. MPC performs rolling optimization within the time window of Hp; j is the period index in the prediction time domain, and the value of j is 0, 1, 2, ..., H. p -1; MPC performs rolling optimization within the future time window corresponding to this number of cycles;

[0133] v alk (j) represents the alkali replenishment flow rate in the j-th cycle of the prediction time domain;

[0134] T sp (j) represents the temperature setpoint for the j-th period in the prediction time domain;

[0135] T opt The preset reference pickling temperature is determined by process experience; the first objective function tends to maintain the temperature around this value to balance the reaction rate and energy consumption.

[0136] The first constraint condition, based on satisfying the coupled model (i.e., equations (1) to (3)), also includes:

[0137] C bulk,min ≤ C bulk (t) ≤ C bulk,max ;

[0138] T min ≤ T sp (j) ≤ T max ;

[0139] | T sp (j+1) - T sp (j) | ≤ ΔT rate ;

[0140] 0 ≤ v alk (j) ≤ v max ;

[0141] Among them, C bulk (t) represents the alkali concentration at time t, C bulk,min C is the preset minimum alkali concentration. bulk,max T is the preset maximum alkali concentration. min T is the preset minimum temperature. max T is the preset maximum temperature. sp (j+1) is the temperature setpoint for the (j+1)th period in the prediction time domain, ΔT rate v is the preset maximum temperature change. max This is the preset maximum alkali replenishment flow rate; all alkali replenishment flow rates are in L / min, and all temperatures are in °C.

[0142] According to equation (4), we know that: First, the first objective function J k Simultaneously considering pickling quality (first), pickling alkali consumption (second), and energy consumption (third), a unified optimization of multiple objectives was achieved, improving pickling quality while ensuring the economy of the pickling process; secondly, using t fin As a dynamic anchor point, the first model predictive controller can sense whether the current pickling progress meets expectations and adjust the control intensity accordingly to achieve full-process adaptive regulation; third, the first constraint condition corresponds to the actual boundary of the pickling process, ensuring that the optimal solution is always within the physically realizable range and can run autonomously without human intervention.

[0143] After the lower-level MPC controllers corresponding to each pickling device perform independent solutions, they output the corresponding alkali replenishment flow rate request value and temperature setting request value. First, they need to be summarized. Therefore, in some embodiments, the total heating power required by all pickling devices is determined based on the temperature setting request value of each pickling device, and the total alkali replenishment flow rate required by all pickling devices is determined based on the alkali replenishment flow rate request value of each pickling device.

[0144] Specifically, for the k-th pickling device, the corresponding temperature setting request value T is determined. sp,k The required heating power P to maintain this temperature in the pickling device can be estimated using a heat balance model. heat,k (Unit: kW) The total required heating power can be obtained by adding up the heating power requirements of all pickling devices: Similarly, the total alkali replenishment flow rate required for all pickling devices can be obtained as follows: N is the number of pickling devices, N≥2 and is an integer.

[0145] Based on this, in some embodiments, when it is detected that the total heating power required does not exceed the preset upper limit of the total heating power and the total alkali replenishment flow rate required does not exceed the preset upper limit of the total alkali replenishment flow rate, the requested alkali replenishment flow rate value is used as the target alkali replenishment flow rate and the requested temperature setting value is used as the target temperature setting value; when it is detected that the total heating power required exceeds the preset upper limit of the total heating power or the total alkali replenishment flow rate required exceeds the preset upper limit of the total alkali replenishment flow rate, the target alkali replenishment flow rate and target temperature setting value of each pickling device are obtained by rolling optimization solution through model predictive control method based on the total heating power required and the total alkali replenishment flow rate required.

[0146] The above solution is based on the pre-determined resources of the pickling system: two-way branch processing with a preset upper limit for total heating power and a preset upper limit for total alkali replenishment flow rate.

[0147] Path not exceeding limits: When both of the following conditions are met simultaneously: the total heating power demand does not exceed the preset total heating power limit and the total alkali replenishment flow rate does not exceed the preset total alkali replenishment flow rate limit, it indicates that the system resources are sufficient and the independent optimization results of each pickling device can be fully executed. The requested value can be directly sent as the target value without the need for intervention from the upper-level group control MPC, thus reducing computational overhead.

[0148] Over-limit path: When at least one of the two constraints is not met, it indicates that the current system resources are insufficient to simultaneously meet the requests of all pickling devices. The upper-level group control MPC is activated and intervened to reallocate the resource quotas of each pickling device. This ensures that the alkali replenishment and heating resources are allocated globally optimally under the total resource constraints, avoiding the failure of some pickling devices to execute control commands due to overload of a certain resource.

[0149] For the upper-level group control MPC, in some embodiments, the rolling optimization solution is performed based on the total demand heating power and the total demand alkali replenishment flow rate, including: inputting the total demand heating power and the total demand alkali replenishment flow rate into the second model predictive controller, and solving the problem based on the second constraint condition with the goal of minimizing the second objective function;

[0150] Second objective function J global for:

[0151] (5)

[0152] The second constraint includes the first constraint, and:

[0153] ;

[0154] ;

[0155] Among them, t fin,kThe predicted gelation degree of the k-th pickling device first reaches the target gelation degree G. target The estimated time, T ref To determine the ideal curing time for poultry eggs, the group control MPC, i.e., the second model predictive controller, forces the curing device to achieve the desired curing time through resource allocation. fin,k To T ref Convergence; ρ is J k The weighting factor is used to adjust the balance between "chickening group synchronicity" and "optimal performance of each pickling device"; P total The maximum preset heating power can be determined based on the transformer capacity or the total power of the heating system; F total The upper limit of the total flow rate for alkali replenishment can be determined by the flow rate or capacity of the centralized alkali supply pipeline.

[0156] In summary, the goal of the second model predictive controller (i.e., the upper-level group control MPC) is to minimize the dispersion of the expected output time of each tank under the resource constraints of the second constraint: after solving this quadratic programming problem, the target temperature setpoint and target alkali replenishment flow rate (i.e., alkali replenishment quota) of each pickling device can be obtained after global coordination, so that the pickling devices with faster progress can "wait" appropriately, and the pickling devices with slower progress can obtain more resources, ultimately achieving synchronous maturation of poultry eggs in the entire group of pickling devices.

[0157] To more intuitively illustrate the synergistic effect of two-level model predictive control, please refer to one embodiment. Figure 3 The provided control logic diagram for the two-layer MPC group control scheduling is illustrated for ease of description. The control terminal currently controls the pickling processes of the first and second pickling devices, each configured with a digital twin 1 and a digital twin 2. Digital twin 1 outputs state estimation data based on the pickling data sent by the first pickling device to the single-cylinder optimizer 1 of the lower-level single-cylinder MPC. The single-cylinder optimizer 1 performs rolling optimization to obtain the alkali replenishment flow rate request value v of the first pickling device. alk 1 and temperature setting request value T sp 1; Similarly, the single-cylinder optimizer 2 outputs the alkali replenishment flow rate request value v of the second pickling device. alk 2 and temperature setting request value T sp 2.

[0158] The upper-layer group control MPC aggregates resource requests and, based on v alk 1 and v alk 2. Obtain the total required alkali replenishment flow rate, based on T. sp 1 and T sp 2. If the total heating power required is obtained, and the resource request is not exceeded (i.e., the total alkali replenishment flow rate required does not exceed the preset total alkali replenishment flow rate limit, and the total heating power required does not exceed the preset total heating power limit), then the target alkali replenishment flow rate v is directly issued. alk 1'=v alk1. Target temperature setpoint T sp 1' = T sp 1, as the execution unit 1 for allocating resources to the first pickling device, issues the target alkali replenishment flow rate v. alk 2'= v alk 2. Target temperature setpoint T sp 2' = T sp 2 serves as the execution unit 2 for allocating resources to the second pickling device; if the resource request exceeds the limit, the global coordination optimizer (i.e., the upper-level group control MPC) is activated, and rolling optimization is performed based on equation (5) to calculate the target alkali replenishment flow rate v. alk 1' and target temperature setpoint T sp 1' is allocated as a resource quota to execution unit 1, and the target alkali replenishment flow rate v is calculated. alk 2' and target temperature setpoint T sp 2' is allocated as a resource quota to execution unit 2.

[0159] The dual-layer MPC optimization design in the above embodiments, by configuring an independent lower-layer MPC controller for each pickling device and performing rolling optimization solutions based on the state estimation of their respective digital twin models, can obtain alkali replenishment flow rate request values ​​and temperature setting request values ​​that better match the actual alkali concentration and gelation rate of each pickling device. This can better solve the quality fluctuation problem caused by the nonlinearity of "alkali-gel coupling". On this basis, the upper-layer group control MPC controller is configured to aggregate the alkali replenishment flow rate request values ​​and temperature setting request values ​​of each pickling device for overall control. Under the premise of meeting the overall resource constraints, the resource allocation of each pickling device is optimized, and the pickling progress of each pickling device is coordinated to keep it basically consistent. This effectively solves the problem of control failure and quality fluctuation caused by unreasonable resource allocation.

[0160] In some embodiments, after the pickling device performs pickling control according to the target alkali replenishment flow rate and target temperature setpoint, the control method further includes: after the pickling device completes pickling, acquiring the measured gelation distribution of the pickled poultry eggs; and adjusting the model parameters of the digital twin model according to the deviation between the measured gelation distribution and the predicted gelation distribution.

[0161] By sampling and testing the gel content of cured poultry eggs, for example using a texture analyzer and machine vision cutting inspection, the measured gel content distribution G of the cured poultry eggs can be obtained. true (r); By analyzing the measured gelation distribution G true (r) and the final predicted gel distribution output by the digital twin model The discrepancy between the two can be used to verify and correct the prediction accuracy of the digital twin model.

[0162] In some embodiments, the measured gelation distribution G is calculated.true (r) and predicted gelation distribution The root mean square error between the two is used as the objective function, and the key parameters of the digital twin model are adjusted using the Bayesian optimization algorithm until the root mean square error is minimized. Other non-critical model parameters can either be estimated by empirical formulas or are relatively stable and do not participate in the adjustment. The corrected model parameters can be associated with the batch characteristics of the current batch of pickled poultry eggs (such as place of origin, season and egg weight) and stored in the adaptive model library for the initialization of subsequent batches.

[0163] In summary, the control method for pickling poultry eggs applied to a control terminal provided in the first aspect embodiment has the following main features:

[0164] 1) A digital twin virtual model described by a set of partial differential equations was constructed for each pickling device; the model integrates the measured data of the sensors in real time and estimates the internal state data that cannot be directly measured, including alkali concentration, gelation distribution and gel front position, online through particle filtering algorithm.

[0165] 2) The first model predictive controller (lower-level MPC) for each pickling device continuously optimizes the alkali replenishment flow rate request value and temperature setting request value based on its own digital twin model; when it detects that the total heating power or total alkali replenishment flow rate of all pickling devices is limited, the second model predictive controller, as the upper-level group control MPC, automatically intervenes to optimize the target alkali replenishment flow rate and target temperature setting value of each pickling device, thereby coordinating the progress of each pickling device and making the expected pickling completion time of all pickling devices converge to the preset time window, so as to achieve consistent maturity of the entire batch of pickled products.

[0166] 3) After each batch of poultry eggs is cured, the model parameters are adjusted or corrected so that the digital twin model can more accurately depict the characteristics of raw materials (such as the differences in parameters of the digital twin model caused by different origins, seasons, egg weights, etc.). As batches accumulate, the model prediction error continues to decrease, the control precision continues to improve, forming a positive self-learning closed loop and improving the system's ability to quickly adapt to the curing of new batches of poultry eggs.

[0167] Secondly, based on the same inventive concept, please refer to some embodiments. Figure 4 A control terminal 400 for pickling poultry eggs is provided, comprising: a memory 410 for storing a computer program 411; and a processor 420 for executing the computer program 411 to implement the control method of any one of the first aspects of the embodiments.

[0168] The control terminal 400 can be a PC or a central server, used to receive pickling data sent by the pickling device. The control terminal 400 integrates the digital twin solver, upper-level MPC optimizer and lower-level MPC optimizer corresponding to the pickling device, calls the model parameters in the adaptive model library to perform calculations, and outputs the target alkali replenishment flow rate and target temperature setpoint to the pickling device to execute pickling control.

[0169] Thirdly, based on the same inventive concept, in some embodiments please refer to Figure 5 A control system for pickling poultry eggs is provided, including a control terminal 400 and at least one pickling device 100. The pickling device 100 is used to load poultry eggs; the pickling device 100 is used to acquire pickling data and send the pickling data to the control terminal 400; receive target alkali replenishment flow rate and target temperature setpoint sent by the control terminal 400, and perform pickling control according to the target alkali replenishment flow rate and target temperature setpoint; the control terminal 400 is used to receive pickling data; determine the corresponding state estimation data of the pickling device 100 over time according to the pickling data; determine the target alkali replenishment flow rate and target temperature setpoint of the corresponding pickling device 100 according to the state estimation data; and send the target alkali replenishment flow rate and target temperature setpoint to the corresponding pickling device 100.

[0170] The pickling data includes measured alkali concentration, measured temperature, and measured gel characteristic data within the pickling device 100. The measured gel characteristic data includes at least the first wave velocity and characteristic frequency band attenuation coefficient of the non-invasive transmission signal, which is obtained by exciting the pickling device 100. The state estimation data includes the predicted alkali concentration inside the albumen of each egg, the predicted gelation distribution in the radial space inside the albumen, and the predicted gel front position. The predicted gel front position is the radial position corresponding to the first time the predicted gelation is lower than a preset threshold in the direction from the inner surface of the eggshell to the yolk in the predicted gelation distribution.

[0171] In some embodiments, the pickling apparatus 100 includes a pickling container for holding poultry eggs and a sensor assembly disposed on the pickling container for acquiring pickling data; a first communication unit for transmitting pickling data and receiving a target alkali replenishment flow rate and a target temperature setpoint; and an execution unit for performing pickling control according to the target alkali replenishment flow rate and the target temperature setpoint.

[0172] In some embodiments, the sensor assembly includes a multimodal sensor array, with each pickling device 100 independently configured with:

[0173] Strong alkali resistant ion-selective electrode for real-time measurement of OH in pickling solutions. - concentration.

[0174] A thermometer used to measure the temperature T of the pickling solution.

[0175] An ultrasonic sensor is installed on the outer wall of the pickling container. The ultrasonic sensor includes a transmitting probe, a receiving probe, and a processing unit. The transmitting probe is located on one side of the outer wall, and the receiving probe is located on the other side. The transmitting probe is used to transmit an ultrasonic excitation signal, which excites the outer wall to form a guided wave. The guided wave propagates within the pickling container to generate an ultrasonic wave pattern, which includes a direct wave from the outer wall, multiple reflected waves, and a transmitted wave passing through the egg. The receiving probe is used to receive the ultrasonic wave pattern. The processing unit is used to remove the direct wave from the outer wall and the multiple reflected waves, and to select the transmitted wave within the first wave arrival time window for feature extraction to obtain measured gel feature data.

[0176] In some embodiments, the control terminal 400 includes: a second communication unit for receiving pickling data and sending a target alkali replenishment flow rate and a target temperature setpoint; and a processor for determining, based on the pickling data, the corresponding state estimation data of the pickling apparatus 100 over time; and determining the target alkali replenishment flow rate and the target temperature setpoint based on the state estimation data.

[0177] In some embodiments, the processor is configured to: input the pickling data as observations into a particle filter state estimator that uses a digital twin model as the state transition operator to obtain state estimation data.

[0178] In some embodiments, the digital twin model includes a mapping relationship between the effective diffusion coefficient of the base and the gelation degree of the protein, and a differential equation for the evolution of gelation degree; the mapping relationship corresponds to formula (2), and the differential equation corresponds to formula (3). Please refer to the relevant content in the first aspect embodiment, which will not be repeated here.

[0179] In some embodiments, the processor is configured to: employ a particle filtering algorithm to use a digital twin model as a state transition model and measured gel feature data as observations; use the observation model to map the state represented by each particle to predicted gel feature data; calculate the deviation between the predicted gel feature data and the measured gel feature data; and update the weight of each particle according to the deviation to correct the state estimate represented by each particle.

[0180] In some embodiments, the processor is configured to: determine the average gelation degree of the protein in each egg based on the predicted gelation degree distribution; determine the gel layer thickness of the protein in each egg based on the predicted gel front position and the size of the egg; input the average gelation degree and gel layer thickness of all eggs into the observation model, and output predicted gel feature data corresponding to the measured gel feature data.

[0181] In some embodiments, the processor is configured to: take state estimation data as input, perform rolling optimization solution using a model predictive control method to obtain the alkali replenishment flow rate request value and the temperature setting request value; and determine the target alkali replenishment flow rate and the target temperature setting value based on the alkali replenishment flow rate request value and the temperature setting request value.

[0182] In some embodiments, the processor is configured to: for the k-th pickling device, in the prediction time domain, using a first model prediction controller corresponding to the k-th pickling device, minimize a first objective function as the objective, and solve based on first constraints; the first objective function J k Please refer to the relevant content in the first aspect embodiment for the first constraint condition, which will not be repeated here.

[0183] In some embodiments, the processor is configured to: determine the total required heating power of all marinating devices 100 based on the temperature setting request value of each marinating device 100, and determine the total required alkali replenishment flow rate of all marinating devices 100 based on the alkali replenishment flow rate request value of each marinating device 100; when it is detected that the total required heating power does not exceed the preset total heating power limit and the total required alkali replenishment flow rate does not exceed the preset total alkali replenishment flow rate limit, the processor uses the alkali replenishment flow rate request value as the target alkali replenishment flow rate and the temperature setting request value as the target temperature setting value; when it is detected that the total required heating power exceeds the preset total heating power limit or the total required alkali replenishment flow rate exceeds the preset total alkali replenishment flow rate limit, the processor performs rolling optimization solution using a model predictive control method based on the total required heating power and the total required alkali replenishment flow rate to obtain the target alkali replenishment flow rate and the target temperature setting value for each marinating device 100.

[0184] In some embodiments, the processor is configured to: input the total demand heating power and the total demand alkali replenishment flow rate into the second model predictive controller, and solve for the second objective function based on the second constraints, with the objective of minimizing the second objective function J; global For the second constraint, please refer to the relevant content in the first aspect embodiment, which will not be repeated here.

[0185] After the pickling device 100 performs pickling control according to the target alkali replenishment flow rate and target temperature setpoint, in some embodiments, the processor is also used to: after the pickling device 100 completes pickling, acquire the measured gelation distribution of the pickled poultry eggs; and adjust the model parameters of the digital twin model according to the deviation between the measured gelation distribution and the predicted gelation distribution.

[0186] For details on the technical effects of the above solutions, please refer to the relevant content of the first aspect embodiment.

[0187] To illustrate the above solution more intuitively, a detailed explanation will be provided in the following embodiment, using a case of preserved egg pickling.

[0188] Please see Figure 6 The provided preserved egg curing system includes a curing unit group and a control terminal 400:

[0189] The pickling equipment complex includes:

[0190] 1) N parallel food-grade ceramic jars (11, 21, ..., N1) are used as pickling containers. Each ceramic jar has a volume of 100L~500L and contains fresh duck eggs and pickling liquid.

[0191] 2) Sensor groups (12, 22, ..., N2) employ a multimodal sensor array, independently configured at each ceramic jar:

[0192] Strong alkali resistant ion-selective electrode: range of 0~10 4 mol / m 3 NaOH, with an accuracy of ±0.5%, is used for real-time measurement of OH content in the pickling solution. - concentration.

[0193] PT1000 platinum resistance thermometer: Measurement accuracy ±0.1℃, used to determine the temperature T of pickling solution.

[0194] Non-invasive gel state sensing device: Each ceramic jar is equipped with a pair of air-coupled ultrasonic transducers with a center frequency of 100 kHz, respectively attached to both ends of the same horizontal axis on the outer wall of the ceramic jar; one serves as the transmitting probe, and the other as the receiving probe; please refer to Figure 7 The provided schematic diagram illustrates the working principle of the air-coupled ultrasonic transducer. The transmitting probe emits a narrow-pulse ultrasonic signal with a center frequency of 100kHz, which excites the left cylinder wall to generate a guided wave. The guided wave propagates along the cylinder wall, generating a direct wave that travels along the cylinder wall. This wave then propagates within the cylinder, forming multiple reflected waves. Some of the energy is transmitted into the liquid and through the egg cluster, forming a transmitted wave carrying gel state information. This transmitted wave couples out from the right cylinder wall and is collected by the receiving probe. The receiving probe also receives the direct wave and multiple reflected waves. The receiving probe converts these waveforms into electrical signals and transmits them to the data acquisition card. The data acquisition card uses time-domain gating technology to remove the direct wave and multiple reflected waves. Then, it selects the transmitted wave signal within the first wave arrival time window for feature extraction and pre-establishes an empty cylinder reference signal for background subtraction. Through FFT transformation of the transmitted wave electrical signal and bandpass filtering from 80 to 120kHz, an ultrasonic feature vector is obtained, including the first wave velocity v and the characteristic frequency band attenuation coefficients α1 and α2. This vector serves as the measured gel characteristic data, indirectly reflecting the average gelation degree and gel layer thickness within the egg.

[0195] 3) Distributed execution units (13, 23, ..., N3), one set for each ceramic jar, including:

[0196] Alkali metering pump: accurately adds high-concentration NaOH solution (20 wt%).

[0197] Water replenishment metering pump: Adds process water to dilute the solution.

[0198] Electric heating / cooling module: A flexible electric heating film and air-cooled pipes that wrap around the outer wall of the cylinder, with adjustable power.

[0199] Low-speed stirring device: paddle stirrer, 10~30 rpm, to maintain uniform concentration and temperature without damaging the eggshell.

[0200] 4) Cylinder-on-cylinder control unit (14,24,…,N4): An embedded controller based on ARM Cortex-M7 is configured at each ceramic cylinder, including an analog acquisition front end and an EtherCAT slave station, which is responsible for signal conditioning, acquisition, the underlying PID closed loop of the actuator and communication.

[0201] The control terminal 400 adopts a central intelligent computing and group control server, such as an industrial computer, which is internally deployed with a digital twin solver 401 (finite difference / finite element engine), a particle filter state estimator 402, a two-layer MPC optimizer 403 (including a lower-level single-cylinder MPC solver and an upper-level MPC group control solver) and an adaptive model parameter library 404. It should be noted that each ceramic vat is independently configured with a set of digital twin solvers 401, particle filter state estimators 402 and lower-level MPC controllers.

[0202] The core mechanism model of digital twin can be found in equations (1) to (3) and their boundary conditions in the first aspect embodiment. The digital twin model can be solved online using the Crank-Nicolson implicit difference method or the finite element method. The number of radial discrete points is 50 to 100 to meet the control requirements.

[0203] A flowchart illustrating the process control for pickling preserved eggs can be found here. Figure 8 The process includes steps S810 to S880, as detailed below:

[0204] S810: System initialization and digital twin instantiation.

[0205] The control terminal reads the statistical information of this batch of fresh duck eggs (such as average weight ±3%, average eggshell thickness 0.3mm~0.4mm, Hough units 60~80), and retrieves the set of model parameters that best matches the current season and origin from the adaptive model parameter library 404. For each ceramic jar k, create an independent numerical solution engine containing equations (1) to (3) and a digital twin solver instance M of the current estimated state. k (t0), and set a uniform target gelation degree G at the outlet. target=0.85 and ideal cylinder exit time window [T min , T max (e.g., T) min =25 days, T max =35 days); the execution unit of each tank prepares the initial pickling solution to an alkali concentration of C. bulk ,0 (e.g., 0.55×10 3 mol / m 3 Heat to the initial temperature T0 (e.g., 25°C).

[0206] S820: Synchronous acquisition and feature extraction of multi-source data.

[0207] At a fixed period Δt = 10 minutes, each cylinder control unit synchronously triggers the sensor group to collect marinating data, including:

[0208] 1) Alkali concentration C bulk (k,i); k represents the ceramic jar number, and i represents the sampling time number;

[0209] 2) Temperature T(k,i);

[0210] 3) The feature vector of the ultrasonic transmission signal waveform is extracted after fast Fourier transform and bandpass filtering. , where v(k,i) is the speed of the first wave of sound, and α1(k,i) and α2(k,i) are the attenuation coefficients at 80kHz and 120kHz, respectively.

[0211] The pickling data is transmitted to the control terminal 400 via the EtherCAT bus.

[0212] S830: Internal state estimation based on digital twin and particle filtering, including:

[0213] Digital twin: The server executes a particle filtering algorithm on each cylinder k and fuses the measured ultrasonic feature vector y. us (k,i) and the digital twin model M k Prediction, real-time estimation of the complete state vector .

[0214] State prediction: Based on the posterior state of the previous moment and the model (1)~(3), Δt is recursively extrapolated forward to obtain the prior particle set.

[0215] Weight update: The likelihood of each particle is calculated using the observation model h(x); the observation model employs a three-layer feedforward neural network pre-trained with experimental data, and the input data is the average gel strength. and gel layer thickness L gel =RR gel The output is the predicted ultrasound feature vector. Particle degradation is avoided through resampling.

[0216] S840: Solving the MPC of the lower single cylinder.

[0217] For each ceramic jar k, its digital twin M k and the current estimated state vector Starting from the prediction time domain Hp (48 cycles = 8 hours), solve the rolling optimization problem of equation (4) and output the first element of the optimal control sequence, namely the alkali replenishment flow rate request value v. alk and temperature setting request value T sp As the current control command.

[0218] S850: Determine if the total resource requests have exceeded the limit.

[0219] Assume the factory's preset maximum total heating power is P. total The preset upper limit of the total alkali replenishment flow rate is F. total ;P heat,k The heating power requested by the k-th ceramic jar at the current moment can be determined according to T. sp Based on the thermal balance model estimation; when the sum of the heating power requested by all single-cylinder MPCs exceeds P total And / or the sum of the alkali replenishment flow rate requests of all single-cylinder MPCs exceeds F total If the total resource request exceeds the limit, proceed to step S851; otherwise, if the total resource request does not exceed the limit, proceed to step S852.

[0220] S851: Determine the target alkali replenishment flow rate v alk '=v alk Target temperature setpoint T sp =T sp .

[0221] S852: Initiate upper-level group control MPC optimization solution to obtain the target alkali replenishment flow rate v. alk 'and target temperature setpoint T' sp '.

[0222] The goal of optimizing the upper-level MPC group control is to achieve P total and F total Under the constraint of minimizing the dispersion of the expected cylinder ejection time of each cylinder, v is obtained by solving equation (5). alk 'and T sp After solving this quadratic programming problem, the temperature settings and alkali replenishment quotas for each tank are obtained in a globally coordinated manner, allowing tanks with faster progress to "wait" appropriately, while those with slower progress receive more resources, and ultimately the entire group matures synchronously.

[0223] After obtaining v alk 'and T sp Then, it is sent as a control command to the cylinder-mounted actuator for execution.

[0224] S860: Determine if any ceramic jars meet the discharge criteria. The discharge criteria include: based on the predicted gelation distribution. The average gel strength obtained Has the target gelation degree (G) been achieved upon discharge from the vat? target =0.85 and whether the pickling time is within the ideal time window for removing from the jar [T] min , T max Within; in some embodiments, it may also include: predicting the gel front position R gel The gel content should not exceed 95% of the surface r0 of the egg yolk, and the difference between the maximum and minimum gel content within the gel layer should not exceed 0.15; if so, proceed to step S870, otherwise return to S820 to continue control.

[0225] S870: Egg removal detection and adaptive correction of model parameters.

[0226] Once a ceramic jar that meets the conditions for removal from the vat is opened, 3-5 eggs are sampled and cut and inspected using a texture analyzer and machine vision to obtain the radial distribution of the actual gelation degree (G). true (r), and compare it with the final distribution G predicted by the digital twin. pred (r) is compared, and the root mean square error between the two is calculated.

[0227] Using this error as the objective function, the Bayesian optimization algorithm is used to adjust the key parameters of the model corresponding to this batch. For other model parameters, D0 can be estimated using the empirical formula for eggshell thickness, and k c E a t ind It is relatively stable and does not need to be adjusted, which can improve optimization efficiency until the prediction error is minimized.

[0228] The optimized model parameters are associated with the characteristics of the eggs in this batch (including origin, season, egg weight, etc.) and stored in the adaptive model parameter library 404, which can be used for subsequent batches to achieve system self-evolution.

[0229] S880: Determine whether all earthenware jars have completed pickling, i.e., whether all earthenware jars meet the conditions for being removed from the jars; if yes, end the control; otherwise, return to S820 to continue.

[0230] Therefore, the technical framework of the control system for pickling poultry eggs provided in this application embodiment is as follows:

[0231] 1) Online sensing of gel characteristics: Low-frequency air-coupled ultrasonic guided wave technology is used to excite and receive ultrasonic transmitted waves on the outer wall of the ceramic jar in a non-contact manner. Then, the sound velocity and attenuation coefficient of the transmitted wave are analyzed to obtain the average gel state of multiple poultry eggs in the jar in a non-destructive and continuous manner.

[0232] 2) Digital Twin: A virtual prediction model described by a set of partial differential equations is established for each ceramic jar; the prediction model fuses sensor data in real time and estimates the internal gel distribution and gel front position that cannot be directly measured through a particle filtering algorithm.

[0233] 3) Dual-layer optimization control: The Model Predictive Controller (MPC) of each ceramic vat in the lower layer continuously optimizes the alkali replenishment flow rate and temperature setpoint based on its own digital twin; when public resources such as total heating power or total alkali replenishment flow rate are limited, the upper-layer group control MPC automatically intervenes to coordinate the progress of each vat, so that the expected discharge time of all ceramic vats converges to the preset time window, and the consistent maturity of the entire batch of products is achieved.

[0234] 4) Model Adaptive Evolution: After the eggs are removed from the vat, the deviation between the actual gelation degree measured by the texture analyzer and the gelation degree predicted by the digital twin is used to automatically correct the key parameters in the digital twin model, so that the control system can adapt to changes in raw material batches and seasons.

[0235] In summary, the control method and control system for poultry egg pickling provided in this application are the first to clearly define and quantify the core mechanism of "alkali-gel coupling" in the poultry egg pickling process. An adaptive control system of "nonlinear mechanism model + non-invasive online sensing + two-layer model predictive control" is constructed around it, and a digital twin capable of online evolution is established for each ceramic jar, fundamentally solving the three fundamental defects of poultry egg pickling mentioned above.

[0236] Fourthly, based on the same inventive concept, please refer to the following in some embodiments: Figure 9 A computer-readable storage medium 900 is provided, including a computer program 911, which can be executed by a processor to implement the control method of any of the first aspect embodiments.

[0237] In the above embodiments, the technical solutions can be implemented wholly or partially through software, hardware, firmware, or any combination thereof. Furthermore, as those skilled in the art will understand, the principles herein can be reflected in a computer program product on a computer-readable storage medium pre-loaded with computer-readable program code. Any tangible, non-transitory computer-readable storage medium may be used, including magnetic storage devices (hard disks, floppy disks, etc.), optical storage devices (CDs, DVDs, Blu-ray discs, etc.), flash memory, and / or the like. These computer program instructions can be loaded onto a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to form a machine, such that instructions executing on the computer or other programmable data processing apparatus can generate means for performing a specified function. These computer program instructions can also be stored in a computer-readable storage medium that can instruct the computer or other programmable data processing apparatus to operate in a particular manner, such that instructions stored in the computer-readable storage medium can form an article of manufacture, including means for implementing the specified function. The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to perform a series of operational steps on the computer or other programmable apparatus to produce a computer-implemented process, such that instructions executing on the computer or other programmable apparatus can provide steps for implementing the specified function.

[0238] This document describes various exemplary embodiments with reference to them. However, those skilled in the art will recognize that changes and modifications can be made to the exemplary embodiments without departing from the scope of this document. For example, various operational steps and components for performing operational steps can be implemented in different ways depending on the specific application or considering any number of cost functions associated with the operation of the system (e.g., one or more steps can be deleted, modified, or combined with other steps).

[0239] While the principles herein have been illustrated in various embodiments, numerous modifications to the structures, arrangements, proportions, elements, materials, and components, particularly suited to specific environments and operational requirements, may be used without departing from the principles and scope of this disclosure. These modifications and other alterations or alterations will be included within the scope of this document. Those skilled in the art will recognize that many changes can be made to the details of the above embodiments without departing from the fundamental principles of this application.

Claims

1. A method for controlling the pickling of poultry eggs, characterized in that, The control method, applied to a control terminal, includes: The system receives pickling data sent by a pickling device containing poultry eggs. The pickling data includes measured alkali concentration, measured temperature, and measured gel characteristic data within the pickling device. The measured gel characteristic data includes at least the first wave velocity and characteristic frequency band attenuation coefficient of a non-invasive transmission signal, which is obtained by exciting the pickling device. Based on the pickling data, state estimation data over time is determined within the pickling device. The state estimation data includes the predicted alkali concentration inside the albumen of each egg, the predicted gelation distribution in the radial space inside the albumen, and the predicted gel front position. The predicted gel front position is the radial position corresponding to the first time the predicted gelation is lower than a preset threshold in the direction from the inner surface of the eggshell to the yolk, according to the predicted gelation distribution. Based on the state estimation data, the target alkali replenishment flow rate and target temperature setpoint of the pickling device are determined; The target alkali replenishment flow rate and the target temperature setpoint are sent to the pickling device so that the pickling device can perform pickling control according to the target alkali replenishment flow rate and the target temperature setpoint.

2. The control method as described in claim 1, characterized in that, The step of determining the time-varying state estimation data within the pickling apparatus based on the pickling data includes: The pickling data is used as an observation and input into a particle filter state estimator that uses a digital twin model as the state transition operator to obtain the state estimation data.

3. The control method as described in claim 2, characterized in that, The digital twin model includes the mapping relationship between the effective diffusion coefficient of the base and the gelation degree of the protein, as well as the differential equation for the evolution of gelation degree; The mapping relationship satisfies: ; The differential equation is: ; Among them, D eff Where D0 is the effective diffusion coefficient of the base, G is the intrinsic diffusion coefficient of the base ions in the uncoagulated protein, β is the gelation degree of the protein, and β > 1 is the gelation barrier index; C is the base concentration in mol / m³. 3 t is time, k g (T) is the rate constant dependent on temperature T, m is the reaction order of the alkali concentration, and p is the gelation reaction order.

4. The control method as described in claim 2, characterized in that, The particle filter state estimator is used to: employ a particle filter algorithm, using the digital twin model as a state transition model, and the measured gel feature data as observations; The state represented by each particle is mapped to predicted gel feature data using an observation model; the deviation between the predicted gel feature data and the measured gel feature data is calculated, and the weight of each particle is updated according to the deviation to correct the state estimate represented by each particle.

5. The control method as described in claim 4, characterized in that, The process of mapping the state represented by each particle to predicted gel feature data using an observation model includes: Based on the predicted gelation distribution, the average gelation degree of the protein in each of the poultry eggs is determined; The thickness of the protein gel layer in each egg is determined based on the predicted gel front position and the size of the egg. The average gelation degree and gel layer thickness of each egg are input into the observation model, and the predicted gelation feature data corresponding to the measured gelation feature data are output.

6. The control method as described in claim 1, characterized in that, The step of determining the target alkali replenishment flow rate and target temperature setpoint of the pickling device based on the state estimation data includes: Using the state estimation data as input, the rolling optimization solution is performed through the model predictive control method to obtain the alkali replenishment flow rate request value and the temperature setting request value; The target alkali replenishment flow rate and the target temperature setting value are determined based on the requested alkali replenishment flow rate and the requested temperature setting value.

7. The control method as described in claim 6, characterized in that, The number of the pickling devices is multiple, and the rolling optimization solution using the model predictive control method includes: For the k-th pickling device, in the prediction time domain, the first model prediction controller corresponding to the k-th pickling device is used to minimize the first objective function as the objective, and the solution is performed based on the first constraint condition. The first objective function J k for: ； The first constraint includes: C bulk,min ≤ C bulk (t) ≤ C bulk,max ; T min ≤ T sp (j) ≤ T max ; | T sp (j+1) - T sp (j) | ≤ ΔT rate ; 0 ≤ v alk (j) ≤ v max ; Where ω1, ω2, and ω3 are weighting coefficients. For in t fin Average gelation degree at time t fin To predict the first time the gelation degree reaches the target gelation degree G target The estimated time, Hp is the number of periods corresponding to the predicted time domain, v alk (j) represents the alkali replenishment flow rate in the j-th period of the predicted time domain, T sp (j) is the temperature setpoint for the j-th period in the prediction time domain, T opt The preset reference marinating temperature, C bulk (t) represents the alkali concentration at time t, C bulk,min C is the preset minimum alkali concentration. bulk,max T is the preset maximum alkali concentration. min T is the preset minimum temperature. max T is the preset maximum temperature. sp (j+1) is the temperature setpoint for the (j+1)th period in the prediction time domain, ΔT rate v is the preset maximum temperature change. max The above values ​​represent the preset maximum alkali replenishment flow rate. All alkali replenishment flow rates are in L / min, and all temperatures are in °C.

8. The control method as described in claim 7, characterized in that, The step of determining the target alkali replenishment flow rate and the target temperature setting value based on the requested alkali replenishment flow rate and the requested temperature setting value includes: Based on the temperature setting request value of each of the pickling devices, determine the total heating power required for all the pickling devices, and based on the alkali replenishment flow rate request value of each of the pickling devices, determine the total alkali replenishment flow rate required for all the pickling devices. When it is detected that the total heating power required does not exceed the preset total heating power limit and the total alkali replenishment flow rate required does not exceed the preset total alkali replenishment flow rate limit, the alkali replenishment flow rate request value is taken as the target alkali replenishment flow rate and the temperature setting request value is taken as the target temperature setting value. When the total heating power required exceeds the preset upper limit of the total heating power, or the total alkali replenishment flow rate required exceeds the preset upper limit of the total alkali replenishment flow rate, the target alkali replenishment flow rate and target temperature setpoint for each pickling device are obtained by rolling optimization solution through model predictive control method based on the total heating power required and the total alkali replenishment flow rate required.

9. The control method as described in claim 8, characterized in that, The step of performing rolling optimization based on the required total heating power and the required total alkali replenishment flow rate using a model predictive control method includes: The total heating power and the total alkali replenishment flow rate are input into the second model prediction controller, and the solution is obtained based on the second constraint condition with the goal of minimizing the second objective function. The second objective function J global for: The second constraint includes the first constraint, and: ； ； Where N is the number of pickling devices, N≥2 and is an integer, t fin,k For the k-th pickling device, the predicted gelation degree first reaches the target gelation degree G. target The estimated time, T ref The ideal curing time for poultry eggs is given by ρ, where ρ is a weighting factor. The total heating power required is [the specified value]. To meet the required alkali replenishment total flow rate, P total F is the preset maximum total heating power. total The preset upper limit of the total flow rate for alkali replenishment.

10. The control method as described in claim 2, characterized in that, After the pickling device performs pickling control based on the target alkali replenishment flow rate and the target temperature setpoint, the control method further includes: After the pickling device completes the pickling process, the measured gel density distribution of the pickled poultry eggs is obtained. The model parameters of the digital twin model are adjusted based on the deviation between the measured gelation distribution and the predicted gelation distribution.

11. A control terminal for pickling poultry eggs, characterized in that, include: Memory, used to store computer programs; A processor for implementing the control method as described in any one of claims 1 to 10 by executing the computer program.

12. A control system for pickling poultry eggs, characterized in that, It includes a control terminal and at least one marinating device for loading poultry eggs; The pickling device is used to acquire pickling data and send the pickling data to the control terminal; receive the target alkali replenishment flow rate and target temperature set value sent by the control terminal, and perform pickling control according to the target alkali replenishment flow rate and target temperature set value; The control terminal is used to receive the pickling data; determine the state estimation data of the corresponding pickling device over time based on the pickling data; determine the target alkali replenishment flow rate and the target temperature setpoint of the corresponding pickling device based on the state estimation data; and send the target alkali replenishment flow rate and the target temperature setpoint to the corresponding pickling device. The pickling data includes measured alkali concentration, measured temperature, and measured gel characteristic data within the pickling device. The measured gel characteristic data includes at least the first wave velocity and characteristic frequency band attenuation coefficient of a non-invasive transmission signal, which is obtained by exciting the pickling device. The state estimation data includes the predicted alkali concentration within the albumen of each egg, the predicted gelation distribution in the radial space within the albumen, and the predicted gel front position. The predicted gel front position is the radial position corresponding to the first time the predicted gelation falls below a preset threshold in the direction from the inner surface of the eggshell towards the yolk, within the predicted gelation distribution.

13. The control system as described in claim 12, characterized in that, The pickling apparatus includes a pickling container for holding poultry eggs and a device disposed on the pickling container: Sensor components are used to acquire the pickling data; The first communication unit is used to send the pickling data and receive the target alkali replenishment flow rate and the target temperature setpoint. as well as, The execution unit is used to perform pickling control according to the target alkali replenishment flow rate and the target temperature setpoint; The control terminal includes: The second communication unit is used to receive the pickling data and send the target alkali replenishment flow rate and the target temperature set value. as well as, The processor is configured to determine, based on the pickling data, the corresponding state estimation data of the pickling device over time; and to determine, based on the state estimation data, the target alkali replenishment flow rate and the target temperature setpoint.

14. The control system as described in claim 13, characterized in that, The sensor assembly includes an ultrasonic sensor disposed on the outer wall of the pickling container; the ultrasonic sensor includes a transmitting probe, a receiving probe, and a processing unit, the transmitting probe being disposed on one side of the outer wall, and the receiving probe being disposed on the other side of the outer wall; The transmitting probe is used to transmit an ultrasonic excitation signal, which excites the outer wall to form a guided wave. The guided wave propagates within the pickling container to generate an ultrasonic wave, which includes a direct wave from the outer wall, multiple reflected waves, and a transmitted wave that passes through the egg. The receiving probe is used to receive the ultrasonic wave pattern; The processing unit is used to remove the direct wave from the outer wall and the multiple reflected waves, and to extract features from the transmitted wave within the first wave arrival time window to obtain the measured gel feature data.

15. A computer-readable storage medium, characterized in that, It includes a computer program that can be executed by a processor to implement the control method as described in any one of claims 1 to 10.

Images