A method and system for controlling the crystallization morphology of beef tallow based on ultrasonic assistance

Abstract

The application discloses a kind of based on ultrasonic auxiliary control tallow crystalline morphology regulation and control method and system, belong to food production control technical field, the regulation and control method includes: in tallow cooling crystallization tank upper deployment sensor cluster, in tallow crystallization process, real-time collection tallow crystallization state's process data, and upload to processing equipment;Processing equipment is preprocessed to the data collected by sensor cluster, and the data after pre-processing is aligned time stamp, and is integrated and is constructed into sensor historical database;Call processing equipment to analyze the data in sensor historical database, construct the key index dynamic model capable of reflecting the internal correlation between regulation and control parameters and crystalline morphology, and obtain optimization regulation and control algorithm based on the key index dynamic model design;Based on the model design optimization regulation and control algorithm, the technical problem that the single regulation and control means in traditional method cannot be considered simultaneously to solve the crystallization rate and beta' crystal form control.

Description

Technical Field

[0001] This invention belongs to the field of food production control technology, specifically, it relates to a method and system for controlling the crystallization morphology of butter based on ultrasound-assisted control. Background Technology

[0002] As one of the three essential nutrients for the human body, fat not only provides energy for bodily functions but is also crucial for maintaining healthy cell function. Furthermore, it serves as a carrier of fat-soluble trace elements and bioactive molecules, imparting excellent texture and sensory properties to food. However, in recent years, per capita fat consumption has been steadily increasing, leading to a rise in the incidence of chronic diseases such as obesity, diabetes, heart disease, and some cancers caused by excessive consumption of high-fat foods, placing a heavy burden on the healthcare system. Therefore, rationally controlling fat intake is of great significance in improving this situation.

[0003] The most common crystal forms of fat are β and β'. Different crystal forms have different effects on human digestion and absorption. β' crystals are small with a large specific surface area, making them easier to contact and break down with lipases; while the large particle structure of β crystals may reduce the efficiency of enzyme activity. During digestion, changes in the structure and digestibility of fat crystals significantly affect fat digestion and absorption, thus impacting human health.

[0004] Existing technology shows that introducing ultrasonic-assisted treatment into the oil crystallization process, including factors such as ultrasonic power, ultrasonic time, treatment temperature, and treatment volume, all affect the crystallization process. Introducing ultrasound can alter the crystal structure of oils, not only shortening the crystallization induction period but also significantly increasing the crystallization rate. Therefore, using ultrasonic-assisted crystallization technology in the cooling process of tallow can greatly accelerate oil crystallization and change the crystal shape. This demonstrates the broad application prospects of ultrasound in the field of assisted oil crystallization.

[0005] However, most current research focuses on verifying the macroscopic effects of ultrasound on crystallization behavior or exploring the effects of single process parameters such as ultrasonic power or time. In actual industrial production, the crystallization process of tallow is a complex, multi-factor coupled dynamic phase transition process. Ultrasonic effects, temperature changes, material flow, and other factors are intertwined and jointly determine the final crystal form distribution.

[0006] Current technologies lack a systematic approach that can comprehensively consider these factors and dynamically and precisely control them based on the real-time status of the crystallization process. Current control methods still heavily rely on offline analysis and the experience of skilled workers, failing to achieve online optimization of complex crystallization processes. This results in large batch-to-batch quality fluctuations, indicating room for improvement in the cooling crystallization time of tallow and the proportion of β' crystals in tallow.

[0007] Therefore, in order to solve the above problems, this application provides a method and system for controlling the crystal morphology of tallow based on ultrasound assistance. Summary of the Invention

[0008] To address the deficiencies in the aforementioned technical solutions, the first aspect of this invention provides a method for controlling the crystallization morphology of tallow based on ultrasound-assisted control, thereby solving problems such as the low proportion of β' crystal form and slow production rate in the existing tallow production process.

[0009] This invention is achieved through the following technical solution: a method for controlling the crystallization morphology of tallow based on ultrasound assistance, the method comprising the following steps:

[0010] S1. Deploy a sensor cluster on the tallow cooling crystallization tank to collect process data on the tallow crystallization state in real time during the tallow crystallization process and upload it to the processing equipment;

[0011] S2. The processing equipment preprocesses the data collected by the sensor cluster, maps the preprocessed data to a unified spatial coordinate system, and integrates it to build a sensor historical database.

[0012] S3. Call the processing device to analyze the data in the sensor's historical database, construct a dynamic model of key indicators that can reflect the intrinsic relationship between control parameters and crystal morphology, and design and optimize the control algorithm based on the model.

[0013] S4. Based on the process data transmitted back to the processing equipment in real time from the sensor cluster, and combined with the optimization and control algorithm, the predicted values ​​of crystallization rate and β' crystal form ratio are calculated. The predicted values ​​are compared with the real-time process data. Based on the comparison results, the parameters of the dynamic model of the key indicators are dynamically adjusted, and an optimized operation combination for adjusting the ultrasonic parameters and cooling system parameters is generated to control the final crystallization form of the butter.

[0014] Furthermore, the sensor cluster is constructed through the following sub-steps: S11, Select different types of sensors based on the main factors affecting the butter crystallization process and collect data; the data includes: butter temperature, ultrasonic power, butter viscosity, total crystallinity, and β' crystal form ratio;

[0015] S12. Use a near-infrared spectrometer to analyze spectral data and calculate the total crystallinity and the proportion of β' crystal form. Use a temperature sensor to collect the real-time temperature of the tallow. Use an ultrasonic power meter to monitor the ultrasonic power applied to the tallow. Use a rotational viscometer to monitor the effective viscosity of the tallow.

[0016] S13. Install the different types of sensors in key positions in the tallow cooling crystallization tank. The different types of sensors will transmit the collected data to the processing device in real time through the Internet of Things interface. The processing device will integrate the data of different types of sensors to achieve the synchronization of sensor data.

[0017] Furthermore, the preprocessing includes: denoising, completion, and standardization of the process data. Specifically, a window moving average method is used to perform moving average filtering on time-series data such as temperature, viscosity, and ultrasonic power. Wavelet decomposition is used to filter out background noise from the total crystallinity and β' crystal form ratio data calculated by the near-infrared spectrometer. Outlier detection is used to identify and remove data that exceeds the mean ± 2 standard deviations. After denoising, the data is repaired and completed using linear interpolation to repair data loss caused by transient sensor malfunctions or signal transmission loss. Finally, Z-Score standardization is used to normalize the data.

[0018] The key indicator dynamic model consists of two coupled equations for the influencing factors of the tallow crystallization process, including: a dynamic model for the crystallization rate describing the change in the crystallization rate, as shown in the following equation:

[0019]

[0020] in, Total crystallinity; Supercooling; This is the real-time temperature of the butter; Real-time ultrasonic power; This refers to the effective viscosity of butter. The thermodynamic driving efficiency coefficient. This is the ultrasonic energy input efficiency coefficient.

[0021] The dynamic model of crystal morphology used to describe the change in the proportion of β' crystal form is shown in the following equation:

[0022]

[0023] in, The proportion of the β' crystal form in the total crystal; This is the optimal growth temperature for the β' crystal form; Runtime; The intensity coefficient of the temperature selectivity effect. This is the width factor of the optimal temperature window.

[0024] Furthermore, the optimization control algorithm includes an objective function and constraints, wherein the objective function is:

[0025]

[0026] in, Ultrasonic power setting value The setpoint for the jacket temperature of the tallow cooling crystallizer is a control variable to be optimized. The predicted crystallization rate; For prediction Crystal form ratio; , , The weighting coefficients and constraints are set based on actual production conditions and include: control variable constraints, which characterize the physical limits of various control parameters of the ultrasonic generator and cooling system; and process constraints, which characterize the performance-related limitations of the tallow crystallization process.

[0027] Furthermore, step S4 also includes a closed-loop control step, in which the sensor cluster continuously collects new data, inputs the latest data into the error function, obtains updated parameters through gradient descent, and uses the updated parameters to dynamically calibrate the dynamic model of key indicators.

[0028] Furthermore, the error function is shown in the following equation:

[0029]

[0030] in, It is the vector of model parameters to be calibrated. ;

[0031] These are the model's predicted values; This is the actual measured value obtained by differential calculation of real-time sensor data; These are the weighting coefficients used to balance the two error terms;

[0032] Adjust parameters iteratively The error is gradually reduced, and its parameter update mechanism is expressed as follows:

[0033]

[0034] in, The parameter value after the (k+1)th iteration; This represents the parameter value for the current k-th iteration; The learning rate is used to control the step size of each adjustment. This is the gradient of the error function at the current parameters.

[0035] The second aspect of the present invention provides a control system for controlling the crystallization morphology of tallow based on ultrasound assistance, the system comprising a tallow cooling crystallization tank, an ultrasonic generator, a cooling system, and processing equipment;

[0036] The tallow cooling crystallization tank is equipped with a sensor cluster module; the cooling system includes a heating / cooling jacket for temperature control and a low-speed stirring device; the processing equipment is used to process the data collected by the sensor cluster module and adjust the parameters of the ultrasonic generator and the cooling system according to the data processing results.

[0037] The ultrasonic frequency emitted by the ultrasonic generator is between 20-100kHz.

[0038] The low-speed stirring device can stir at a speed of 20-50 rpm to ensure a uniform distribution of energy and temperature fields in the tallow slurry system;

[0039] The control system also includes a data preprocessing module, a data analysis module, and a control module.

[0040] The data preprocessing module is configured to preprocess the data collected by the sensor cluster module, including performing noise reduction, completion and standardization operations.

[0041] The control system also includes a data preprocessing module, a data analysis module, and an optimization control module. The data preprocessing module is configured to preprocess the data collected by the sensor cluster module, align the preprocessed data with timestamps, integrate and construct a sensor historical database.

[0042] The data analysis module is connected to the data preprocessing module and is configured to analyze the data in the sensor historical database, construct a dynamic model of key indicators, and design an optimized control algorithm based on the dynamic model of key indicators.

[0043] The optimization and control module is connected to the data analysis module and the sensor cluster module. It is configured to calculate the predicted value based on the real-time data transmitted back from the sensor cluster module and the optimization and control algorithm, compare the predicted value with the real-time sensor data, and dynamically adjust the parameters of the dynamic model of key indicators based on the comparison results.

[0044] Furthermore, the sensor cluster module is configured to install different types of sensors at key locations in the tallow cooling crystallization tank. The data collected by the different types of sensors is transmitted to the processing device in real time via an IoT interface. The processing device integrates the data from different types of sensors to achieve synchronization of sensor data. A temperature sensor is used to collect the temperature of the tallow slurry, an ultrasonic power meter is used to monitor the real-time ultrasonic output power, a viscometer is used to measure the effective viscosity, and a near-infrared spectrometer is used to monitor the total crystallinity and β' crystal form ratio of the tallow.

[0045] The beneficial effects of this invention are as follows:

[0046] 1. This invention constructs a data-driven dynamic model of key indicators by deploying a sensor cluster, and designs a closed-loop optimization control algorithm based on this model. By analyzing multi-source sensor data and calibrating model parameters online, it can calculate the optimal control command that can achieve the desired increase in crystallization rate and β' crystal form ratio. This intelligent control method transforms the complex, multi-variable coupled crystallization process into a predictable and optimizable control, thereby avoiding the shortcomings of existing technologies, such as blind control and large batch-to-batch quality fluctuations, caused by the inability to handle the coupling relationship of multiple factors.

[0047] 2. This invention provides a complete system solution for quantitative, automated, and optimized control of complex phase transition processes. Based on this model, an optimized control algorithm is designed to generate tallow products with a higher proportion of healthy β' crystal form in a shorter crystallization time. This solves the technical problem of controlling both crystallization rate and β' crystal form in traditional methods due to the single control method. Attached Figure Description

[0048] Figure 1 This is a flowchart of the control method of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. The illustrative embodiments and descriptions of this invention are for explanation only and are not intended to limit the invention. Furthermore, regarding numerical ranges in this invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Any stated value or intermediate value within a stated range, as well as each smaller range between any other stated value or intermediate value within said range, is also included in this invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0050] Unless otherwise stated, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. While only preferred methods and materials have been described herein, any methods and materials similar or equivalent to those described herein may be used in the implementation or testing of this invention. All references to this specification are incorporated by way of citation to disclose and describe methods and / or materials associated with those references. In the event of any conflict with any incorporated reference, the content of this specification shall prevail.

[0051] Various modifications and variations can be made to the specific embodiments described in this specification without departing from the scope or spirit of the invention, as will be apparent to those skilled in the art. Other embodiments derived from this specification will also be apparent to those skilled in the art. This specification and embodiments are merely exemplary.

[0052] The terms “include,” “including,” “have,” “contain,” etc., used in this article are all open-ended terms, meaning that they include but are not limited to.

[0053] In the following examples, "parts" refers to parts by weight.

[0054] Example 1

[0055] This embodiment provides a method for controlling the crystallization morphology of tallow based on ultrasound assistance. The tallow crystallization process in this embodiment aims to induce the formation of a semi-solid product with a predominantly β' crystal form from molten tallow, whose main component is triglycerides, through precise control of physicochemical conditions. This process is carried out in a tallow cooling crystallization tank equipped with a jacketed temperature control and stirring function. First, refined tallow is heated to complete melting to eliminate all crystal memory. Then, it undergoes cooling and physical treatment according to a specific procedure, ultimately obtaining a tallow product with a higher proportion of β' crystals in the shortest possible cooling time.

[0056] In this embodiment, by considering the characteristics of ultrasonic assistance and gradient temperature control used in the butter crystallization process, the key parameters that need to be monitored are determined. A corresponding sensor cluster is installed on the butter cooling crystallization tank to collect sensor data. An optimized control algorithm is constructed based on the data collected during production. Then, in the subsequent production process, the butter crystallization production process is controlled based on the real-time data collected by the sensors and the constructed optimized control algorithm, thereby achieving process optimization that significantly increases the proportion of β' crystal form in the butter product. Figure 1 The flowchart of the control method in this embodiment is shown. As can be seen from the figure, this embodiment includes the following steps:

[0057] S1. Deploy a sensor cluster on the tallow cooling crystallization tank to collect process data on the tallow crystallization state in real time during the tallow crystallization process and upload it to the processing equipment;

[0058] Specifically, the deployed sensor cluster includes:

[0059] 1) Analyze the key data in the butter crystallization process to determine the key parameters that need to be monitored. In the butter crystallization process, the parameters that affect the crystallization rate and the proportion of β' crystal form mainly include: butter temperature, ultrasonic power, butter viscosity, total crystallinity of butter and the proportion of β' crystal form.

[0060] 2) Based on the main factors affecting the crystallization process of tallow mentioned above, select appropriate sensors; use a near-infrared spectrometer to analyze the spectral data by utilizing its sensitivity to the vibrational absorption of molecular functional groups, in order to calculate the total crystallinity and the proportion of β' crystal form; use a temperature sensor to collect the real-time temperature of the tallow; use an ultrasonic power meter to monitor the magnitude of the ultrasonic power applied to the tallow; and use a rotational viscometer to monitor the effective viscosity of the tallow.

[0061] 3) After selecting the appropriate sensor, install it in a key position in the tallow cooling crystallization tank, so that the temperature probe and the near-infrared spectrometer probe are placed inside the tallow, and the rotor of the viscometer is also immersed in the tallow, to ensure that the measured data are representative.

[0062] 4) The data collected by the sensor is transmitted to the processing device in real time through the Internet of Things (IoT) interface to achieve synchronous acquisition of the source data.

[0063] In step S2, the processing device preprocesses the data collected by the sensor cluster, aligns the preprocessed data with timestamps, maps it to a unified spatial coordinate system, and integrates it to build a sensor historical database.

[0064] Specifically, the data preprocessing operations include denoising, completion, and standardization of the process data;

[0065] First, a windowed moving average method is used to smooth short-term fluctuations by applying a moving average filter to the time-series data, processing temperature, viscosity, and ultrasonic power data. Wavelet decomposition is then used to filter out background noise, such as the total crystallinity and β' crystal form percentage calculated from the near-infrared spectrometer, while preserving the abrupt change characteristics of data mutations. Outlier detection and removal are then implemented; data points exceeding the mean ± 2 standard deviations are considered outliers and removed. Next, linear interpolation is used to repair and complete the denoised data, correcting data loss caused by transient sensor malfunctions or signal transmission loss. Finally, Z-score normalization is applied to normalize process data of different dimensions, making the sensor data usable for subsequent model training and optimization.

[0066] Step S3: Call the processing device to analyze the data in the sensor historical database. By analyzing the data in the historical database, the key indicators that have the greatest impact on the butter crystallization process are obtained, namely: crystallization rate and β' crystal form ratio. Based on the coupling relationship between these two key indicators, a dynamic model is established, and an optimized control algorithm is designed based on the constructed key indicator dynamic model.

[0067] Specifically, in the second step of the sensor cluster deployed in step S1, the near-infrared spectrometer is used to analyze spectral data by utilizing its sensitivity to the vibrational absorption of molecular functional groups in order to calculate the total crystallinity and the proportion of β' crystal form.

[0068] The physical principle is that although different crystal forms of fat (β' and β) have the same chemical composition, their molecules are arranged in different ways and have different spatial structures in the crystal lattice. This difference in microstructure leads to different modes of molecular vibration. When near-infrared light shines on butter, different crystal forms will produce characteristic absorption of light at their specific wavelengths. Therefore, by analyzing the spectral data of a near-infrared spectrometer, data including the proportion of crystal forms and the total crystallinity can be obtained.

[0069] Specifically, the data analysis of near-infrared spectrometers is based on chemometrics. Through multivariate correction methods such as partial least squares or principal component regression, a mathematical model is established between the spectral data and the β' crystal form ratio and total crystallinity measured by laboratory standard methods. By substituting the collected spectral data into this model, the crystal form ratio and total crystallinity can be quickly calculated.

[0070] Among them, the crystallization rate is a dynamic value that represents how fast the crystallinity changes with time. It cannot be directly measured. The crystallization rate is the derivative of the total crystallinity with respect to time, which can be obtained by differential calculation.

[0071] Specifically, to precisely control the butter crystallization process and produce butter products with high β' crystal content more quickly, it is necessary to express the dynamic changes of various key physical quantities affecting the crystallization rate and the selection of the β' crystal content using mathematical equations. The core of the key indicator dynamic model is to analyze the data collected during the butter crystallization process and, in conjunction with the physicochemical principles of the butter crystallization process, construct a set of coupled equations that can reflect the actual butter production process. This set of coupled equations may specifically include the following:

[0072] 1) The crystallization process of tallow is influenced by a combination of factors, including temperature, viscosity, and ultrasonic power, and its final crystallization rate is the result of the combined effect of these factors. Furthermore, the proportion of β' crystal form in the final tallow product is also regulated by the current temperature and the early nucleation environment. The crystallization rate can be described by the following equation:

[0073]

[0074] in, The total crystallinity is dimensionless and is measured in real time by a near-infrared spectrometer. For supercooling, K, By analyzing the initial crystallization temperature of tallow Real-time temperature of butter The difference was calculated; the temperature of the butter was measured by a temperature sensor; The real-time ultrasonic power, W, is monitored by the power meter of the ultrasonic generator; The effective viscosity of the tallow is Pa·s, measured by a rotational viscometer. The thermodynamic driving efficiency coefficient. This is the ultrasonic energy input efficiency coefficient.

[0075] It should be noted that in this equation The driving term is used to characterize the crystallization driving force provided by thermodynamics, namely temperature and ultrasound; where The thermodynamic driving efficiency coefficient characterizes how fast the crystallization rate can be provided for each degree of temperature reduction. The ultrasonic energy input efficiency coefficient characterizes how fast the crystallization rate can be provided for every 1W of ultrasonic power. As a fluidity factor, the crystallization rate is directly proportional to the molecular mobility, which in turn is directly proportional to the reciprocal of the system viscosity. The higher the viscosity, the greater the resistance. The available liquid phase fraction is used to characterize the consumption of unsolidified tallow liquid phase during the tallow crystallization process.

[0076] The description of crystallization rate, based on the rate of any phase transition process, depends on how strong the tendency to want the phase transition is, i.e., the driving force, and the influence of whether molecules can move to the designated location, i.e., the mobility, as well as whether there is something else available for the phase transition.

[0077] Among them, the driving factors include thermodynamic driving and ultrasonic driving, the migration rate is affected by the viscosity of the system, and the crystallization process is the process of liquid phase to solid phase. When most of the system is already solid, the available liquid phase fraction will decrease.

[0078] 2) During the crystallization process of butter, the β' crystal form competes with other crystal forms for growth; its final proportion is determined by whether the current temperature is within its optimal growth window and the nucleation environment provided by ultrasound during crystallization; the proportion of the β' crystal form can be described by the following equation:

[0079]

[0080] in, The proportion of the β' crystal form in the total crystal is dimensionless and measured by a near-infrared spectrometer. The optimal growth temperature for the β' crystal form is K; The runtime is in seconds.

[0081] It should be noted that in this equation The Gaussian function chosen for the temperature is based on the real-time temperature of the butter. Exactly equal to Optimal growth temperature of crystal form When the value is at its maximum, it will decrease rapidly if it deviates from the specified value; thus, it can accurately reflect the control principle of gradient temperature control. The term representing the attenuation of the historical effect of ultrasound refers to the nucleation environment provided by ultrasound, the influence of which decreases as the crystallization process progresses; the denominator represents the attenuation of the historical effect of ultrasound. To describe this decay, the longer the time t, or the higher the total crystallinity... The higher the value, the larger the denominator, and the smaller the influence of this term. Among them, The intensity coefficient of the temperature selectivity effect. The width coefficient is the optimal temperature window width factor; both coefficients together characterize and describe the... When the temperature is near the optimal growth temperature of the crystal form, the effect of temperature on The intensity and extent of influence of crystal growth; is the intensity coefficient of the ultrasonic seeding effect, which can characterize the intensity of the influence of the ultrasonic nucleation environment; This is the logistic growth term, used to characterize the competitive growth relationship between different crystal forms.

[0082] The description of the proportion of β' crystal form is based on the logistic growth model, which does not directly describe the absolute amount of β' crystal form, but rather the rate of change of its proportion in the total crystal. This rate of change depends on the growth advantage of β' crystal form relative to β crystal form and the competition for living space between them.

[0083] 3) In the two models above, there are multiple parameters that need to be determined. These parameters reflect the inherent physical properties of specific raw materials and equipment, and can be updated in real time through an online calibration mechanism to ensure that the model accurately reflects actual operating conditions.

[0084] By coupling the two dynamic models mentioned above, a dynamic model is formed that can quantify the key indicators of crystallization rate and β' crystal form ratio. By combining real-time sensor data, these equations can be solved in real time, thereby providing accurate predictions of future crystallization states for subsequent optimization and control algorithms.

[0085] In this embodiment, the purpose of the optimization and control algorithm is to achieve the overall optimal production goal of maximizing the crystallization rate and increasing the proportion of β' crystal form by dynamically adjusting the control parameters while meeting the constraints of actual tallow production. The implementation process of the optimization and control algorithm is highly dependent on the predictive ability of the dynamic model of key indicators, because the optimization and control algorithm needs to dynamically predict the future state based on the real-time solution of the model, and then make the best decision in advance.

[0086] Specifically, the construction of the optimization and control algorithm includes: 1) From the perspective of improving production efficiency and product quality, in order to achieve faster crystallization speed and higher β' crystal form ratio, a comprehensive objective function is constructed.

[0087] In this embodiment, the objective function can be expressed by the following formula:

[0088]

[0089] in, For comprehensive performance indicators, Ultrasonic power setting value The setpoint for the jacket temperature of the tallow cooling crystallizer is a control variable to be optimized. The predicted crystallization rate is obtained by substituting the current sensor measurements and the control variables to be optimized into the dynamic model of the crystallization rate. For prediction Crystal form ratio, by substituting the current sensor measurement value and the control variable to be optimized into The dynamic model of crystal form ratio was obtained; , , These are weighting coefficients that can be adjusted according to production needs, used to balance the relationship between production efficiency, product quality, and production costs.

[0090] Constraints are set for the objective function based on actual production conditions; these constraints ensure that the control commands given by the algorithm do not exceed the safe operating range of the equipment, guaranteeing the stability and safety of the production process. The above parameters are subject to the following constraints: , , ;

[0091] In each control cycle, the optimized control algorithm obtains a set of optimal control commands by solving the aforementioned constrained minimization problem. It is then sent to the corresponding controllers for optimization.

[0092] S4. Based on the process data transmitted back to the processing equipment in real time by the sensor cluster during the butter crystallization process, and combined with the optimization and control algorithm in step 3, the predicted values ​​of crystallization rate and β' crystal form ratio are calculated. The predicted values ​​are compared with the real-time data of the sensors. Based on the comparison results, the parameters of the dynamic model of key indicators are dynamically adjusted. Based on the adjusted model parameters, an optimized operation combination for adjusting the ultrasonic parameters and cooling system parameters is generated to control the final crystallization form of the butter.

[0093] Specifically, step S4 also includes a closed-loop control step, in which the sensor cluster continuously collects new data, inputs the latest data into the error function, obtains updated parameters through gradient descent, and uses the updated parameters to dynamically calibrate the dynamic model of key indicators.

[0094] In the actual implementation process, since the dynamic model of key indicators is an idealized model, factors such as batch differences in raw materials, fluctuations in ambient temperature, and equipment aging may cause the model parameters to deviate during actual production, thus affecting the accuracy of prediction. Therefore, it is necessary to establish an online calibration mechanism to dynamically correct the model parameters based on real-time sensor data.

[0095] In this embodiment, an error function is constructed to dynamically calibrate the model parameters based on real-time sensor data; this error function is achieved by setting a set of model parameters. This ensures that the predicted values ​​of the dynamic model for key indicators are as close as possible to the sensor measurements.

[0096] In this embodiment, the error function is as follows:

[0097]

[0098] in, It is the vector of model parameters to be calibrated. ; The total crystallinity was obtained by differential calculation from the time-series data of the total crystallinity measured by near-infrared spectroscopy; the specific calculation is as follows:

[0099]

[0100] Measured by a near-infrared spectrometer Obtained by differencing time series data; The predicted value of the model can be calculated by substituting the sensor measurements and control inputs from the previous moment into the formulas of the crystallization rate model and the β' crystal form ratio model. This is the actual measured value obtained by differential calculation of real-time sensor data; The weighting coefficients are used to balance the two error terms; the model parameters are adjusted iteratively. Gradually reduce the error Its parameter update mechanism can be expressed by the following formula:

[0101]

[0102] in, The parameter value after the (k+1)th iteration; This represents the parameter value for the current k-th iteration; The learning rate is used to control the step size of each adjustment; The gradient of the error function with respect to the parameter vector can be obtained by taking the partial derivative of the error function formula.

[0103] Meanwhile, the sensor cluster continues to collect new data, and if it finds that the actual crystallization state is still not up to standard, it immediately initiates a new round of optimization and calibration. The sensor cluster collects data every [period]. Data is collected periodically, and the latest data is input into the error function. Updated parameters are then obtained using gradient descent. The new parameters are then... By substituting the key indicator into the dynamic model, more accurate predictions are generated. Based on the updated predictions, the algorithm is optimized to calculate new control commands and complete closed-loop control.

[0104] Example 2

[0105] This embodiment discloses a control system based on ultrasound-assisted control of the crystallization morphology of tallow; the system includes a tallow cooling crystallization tank, an ultrasonic generator, a cooling system, and processing equipment;

[0106] The tallow cooling crystallization tank is equipped with a sensor cluster module; the cooling system includes a heating / cooling jacket for temperature control and a low-speed stirring device; the processing equipment is used to process the data collected by the sensor cluster module and adjust the parameters of the ultrasonic generator and the cooling system according to the data processing results.

[0107] The ultrasonic frequency emitted by the ultrasonic generator is between 20-100kHz.

[0108] The low-speed stirring device can stir at a speed of 20-50 rpm to ensure a uniform distribution of energy and temperature fields in the tallow slurry system;

[0109] The control system also includes a data preprocessing module, a data analysis module, and a control module.

[0110] The data preprocessing module is configured to preprocess the data collected by the sensor cluster module, including performing noise reduction, completion and standardization operations.

[0111] The data preprocessing module is connected to the data analysis module. The data preprocessing module is configured to analyze the data in the sensor historical database, construct a dynamic model of key indicators, and design an optimized control algorithm based on the dynamic model of key indicators.

[0112] The optimization and control module is connected to the data analysis module and the sensor cluster module. It is configured to calculate the predicted value based on the real-time data transmitted back from the sensor cluster module and the optimization and control algorithm, compare the predicted value with the real-time sensor data, and dynamically adjust the parameters of the dynamic model of key indicators based on the comparison results according to the online calibration mechanism.

[0113] Furthermore, the sensor cluster module is configured to install different types of sensors at corresponding positions in the tallow cooling crystallization tank. The different types of sensors will transmit the collected data to the processing device in real time through the Internet of Things interface. The processing device will integrate the data of different types of sensors to achieve sensor data synchronization and send the synchronized data to the subsequent processing module.

[0114] Specifically, a temperature sensor was used to collect the temperature of the tallow slurry, an ultrasonic power meter was used to monitor the real-time ultrasonic output power, a viscometer was used to measure the effective viscosity, and a near-infrared spectrometer was used to monitor the total crystallinity and the proportion of β' crystal form of the tallow.

[0115] Specifically, the optimization and control module also includes closed-loop control. Based on the new data continuously collected by the sensor cluster, the latest data is input into the error function, and updated parameters are obtained through gradient descent. The updated parameters are then used to dynamically calibrate the dynamic model of key indicators.

[0116] The error function is shown in the following formula:

[0117]

[0118] in, For model parameters, including ;

[0119] Adjust parameters iteratively The error is gradually reduced, and its parameter update mechanism is expressed as follows:

[0120]

[0121] in, The parameter value after the (k+1)th iteration; This represents the parameter value for the current k-th iteration; These are control parameters used to control the step size for each adjustment; This is the gradient of the error function at the current parameters.

[0122] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for controlling the crystallization morphology of tallow based on ultrasound-assisted control, characterized in that, The control method includes the following steps: S1. Deploy a sensor cluster on the tallow cooling crystallization tank to collect process data on the tallow crystallization state in real time during the tallow crystallization process and upload it to the processing equipment; S2. The processing device preprocesses the data collected by the sensor cluster, aligns the preprocessed data with timestamps, and integrates them to build a sensor historical database. S3. The processing device is invoked to analyze the data in the historical database of the sensor, and a dynamic model of key indicators that can reflect the intrinsic relationship between the control parameters and the crystal morphology is constructed. Based on the dynamic model of key indicators, an optimized control algorithm is designed. S4. Based on the process data transmitted back in real time by the sensor cluster, and combined with the optimized control algorithm, the predicted values ​​of crystallization rate and β' crystal form ratio are calculated. The predicted values ​​are compared with real-time sensor data. Based on the comparison results, the parameters of the dynamic model of key indicators are dynamically adjusted. Based on the adjusted model parameters, an optimized combination of operations is generated to adjust the ultrasonic parameters and cooling system parameters in order to control the final crystallization form of butter. The dynamic model for key indicators consists of two coupled equations relating to factors influencing the tallow crystallization process, including: The dynamic model for describing the change in crystallization rate is shown in the following equation: in, Total crystallinity; Supercooling; This is the real-time temperature of the butter; Real-time ultrasonic power; This refers to the effective viscosity of butter. The thermodynamic driving efficiency coefficient. The ultrasonic energy input efficiency coefficient; The equation used to describe the proportion of β' crystal form during the crystallization process of tallow is shown below: in, The proportion of the β' crystal form in the total crystal; Runtime; This is the optimal growth temperature for the β' crystal form; This represents the intensity coefficient of the temperature selectivity effect; This is the width coefficient of the optimal temperature window; The intensity coefficient of the ultrasonic seeding effect; The objective function in the optimization and control algorithm is shown in the following equation: in, Ultrasonic power setting value The setpoint for the jacket temperature of the tallow cooling crystallizer is a control variable to be optimized. The predicted crystallization rate; For prediction Crystal form ratio; , , These are the weighting coefficients; Step S4 also includes a closed-loop control step, in which the sensor cluster continuously collects new data, inputs the latest data into the error function, obtains updated parameters through gradient descent, and uses the updated parameters to dynamically calibrate the dynamic model of key indicators. The error function is shown in the following equation: in, It is the vector of model parameters to be calibrated. ; These are the model's predicted values; This is the actual measured value obtained by differential calculation of real-time sensor data; The weighting coefficients are used to balance the two error terms; the parameters are adjusted iteratively. The error is gradually reduced, and its parameter update mechanism is expressed as follows: in, The parameter value after the (k+1)th iteration; This represents the parameter value for the current k-th iteration; The learning rate is used to control the step size of each adjustment. This is the gradient of the error function at the current parameters.

2. The method for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 1, characterized in that, The sensor cluster is constructed through the following sub-steps: S11. Based on the main factors affecting the butter crystallization process, select different types of sensors and collect data; The data includes: butter temperature, ultrasonic power, butter viscosity, total crystallinity, and the proportion of β' crystal form; S12. Use a temperature sensor to collect the real-time temperature of the butter, use an ultrasonic power meter to monitor the ultrasonic power applied to the butter, use a rotational viscometer to monitor the effective viscosity of the butter, and use a near-infrared spectrometer to analyze the spectral data to calculate the total crystallinity and the proportion of β' crystal form. S13. Install the different types of sensors at key locations in the tallow cooling crystallization tank. The different types of sensors will transmit the collected data to the processing device in real time through the Internet of Things interface. The data from the different types of sensors will be integrated at the processing device to achieve synchronization of sensor data.

3. The method for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 1, characterized in that, The preprocessing includes: denoising, completion, and standardization of the process data; The denoising method involves using a window moving average method to perform moving average filtering on the time series data, and using wavelet decomposition to filter out background noise from the data calculated by the near-infrared spectrometer. The completion refers to repairing and completing the data after denoising, using linear interpolation to repair data loss caused by temporary sensor failure or signal transmission loss. The standardization process involves using Z-Score standardization to normalize the data.

4. The method for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 1, characterized in that, The optimization and control algorithm also includes constraints, which are set based on the actual production situation and include: control variable constraints, used to characterize the physical limits of various control parameters of the ultrasonic generator and cooling system; and process constraints, used to characterize the performance-related limitations of the tallow crystallization process.

5. A control system for controlling the crystal morphology of tallow based on ultrasound assistance, wherein the control method for controlling the crystal morphology of tallow based on ultrasound assistance as described in any one of claims 1 to 4 is characterized in that, The control system includes a tallow cooling crystallization tank, an ultrasonic generator, a cooling system, and processing equipment; The tallow cooling crystallization tank is equipped with a sensor cluster module; the cooling system includes a heating / cooling jacket for temperature control and a low-speed stirring device; the processing equipment is used to process the data collected by the sensor cluster module and adjust the parameters of the ultrasonic generator and the cooling system according to the data processing results. The control system also includes a data preprocessing module, a data analysis module, and a control module; The data preprocessing module is configured to preprocess the data collected by the sensor cluster module, including performing noise reduction, completion and standardization operations. The control system also includes a data preprocessing module, a data analysis module, and an optimization control module. The data preprocessing module is configured to preprocess the data collected by the sensor cluster module, align the preprocessed data with timestamps, integrate and construct a sensor historical database. The data analysis module is connected to the data preprocessing module and is configured to analyze the data in the sensor historical database, construct a dynamic model of key indicators, and design an optimized control algorithm based on the dynamic model of key indicators. The optimization and control module is connected to the data analysis module and the sensor cluster module. It is configured to calculate the predicted value based on the real-time data transmitted back from the sensor cluster module and the optimization and control algorithm, compare the predicted value with the real-time sensor data, and dynamically adjust the parameters of the dynamic model of key indicators based on the comparison results.

6. The control system for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 5, characterized in that, The ultrasonic generator emits ultrasonic frequencies between 20 and 100 kHz.

7. The control system for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 5, characterized in that, The low-speed stirring device can stir at a speed of 20-50 rpm to ensure a uniform distribution of energy and temperature fields in the tallow slurry system.

8. The control system for controlling the crystallization morphology of tallow based on ultrasound assistance according to claim 5, characterized in that, The sensor cluster module is configured to install different types of sensors at key locations in the tallow cooling crystallization tank. The different types of sensors transmit the collected data to the processing device in real time via an IoT interface. The processing device integrates the data from different types of sensors to achieve sensor data synchronization. A temperature sensor is used to collect the temperature of the tallow slurry, an ultrasonic power meter is used to monitor the real-time ultrasonic output power, a viscometer is used to measure the effective viscosity, and a near-infrared spectrometer is used to monitor the total crystallinity and the proportion of β' crystal form of the tallow.

Images