Intelligent control system and method for enzymatic animal protein production process

Abstract

The present invention relates to the technical field of protein enzymolysis, and discloses an intelligent control system and method for an enzymatic animal protein production process, comprising the steps of obtaining key attribute parameters of an animal protein raw material, establishing a raw material grading model, and obtaining an initial raw material grading; matching a preset grading correspondence library according to the initial grading to obtain an optimal production parameter range in the initial stage; collecting first production parameters before enzymolysis to obtain adjusted production parameters; starting enzymolysis according to the adjusted parameters, collecting key process parameters in real time, inputting an optimal production parameter prediction model, and predicting the optimal production parameters at the next moment; correcting the optimal production parameters if they exceed a safe range; continuing enzymolysis according to the corrected parameters, and continuously predicting and correcting the parameters until the enzymolysis is completed; the present invention can dynamically regulate process parameters for different batches of raw materials through raw material grading and real-time parameter optimization, thereby maximizing production efficiency while ensuring product quality, and has significant intelligence and adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of protein enzymolysis, and more particularly to an intelligent control system and method for an enzymatically hydrolyzed animal protein production process. Background Art

[0002] Protein enzymatic hydrolysis is an important technical means for preparing functional foods and bioactive peptides. Traditional enzymatic hydrolysis process control relies primarily on experience and manual operation, lacking intelligent management and precise regulation, resulting in low production efficiency and unstable product quality. With the development of biotechnology and information technology, the introduction of intelligent control methods into the protein enzymatic hydrolysis production process, enabling real-time optimization and dynamic adjustment of process parameters, has become an important way to improve enzymatic hydrolysis efficiency and ensure product quality.

[0003] Chinese patent application publication number CN118550263A discloses a control system and method for silicone sealant production. This method uses real-time monitoring equipment to collect key production parameters, dynamically calibrates these parameters based on historical data, and applies hidden Markov models and support vector machines to predict future state transitions and key defects. This method focuses on predicting and intervening in production defects through data-driven modeling, but it lacks consideration for the dynamic optimization of production parameters caused by differences in raw material properties and lacks intelligent grading and parameter matching mechanisms based on raw material characteristics.

[0004] Chinese patent application publication number CN104419697A discloses a method for efficiently regulating enzymatic reactions in aqueous phases using temperature. This method utilizes the temperature-responsive properties of the intelligent polymer gel PNIPAM to control the rate of enzymatic reactions through temperature cycling. This method cleverly utilizes external temperature stimulation to achieve on-off regulation of the enzymatic reaction, but it only considers the single temperature factor and lacks comprehensive optimization of other key parameters such as pH and enzyme concentration. Furthermore, this method focuses on enzymatic reactions in aqueous systems and does not adequately consider the process characteristics of solid-liquid multiphase enzymatic hydrolysis systems.

[0005] Chinese patent application publication number CN115078298A discloses an intelligent control system and method for protein hydrolysis based on spectral monitoring. This system utilizes near-infrared spectroscopy to monitor the hydrolysis process in situ and uses physicochemical indicators such as biological activity to determine the hydrolysis endpoint and the timing of multi-enzyme switching. This method can monitor the hydrolysis process in real time and reduce hydrolysis duration, but it primarily focuses on endpoint control and pays insufficient attention to optimizing production parameters during the raw material feeding and hydrolysis startup phases. It also lacks a dynamic parameter optimization strategy throughout the entire hydrolysis process.

[0006] In summary, most of the existing enzymatic production control methods focus on the optimization control of a specific link or a certain type of parameter, and lack the integrated analysis of multi-source heterogeneous information such as raw material properties, process parameters, and enzymatic hydrolysis process, as well as the optimization of the entire process. Summary of the Invention

[0007] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent control system and method for the production process of enzymatically hydrolyzed animal protein. In view of the large differences in the characteristics of animal protein raw materials and the complex and changeable enzymatic hydrolysis process, the dynamic optimization and precise control of the entire production process from the raw material end to the finished product end are realized. The intelligent control method for the production process of enzymatically hydrolyzed animal protein proposed in the present invention obtains the key attribute parameters of the raw materials to be hydrolyzed and establishes a raw material grading model, matches the optimal production parameter range, collects the key parameters of the enzymatic hydrolysis process and dynamically predicts and corrects the optimal production parameters until the end of the enzymatic hydrolysis process, thereby realizing intelligent grading processing of different batches of raw materials and dynamic optimization of parameters of the entire enzymatic hydrolysis process. It has significant intelligence and adaptability, can significantly improve production efficiency and product quality, and overcomes the shortcomings of the prior art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] The intelligent control method for the production process of enzymatic animal protein includes:

[0010] Acquire key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed, establish a raw material grading model, input the collected key attribute parameters into the raw material grading model to obtain an initial grading of the raw material; based on the initial grading of the raw material, match a preset grading correspondence library to obtain the optimal production parameter ranges for the initial stage of different batches of raw materials; collect the first production parameters before enzymatic hydrolysis, and obtain adjusted production parameters based on the first production parameters and the optimal production parameter ranges;

[0011] The enzymatic hydrolysis reaction is started according to the adjusted production parameters, and the key parameters of the enzymatic hydrolysis process are collected in real time to form a dynamically updated process parameter sequence; the process parameter sequence collected in real time is input into the pre-trained optimal production parameter prediction model, and the optimal production parameter combination at the next moment t+Δt is dynamically predicted; the optimal production parameters in the optimal production parameter combination are compared with the preset process safety threshold. When they exceed the safety range, the optimal production parameters are corrected for safety to obtain a corrected production parameter combination; the enzymatic hydrolysis process is continued according to the corrected production parameter combination; and during the execution process, the key parameters of the enzymatic hydrolysis process are continuously monitored and collected, and the production parameters are predicted and corrected until the end of the enzymatic hydrolysis process, t is the current moment, and Δt is the prediction time step.

[0012] Furthermore, the key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed include the freshness, fat content and tissue structure of the raw material;

[0013] The freshness of the raw material is calculated by weighting the pH value and volatile basic nitrogen content of the raw material.

[0014] Furthermore, the adjusted production parameters obtained according to the first production parameter and the optimal production parameter range include: real-time collection of the first production parameter P1=(pH1, T1, E1) after the raw material is added and before the start of enzymatic hydrolysis, where pH1 is the measured pH value of the enzymatic hydrolysis system, T1 is the measured temperature of the enzymatic hydrolysis system, and E1 is the measured enzyme concentration of the enzymatic hydrolysis system;

[0015] Taking the median of the optimal range of each parameter in the optimal production parameter range as a reference value, respectively calculating the deviations of pH1, T1, and E1 from the reference values, and calculating a first deviation value Δδ based on the deviations of pH1, T1, and E1 from the reference values;

[0016] The first deviation value Δδ is compared with a preset first deviation threshold ε1. When Δδ>ε1, the first adjustment of the production parameters is triggered to obtain the adjusted production parameters.

[0017] Furthermore, the optimal production parameters in the optimal production parameter combination include optimal temperature, optimal pH value and optimal enzyme concentration;

[0018] Comparing the optimal production parameter in the optimal production parameter combination with a preset process safety threshold includes:

[0019] When T' <T min or T'>T max When , it is determined that T' is beyond the safety range;

[0020] When pH <pH min or pH'>pH max When pH' is beyond the safe range;

[0021] When E' <E min or E'>E max When , it is determined that E' is beyond the safety range;

[0022] Among them, T' is the optimal temperature, pH' is the optimal pH value, E' is the optimal enzyme concentration, T min is the lower limit of the temperature safety threshold range, T max is the upper limit of the temperature safety threshold range, pH min The lower limit of the pH safety threshold range, pH max is the upper limit of the pH safety threshold range, E min is the lower limit of the enzyme concentration safety threshold, E max It is the upper limit of the enzyme concentration safety threshold range.

[0023] Furthermore, the safety correction of the optimal production parameters includes safety correction of the optimal temperature T', safety correction of the optimal pH value pH', and safety correction of the optimal enzyme concentration E';

[0024] The safety correction of the optimal temperature T' includes:

[0025] T*=min(max(T',T min ),T max );

[0026] Where T* is the corrected temperature.

[0027] Furthermore, the safety correction of the optimal pH value pH' includes:

[0028] pH*=min(max(pH',pH min ),pH max );

[0029] Wherein, pH* is the corrected pH value.

[0030] Furthermore, the safety correction of the optimal enzyme concentration E' includes:

[0031] E*=min(max(E',E min ),E max );

[0032] Where E* is the corrected enzyme concentration.

[0033] An intelligent control system for an enzymatically hydrolyzed animal protein production process, which is used to implement the above-mentioned intelligent control method for an enzymatically hydrolyzed animal protein production process, comprises:

[0034] Raw material grading module: used to obtain key attribute parameters of animal protein raw materials to be enzymatically hydrolyzed, establish a raw material grading model, input the collected key attribute parameters into the raw material grading model, and obtain the initial classification of the raw materials;

[0035] Parameter adjustment module: used to match the preset classification correspondence library according to the initial classification of raw materials, and obtain the optimal production parameter range for the initial stage of different batches of raw materials; collect the first production parameters before enzymatic hydrolysis, and obtain the adjusted production parameters based on the first production parameters and the optimal production parameter range;

[0036] Parameter prediction module: used to start the enzymatic hydrolysis reaction according to the adjusted production parameters, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence; the real-time collected process parameter sequence is input into the pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment t+Δt, where t is the current moment and Δt is the prediction time step;

[0037] Parameter correction module: used to compare the optimal production parameters in the optimal production parameter combination with the preset process safety threshold. When the optimal production parameters exceed the safety range, the optimal production parameters are corrected for safety to obtain the corrected production parameter combination;

[0038] Loop module: used to continue the enzymatic hydrolysis process according to the revised production parameter combination; and continuously monitor and collect key parameters of the enzymatic hydrolysis process during the execution process, predict and correct production parameters until the enzymatic hydrolysis process is completed.

[0039] An electronic device includes a memory, a central processing unit, and a computer program stored in the memory and executable on the central processing unit. When the central processing unit executes the computer program, the intelligent control method for the production process of enzymatic animal protein is realized.

[0040] A computer-readable storage medium stores a computer program, which, when executed, implements the above-mentioned intelligent control method for the production process of enzymatic animal protein.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention establishes a raw material grading model, comprehensively considers key attributes such as raw material freshness, fat content, and tissue structure, realizes accurate grading of raw materials, and provides a basis for differentiated control of the enzymatic hydrolysis process; matches preset optimal production parameter ranges for raw materials of different grades, provides a reference for the initial setting of process parameters, and reduces trial and error costs; through real-time collection of key parameters of the enzymatic hydrolysis process, inputs a pre-trained parameter prediction model, and dynamically predicts the optimal production parameter combination at the next moment, thereby realizing real-time optimization control of the enzymatic hydrolysis process; compares the predicted optimal parameters with safety thresholds, and makes corrections when they exceed the range, continuously optimizes production parameters and improves production efficiency while ensuring process safety; continuously collects key parameters and predicts and corrects parameters during the enzymatic hydrolysis process until the end of the enzymatic hydrolysis, realizes dynamic optimization of the entire process, and has strong adaptability; through the combination of graded processing and real-time optimization, dynamically regulates enzymatic hydrolysis parameters for raw materials of different qualities, while ensuring product yield and quality, shortens the production cycle to the maximum extent, and improves equipment utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 The figure is a flow chart showing the principle of the intelligent control method for the enzymatic animal protein production process of the present invention;

[0045] Figure 2 This is a flow chart of a method for obtaining adjusted production parameters in the intelligent control method for the enzymatic animal protein production process of the present invention;

[0046] Figure 3 This is a flow chart of a method for obtaining a corrected production parameter combination in the intelligent control method for the enzymatic animal protein production process of the present invention;

[0047] Figure 4 This is a functional module diagram of the intelligent control system for the enzymatic animal protein production process of the present invention. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0049] Example 1

[0050] See also Figure 1 As shown, this embodiment provides an intelligent control method for the production process of enzymatic animal protein, including:

[0051] Step S1000, obtain key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed, establish a raw material grading model, input the collected key attribute parameters into the raw material grading model, and obtain the initial grading of the raw material; according to the initial grading of the raw material, match the preset grading correspondence library to obtain the optimal production parameter range of different batches of raw materials in the initial stage; collect the first production parameter before enzymatic hydrolysis, calculate the first deviation value between the first production parameter and the optimal production parameter range, and trigger the first adjustment when the first deviation value exceeds the preset first deviation threshold to obtain the adjusted production parameter.

[0052] Furthermore, step S1000 includes:

[0053] Step S1100, obtaining basic information and key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed;

[0054] Furthermore, step S1100 includes:

[0055] Step S1110, obtaining basic information such as batch identification ID, source information, and timestamp of the animal protein raw material to be enzymatically hydrolyzed;

[0056] Step S1120, collecting key attribute parameters such as freshness, fat content, and tissue structure of different batches of raw materials;

[0057] The method for obtaining the freshness of the raw material is: measuring the pH value and volatile basic nitrogen content of the raw material, and calculating the freshness of the raw material by weighted calculation of the pH value and volatile basic nitrogen content of the raw material;

[0058] Step S1130: storing the key attribute parameters and basic information of the raw materials into the raw material information database.

[0059] Specifically, in step S1110, the unique ID of the raw material batch is obtained through automatic identification technologies such as scanning QR codes and RFID. Source information such as the origin, supplier, and arrival time of the raw materials is manually recorded to establish a traceability file for the raw material batch. Step S1120 uses sensor technology and physical and chemical testing to obtain key attribute parameters reflecting raw material quality. A pH meter and a volatile basic nitrogen meter are used to measure the pH and volatile basic nitrogen content of the raw materials, respectively, and a weighted algorithm is designed to calculate the raw material freshness. Fat content is determined using a standard Soxhlet extraction method, where fat is extracted from the raw materials and weighed to calculate the percentage. Characterizing tissue structure is complex, requiring the raw material tissue to be sliced and its fiber structure observed under a microscope. Image processing algorithms are then used to extract characteristic parameters such as fiber orientation, thickness, and density. Step S1130 aggregates and stores this collected information, creating a "file" for the raw materials. This allows for precise traceability and attribute query for each batch of raw materials, providing data support for subsequent grading and optimization control. Through information digitization and associative storage, refined raw material management is achieved.

[0060] Step S1200: Establish a raw material classification model, input the collected key attribute parameters into the raw material classification model, and obtain an initial classification of the raw materials.

[0061] The method for establishing a raw material grading model comprises: obtaining raw material historical data, wherein the raw material historical data includes key attribute parameters of different batches of raw materials in the historical enzymatic hydrolysis process, including freshness, fat content and tissue structure; using a K-means clustering algorithm to perform cluster analysis on the raw material historical data, with the number of clusters set to 3, and dividing the raw materials into three grades: high-quality material A, medium-quality material B and low-quality material C, to obtain a raw material grading model.

[0062] The process of clustering analysis of raw material historical data using K-means clustering algorithm is as follows:

[0063] Determine the number of clusters K: the raw materials are divided into 3 levels, so K = 3.

[0064] Initialize cluster centers: Randomly select 3 data points as initial cluster centers.

[0065] Divide the data: Calculate the distance between each data point and the three cluster centers and divide it into the nearest class.

[0066] Update center: Calculate the data centers of the three clusters respectively as the new cluster centers.

[0067] Repeat the above two steps until the cluster center no longer changes significantly or the maximum number of iterations is reached.

[0068] Historical raw material data undergoes necessary data cleaning and preprocessing, converting it into a numerical matrix suitable for cluster analysis. Numerical values for different indicators may vary significantly, requiring standardization for comparability. High-quality raw material A represents the highest quality, characterized by high freshness, moderate fat content, and loose texture; low-quality raw material C represents the lowest quality, characterized by poor freshness, excessively high or low fat content, and dense texture. Therefore, clustering results allow for rapid determination of the raw material's grade, laying the foundation for subsequent differentiated processing. Clustering algorithms such as K-means are used to automatically cluster the existing raw material data. The clustering process uses freshness, fat content, and texture as feature vectors, automatically categorizing raw material samples by calculating the similarity between these vectors. Based on the clustering results, raw materials are categorized into three grades: high-quality raw material A, medium-quality raw material B, and low-quality raw material C. When a new batch of raw materials arrives at the factory, its key attribute parameters are simply input into the trained raw material grading model. The model automatically identifies the most similar clusters for the raw material, thereby determining its initial grade. The establishment of the raw material grading model makes full use of historical big data to form an intelligent decision-making tool for raw material grading, which can significantly improve the accuracy and automation level of grading.

[0069] Step S1300 , based on the initial classification of the raw materials, matching a preset classification correspondence library to obtain the optimal production parameter ranges for the initial stages of different batches of raw materials;

[0070] Specifically, in enzymatic hydrolysis production practice, suitable production parameters vary for raw materials of different quality grades. For high-quality raw materials, efficient enzymatic hydrolysis can be achieved within a wide parameter range. However, for low-quality raw materials, the enzymatic hydrolysis process is more sensitive to production parameters, resulting in a relatively strict parameter range. We manually summarized and organized practical experience in enzymatic hydrolysis production, summarized the optimal production parameter ranges for different raw material grading conditions, and constructed a "raw material grade - optimal production parameter range" classification relationship library. This library uses a structured representation, such as "IF raw material grade = A THEN pH range = [6.5, 7.5], temperature range = [50, 60]°C, enzyme concentration range = [1.5, 2.0] mg / g." These relationships can be used to guide the initial setting and dynamic adjustment of production parameters. Based on the initial raw material grade, pattern matching is performed within the library to quickly retrieve the optimal ranges for various production parameters, such as pH, temperature, and enzyme concentration, for the corresponding grade. The matched ranges of production parameters such as pH value, temperature, and enzyme concentration are used as constraints for production control in the initial stage to ensure that the parameters are set within the optimal range, creating favorable initial conditions for achieving efficient enzymatic hydrolysis. The establishment of a graded correspondence library enables the correlation and matching between raw material grading and production parameter settings. Differentiated control strategies can be adopted for raw materials of different qualities to improve the adaptability and accuracy of the enzymatic hydrolysis process. At the same time, the introduction of a graded correspondence library also reduces the blindness of empirical manual parameter adjustment and provides theoretical support for intelligent production control.

[0071] Step S1400: collecting the first production parameter before enzymatic hydrolysis, calculating the first deviation value between the first production parameter and the optimal production parameter range, triggering the first adjustment when the first deviation value exceeds the preset first deviation threshold, and obtaining the adjusted production parameter.

[0072] Further, if Figure 2 As shown, step S1400 includes:

[0073] Step S1410, collecting in real time the first production parameter P1 = (pH1, T1, E1) after the raw material is added and before the enzymatic hydrolysis begins, where pH1 is the measured pH value of the enzymatic hydrolysis system, T1 is the measured temperature of the enzymatic hydrolysis system, and E1 is the measured enzyme concentration of the enzymatic hydrolysis system;

[0074] Step S1420, using the median of the optimal range of each parameter in the optimal production parameter range as a reference value, respectively calculating the deviations of pH1, T1, and E1 from the reference values, and calculating a first deviation value Δδ based on the deviations of pH1, T1, and E1 from the reference values;

[0075] Step S1430: compare the first deviation value Δδ with a preset first deviation threshold ε1. When Δδ>ε1, trigger the first adjustment of the production parameters to obtain the adjusted production parameters.

[0076] Specifically, due to fluctuations in raw material properties and production errors, actual production parameters may deviate from the optimal range. To ensure that enzymatic hydrolysis proceeds under optimal conditions, parameters must be monitored and dynamically adjusted before enzymatic hydrolysis begins. In step S1410, online sensors for parameters such as pH, temperature, and enzyme concentration are installed in the enzymatic hydrolysis device. These sensors collect the first production parameter P1 in real time after the raw materials are added. pH sensors, temperature sensors, and concentration sensors are installed in the enzymatic hydrolysis tank to continuously monitor the pH value, temperature, and enzyme concentration of the enzymatic hydrolysis system. These parameters are key factors affecting enzymatic hydrolysis efficiency. In step S1420, the deviation of the measured parameters is calculated using the median of the optimal range of each parameter (e.g., the median of the pH range [6.5, 7.5] is 7.0) as a reference benchmark, thereby calculating the first deviation value. In step S1430, the first deviation value Δδ is compared with the preset first deviation threshold ε1. When Δδ ≤ ε1, the first adjustment of the production parameters is not triggered. When Δδ > ε1, the deviation exceeds the allowable range and the adjustment mechanism is triggered. The described first adjustment is based on the positive and negative and size of the deviation from the reference value of pH1, T1, E1, and the PID control algorithm is utilized to calculate the correction amount of each parameter, and control actuator (such as acid pump, alkali pump, heater, cooler, etc.) to adjust each parameter accordingly so that it returns to the optimal range. For example, when the pH measured value is lower than the lower limit of the range, the alkali pump is controlled to continue to add alkali solution until pH reaches the range median. The first adjustment realizes the rapid migration of production parameters from the initial state to the optimal state, reduces the adverse effects of parameter deviation on the enzymolysis effect, and ensures that enzymolysis starts under optimal conditions. Meanwhile, the introduction of sensor monitoring and PID control realizes the automation, intelligence of parameter adjustment, reduces the labor intensity of manual operation, and improves the accuracy and reliability of process control.

[0077] The method for calculating the first deviation value Δδ based on the deviation of pH1, T1, E1 from the reference value includes:

[0078]

[0079] in:

[0080] Δδ represents the first deviation value and is a scalar;

[0081] pH1 represents the measured pH value of the enzymatic hydrolysis system;

[0082] T1 represents the measured temperature of the enzymatic hydrolysis system;

[0083] E1 represents the measured enzyme concentration of the enzymatic hydrolysis system;

[0084] pH0 represents the reference value of pH in the optimal production parameter range (median of the optimal range);

[0085] T0 represents the reference value of temperature in the optimal production parameter range;

[0086] E0 represents the reference value of enzyme concentration in the optimal production parameter range;

[0087] α is the weight coefficient of pH, which indicates the contribution of pH deviation to the total deviation;

[0088] β is the weight coefficient of temperature, which indicates the contribution of temperature deviation to the total deviation;

[0089] γ is the weight coefficient of enzyme concentration, which indicates the contribution of enzyme concentration deviation to the total deviation; α, β, and γ are determined by those skilled in the art based on a large number of experiments;

[0090] λ is a positive scaling parameter that controls the decay rate of the exponential function in the denominator and can be optimized based on historical data fitting;

[0091] e is the base of natural logarithms, approximately equal to 2.71828.

[0092] The numerator reflects the cumulative effect of individual parameter deviations on the total deviation by linearly superimposing the absolute values of the deviations of the three parameters. The denominator introduces an exponential decay function, making Δδ larger when the deviation is small and smaller when the deviation is large, thereby increasing the formula's sensitivity to small deviations. Setting weight coefficients allows the importance of each parameter deviation to be adjusted based on process requirements. The scaling parameter λ controls the rate of exponential decay in the denominator. A larger λ value results in a faster decay of Δδ, thus adjusting the degree of nonlinearity in the formula.

[0093] When pH1, T1, and E1 deviate more from the reference values pH0, T0, and E0, the molecule increases linearly and the Δδ value increases accordingly.

[0094] When the deviation is small, the denominator is close to 1, and the Δδ value is mainly determined by the numerator, showing an approximately linear growth.

[0095] When the deviation is large, the denominator decreases rapidly, the growth rate of the Δδ value slows down, and the value gradually stabilizes.

[0096] By adjusting the relative sizes of α, β, and γ, the influence weights of pH, temperature, and enzyme concentration deviation on Δδ can be adjusted.

[0097] Increasing the value of λ can speed up the decay of the Δδ value, making it respond more quickly to deviations.

[0098] This formula has a good balance between computational efficiency and nonlinear sensitivity and is suitable for real-time monitoring and feedback control.

[0099] Step S1000 starts with raw material information collection, raw material grading, optimal range matching, deviation correction and other aspects to carry out fine classification of enzymatic raw materials and set initial parameters. Information collection realizes the digital characterization of raw material quality, grading correspondence library matching realizes the association between grading and parameter range, and online monitoring realizes real-time correction of deviations. The seamless connection of these links together constitutes an intelligent closed loop of raw material grading and parameter setting, which significantly improves the raw material adaptability and process accuracy of enzymatic production. At the same time, by constructing digital twins of raw materials, grading and parameters, this method has also accumulated rich big data for enzymatic optimization. These data can be used to continuously improve the grading model and control strategy, and promote the intelligent level of enzymatic production to a new level.

[0100] Step S2000, start the enzymatic hydrolysis reaction according to the adjusted production parameters, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence P(t); input the real-time collected process parameter sequence into the pre-trained optimal production parameter prediction model, and dynamically predict the optimal production parameter combination at the next moment; compare the optimal production parameters in the optimal production parameter combination with the preset process safety threshold. When the safety range is exceeded, the predicted optimal production parameters are corrected for safety to obtain a corrected production parameter combination (T*, pH*, E*); execute the enzymatic hydrolysis process according to the corrected production parameter combination (T*, pH*, E*); and continuously monitor and collect key parameters of the enzymatic hydrolysis process during the execution process, and continuously predict and correct the production parameters until the enzymatic hydrolysis process is completed.

[0101] Furthermore, step S2000 includes:

[0102] Step S2100: Initiate an enzymatic hydrolysis reaction, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence P(t); the key parameters of the enzymatic hydrolysis process include temperature, pH value, and enzyme concentration;

[0103] Specifically, after starting the enzymatic hydrolysis reaction, real-time online monitoring of the enzymatic hydrolysis process is achieved by installing various sensors on the enzymatic hydrolysis device. The temperature sensor collects the enzymatic hydrolysis temperature, the pH sensor collects the pH value of the enzymatic hydrolysis system, and the concentration sensor collects the real-time concentrations of the substrate and enzyme. These sensors continuously sample at a certain frequency to obtain time series data that reflects the dynamic changes in the enzymatic hydrolysis process. The collected parameter values are matched one-to-one with the timestamp to form structured time series data P(t), which fully records the changes in key parameters of the entire enzymatic hydrolysis process and provides a data basis for subsequent process optimization and control. Dynamic monitoring of the enzymatic hydrolysis process avoids the traditional offline detection mode and can promptly detect abnormal fluctuations in parameters, creating conditions for precise regulation. At the same time, the accumulation of massive process parameter time series data also provides data support for process optimization and product quality traceability, and has important application value.

[0104] Step S2200: Input the real-time collected process parameter sequence into the pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment;

[0105] Furthermore, step S2200 includes:

[0106] Step S2210, constructing and training an optimal production parameter prediction model;

[0107] The method for constructing and training an optimal production parameter prediction model includes:

[0108] A large amount of historical production data is obtained, including historical process parameter sequences and corresponding optimal production parameter combinations. The historical production data is preprocessed by cleaning and normalization, and divided into training and test sets according to the time series. An LSTM network structure is designed, including an input layer, an LSTM hidden layer, a fully connected layer, and an output layer. The LSTM network is trained using the historical process parameter sequence as input and the corresponding optimal production parameter combination as labels, and the network parameters are optimized using the backpropagation algorithm. The model performance is evaluated using the test set, and the model is tuned until the expected prediction accuracy is achieved. The trained LSTM model is saved as the final optimal production parameter prediction model for online prediction of the optimal production parameter combination for the enzymatic hydrolysis process.

[0109] In step S2220, the process parameter sequence P(t) collected in real time is input into the trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination (T', pH', E') at the next moment t+Δt; wherein T' is the optimal temperature, pH' is the optimal pH value, E' is the optimal enzyme concentration, t is the current moment, and Δt is the prediction time step.

[0110] Specifically, based on historical production data, a long short-term memory (LSTM) network from deep learning was used to construct an optimal production parameter prediction model. LSTM is a recursive neural network suitable for processing time series data. By introducing a gating mechanism, it can effectively capture long-range dependencies in time series. The historical production data is organized as a time series, with the process parameter sequence serving as the model input and the optimal production parameter combination at each moment as the model output. The LSTM network is trained by designing a rational network structure and training strategy, enabling the model to accurately establish the nonlinear mapping relationship between process parameters and the optimal production parameter combination, and to dynamically predict the optimal control strategy for the future. During the enzymatic hydrolysis process, the real-time process parameter sequence is input into the trained optimal production parameter prediction model. The model then dynamically predicts the optimal production parameter combination for the future based on parameter change trends. Deep learning predictive control transforms the optimal control problem of the enzymatic hydrolysis process into a time series prediction problem, fully leveraging the inherent laws of historical production data to achieve self-optimization of the control strategy. This predictive control approach effectively improves control accuracy and timeliness, opening up a new path for intelligent control of enzymatic hydrolysis processes.

[0111] In step S2300, the optimal production parameters predicted by the optimal production parameter prediction model are compared with the preset process safety threshold. When the safety range is exceeded, the predicted optimal production parameters are subjected to safety correction to obtain a corrected production parameter combination (T*, pH*, E*); if the safety range is not exceeded, the correction is not triggered.

[0112] Further, if Figure 3 As shown, step S2300 includes:

[0113] Step S2310: For the key parameters of the enzymatic hydrolysis process, temperature, pH value and enzyme concentration, set the temperature safety threshold range [T min ,T max ]、pH safety threshold range [pH min ,pH max ], enzyme concentration safety threshold range [E min ,E max ]; T min is the lower limit of the temperature safety threshold range, T max is the upper limit of the temperature safety threshold range, pH min The lower limit of the pH safety threshold range, pH max is the upper limit of the pH safety threshold range, E min is the lower limit of the enzyme concentration safety threshold, E max It is the upper limit of the enzyme concentration safety threshold range.

[0114] Step S2320, determining whether the optimal production parameters T', pH', and E' exceed corresponding safety thresholds;

[0115] When T' <T min or T'>T max When , it is determined that T' is beyond the safety range;

[0116] When pH <pH min or pH'>pH max When pH' is beyond the safe range;

[0117] When E' <E min or E'>E max When , it is determined that E' is beyond the safety range.

[0118] Step S2330: When the optimal production parameters exceed the safety range, the predicted optimal production parameters are corrected for safety and a corrected production parameter combination (T*, pH*, E*) is generated; wherein T* is the corrected temperature, pH* is the corrected pH value, and E* is the corrected enzyme concentration.

[0119] The method of performing safety correction on the optimal temperature T' to obtain the corrected temperature T* includes:

[0120] T*=min(max(T',T min ),T max );

[0121] The method for performing safety correction on the optimal pH value pH' to obtain the corrected pH value pH* includes:

[0122] pH*=min(max(pH',pH min ),pH max );

[0123] The method of performing safety correction on the optimal enzyme concentration E' to obtain the corrected enzyme concentration E* includes:

[0124] E*=min(max(E',E min ),E max ).

[0125] Specifically, after the optimal production parameter prediction model predicts the optimal production parameters, the predicted values need to be checked for safety. Safety thresholds are set for key parameters such as temperature, pH, and enzyme concentration based on the performance parameters of the process equipment and production operating procedures. The predicted values are compared with the thresholds to determine whether they exceed the safety margins. If a predicted parameter exceeds the specified limit, an emergency adjustment strategy is promptly implemented to correct it and bring the exceeding parameter within the safety range. This emergency adjustment can employ a simple upper and lower limit truncation method, adjusting values below the lower limit to the lower limit and values above the upper limit to the upper limit. Alternatively, more complex logical judgment and multi-objective optimization methods can be employed to minimize the impact on control performance by aligning the corrected parameters with the predicted values while ensuring safety. These corrected safety parameters are then issued to the actuators as the final control instructions. Safety correction of control parameters is an essential component of intelligent control systems, preventing parameter runaway due to model prediction errors, which can lead to safety incidents such as equipment damage and product failure. This demonstrates that intelligent control is safer and more reliable than traditional control. At the same time, safety threshold data from the production process can be used to continuously optimize the deep learning prediction model, enabling it to learn the safe boundaries of parameter values, reducing the probability of model predictions exceeding the standard, and improving the reliability of intelligent control. For example, when the pH value prediction result is pH' = 5.2, and the pH value safety threshold range is [5.5, 7.5], an emergency adjustment is triggered, correcting pH' to pH* = min(max(5.2, 5.5), 7.5) = 5.5, ensuring that the enzymatic hydrolysis process proceeds within the safe pH range.

[0126] In step S2400, the enzymatic hydrolysis process is continued according to the corrected production parameter combination (T*, pH*, E*), and the process returns to step S2100, and steps S2100 to S2400 are executed in a loop until the enzymatic hydrolysis process is completed.

[0127] Specifically, in step S2400, the enzymatic hydrolysis process is accurately executed according to the production parameter combination (T*, pH*, E*) obtained after the safety threshold is corrected. The heating power of the enzymatic hydrolysis tank is adjusted according to T*, and the temperature is accurately controlled by the PID algorithm with a deviation of within ±0.5°C. The acid-base regulator is automatically added according to pH*, and the dosage is calculated in real time by the PID algorithm, which can control the pH within ±0.1 units. The flow rate of the enzyme preparation dosing pump is adjusted according to E* to achieve dynamic regulation of the enzyme concentration and ensure that its fluctuation is within 5% of the optimal concentration. Through the above process, the dynamic stability of the optimal production parameter combination is maintained together, which provides the necessary conditions for the efficient implementation of the enzymatic hydrolysis process.

[0128] The intelligent control of the enzymatic hydrolysis process is a dynamic closed-loop process that requires continuous monitoring of process parameters, prediction of optimal parameters, and execution of parameter adjustments until the end of the enzymatic hydrolysis process. Therefore, after executing step S2400, it automatically returns to step S2100, re-collects process parameters, inputs the optimal production parameter prediction model, and performs parameter correction and execution based on the prediction results, and so on. This closed-loop control method relies on online sensing technology and data-driven models, and can perform adaptive adjustments to fluctuations in raw materials and processes. On the basis of ensuring process safety, it always maintains production parameters at the dynamic optimum, thereby achieving the purpose of improving quality and increasing efficiency.

[0129] In step S3000, after enzymatic hydrolysis is completed, the quality of the protein product is tested to obtain quality attribute data. The quality attribute data is associated with the initial classification of the batch of raw materials, the process parameter sequence, the optimal production parameter combination, and the revised production parameter combination to form a structured batch production file; the optimal production parameter prediction model is optimized, and the classification correspondence library is updated.

[0130] Specifically, high-performance liquid chromatography (HPLC), gas chromatography-mass spectrometry (GC-MS), and gel electrophoresis were used to qualitatively and quantitatively analyze the enzymatically hydrolyzed protein product. Protein yield was determined using the Coomassie Brilliant Blue method, and the yield percentage was calculated. An automatic amino acid analyzer was used to determine the amino acid composition and content of the protein product. Liquid chromatography-mass spectrometry (LC-MS) was used to analyze the molecular weight distribution of peptides. These test data characterized the key quality attributes of the protein product at both macro and micro levels. The quality attribute data obtained was correlated with production factors such as raw material information (grade, property parameters), process parameters (temperature, pH, etc.), and control decisions (adjustment records, predicted values), establishing a "raw material-process-product" data chain and forming a structured, traceable batch production profile. As a result of the production process, quality attributes inevitably have inherent connections with various production factors. Data analysis methods such as association rule mining and causal inference were used to deeply explore the implicit relationships between quality attributes and production factors, revealing the mechanisms by which each production link influences final quality. Through data collection and correlation analysis, quality feedback forms a closed loop with the production process, guiding process optimization and accumulating optimization knowledge. This process, through continuous iteration and upgrading within the "production-evaluation-optimization" cycle, continuously enhances the refinement and intelligence of the production process. This process introduces product quality testing into the feedback loop of production control, integrating all production links through digital means, forming a data-driven production optimization closed loop, which is key to achieving intelligent regulation of the enzymatic hydrolysis process.

[0131] Using machine learning methods such as incremental learning and online learning, existing models are retrained using newly collected closed-loop production optimization data. New data is divided into training and validation sets, and model parameters are fine-tuned to adapt to dynamic changes in data distribution. For example, for the raw material grading model, new raw material samples can be added to optimize cluster centers. For the optimal production parameter prediction model, more process parameter sequences can be accumulated to expand training data. This continuous learning of the model ensures that its predictive capabilities keep pace with the times and better meet actual production needs. Based on model retraining, cross-validation and other methods are used to evaluate the predictive performance of each model. Grid search and other methods are used to optimize model hyperparameters, control model complexity, and improve generalization. For models with poor performance, their limitations are analyzed and the model algorithm is improved, such as by adopting new neural network structures and loss functions. Through continuous iterative optimization, a highly accurate and robust intelligent algorithm model is ultimately obtained. The "raw material grading - optimal production parameter range" classification correspondence database is regularly maintained and updated through expert review and statistical analysis. Enzymatic production experts summarize and refine new production experience, developing new grading rules and optimization ranges, which are then added to the grading correspondence library. Statistical methods are also used to analyze the confidence and support of the rules in the grading correspondence library, eliminating outdated and invalid entries. The dynamic evolution of the grading correspondence library ensures the advancement and effectiveness of its knowledge, providing reliable theoretical guidance for intelligent production control.

[0132] Example 2

[0133] This embodiment, based on Example 1, provides an intelligent control method for the production process of enzymatically hydrolyzed animal protein, including:

[0134] Step S2330: When the optimal production parameters exceed the safety range, the predicted optimal production parameters are corrected for safety, and a corrected production parameter combination (T*, pH*, E*) is generated.

[0135] The calculation formula of the corrected temperature T* is:

[0136]

[0137] in:

[0138] T * : Corrected temperature;

[0139] T': optimal temperature predicted by the optimal production parameter prediction model;

[0140] T opt : The median of the temperature safety threshold range, used as a reference value for parameter adjustment, is derived from empirical data of the production process or through statistical analysis of historical production data;

[0141] T mid : The middle value of the temperature safety threshold range, used to judge the size of the deviation, usually taken as T min +T max -T min ) / 2;

[0142] λ T : The scaling parameter of the exponential function controls the decay rate of the correction term. A larger λ T Values will make corrections faster;

[0143] δ T : Correction coefficient, which controls the correction strength of the temperature and is usually determined by historical data or empirical values;

[0144] k T : Attenuation coefficient, controls the attenuation speed of the nonlinear offset, the larger the k T Values increase the sensitivity of the correction;

[0145] Δ T : The nonlinear offset added after correction, indicating further adjustment of the parameters when the temperature exceeds the safety threshold range.

[0146] exp(·): exponential function with the base e of the natural logarithm.

[0147] This formula uses exponential decay and nonlinear correction methods to ensure that when the predicted optimal temperature exceeds the temperature safety threshold, the production parameters can be adjusted quickly, and over-correction can be avoided to ensure the stability of the system. T and δ T Provides flexible control of the correction amplitude and speed to meet different process requirements. T The formula gently corrects temperature overshoots, avoiding dramatic parameter fluctuations. The formula's hierarchical structure ensures precise correction of small deviations and rapid response to large deviations, enabling intelligent, automated control of the enzymatic hydrolysis process.

[0148] When T′ approaches T opt When the correction term (T′-T opt ) is very small, the overall correction force is low, T * It is not much different from the predicted value T′.

[0149] As T′ increases beyond the threshold range, the correction term increases accordingly, but due to exponential decay, the growth rate of the correction term gradually slows down to avoid over-adjustment.

[0150] By adjusting the parameter λ T and δ T , can control the smoothness and sensitivity of the correction. Larger λ Tand smaller δ T will make the correction faster, and a smaller λ T This will make the correction smoother, meaning that the speed and magnitude of the correction will be reduced, making the parameter changes more gradual and stable.

[0151] This correction method can balance response speed and stability, and is particularly suitable for complex enzymatic hydrolysis processes, ensuring that production parameters are always within a safe range and improving production efficiency through precise control.

[0152] Example 3

[0153] This embodiment provides an intelligent control system for the enzymatic animal protein production process based on embodiment 1. Figure 4 As shown, including:

[0154] Raw material grading module: used to obtain key attribute parameters of animal protein raw materials to be enzymatically hydrolyzed, establish a raw material grading model, input the collected key attribute parameters into the raw material grading model, and obtain the initial classification of the raw materials;

[0155] Parameter adjustment module: used to match the preset classification correspondence library according to the initial classification of raw materials, and obtain the optimal production parameter range for the initial stage of different batches of raw materials; collect the first production parameters before enzymatic hydrolysis, and obtain the adjusted production parameters based on the first production parameters and the optimal production parameter range;

[0156] Parameter prediction module: used to start the enzymatic hydrolysis reaction according to the adjusted production parameters, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence; the real-time collected process parameter sequence is input into the pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment t+Δt, where t is the current moment and Δt is the prediction time step;

[0157] Parameter correction module: used to compare the optimal production parameters in the optimal production parameter combination with the preset process safety threshold. When the optimal production parameters exceed the safety range, the optimal production parameters are corrected for safety to obtain the corrected production parameter combination;

[0158] Loop module: used to continue the enzymatic hydrolysis process according to the revised production parameter combination; and continuously monitor and collect key parameters of the enzymatic hydrolysis process during the execution process, predict and correct production parameters until the enzymatic hydrolysis process is completed.

[0159] The raw material classification module includes:

[0160] Data acquisition unit: used to obtain basic information and key attribute parameters of the animal protein raw materials to be enzymatically hydrolyzed;

[0161] Grading model building unit: used to establish a raw material grading model, input the collected key attribute parameters into the raw material grading model, and obtain the initial grading of the raw materials.

[0162] In the data acquisition unit, the basic information and key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed are obtained, including:

[0163] Step S1110, obtaining basic information such as batch identification ID, source information, and timestamp of the animal protein raw material to be enzymatically hydrolyzed;

[0164] Step S1120, collecting key attribute parameters such as freshness, fat content, and tissue structure of different batches of raw materials;

[0165] The method for obtaining the freshness of the raw material is: measuring the pH value and volatile basic nitrogen content of the raw material, and calculating the freshness of the raw material by weighted calculation of the pH value and volatile basic nitrogen content of the raw material;

[0166] Step S1130: storing the key attribute parameters and basic information of the raw materials into the raw material information database.

[0167] The parameter adjustment module includes:

[0168] Matching unit: used to match the preset classification correspondence library according to the initial classification of raw materials, and obtain the optimal production parameter range for the initial stage of different batches of raw materials;

[0169] Adjustment unit: used to collect the first production parameter before enzymatic hydrolysis, calculate the first deviation value between the first production parameter and the optimal production parameter range, and trigger the first adjustment when the first deviation value exceeds the preset first deviation threshold to obtain the adjusted production parameter.

[0170] The parameter prediction module includes:

[0171] Real-time data acquisition unit: used to start the enzymatic hydrolysis reaction, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence;

[0172] Prediction unit: used to input the real-time collected process parameter sequence into the pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment.

[0173] Example 4

[0174] This embodiment discloses an electronic device that may include one or more processors and one or more memories. The memories may store computer-readable code that, when executed by the one or more processors, may implement the intelligent control method for an enzymatically hydrolyzed animal protein production process.

[0175] The method or system according to the embodiment of the present application can also be implemented with the help of the architecture of an electronic device. The electronic device may include a bus, one or more CPUs, a read-only memory (ROM), a random access memory (RAM), a communication port connected to a network, an input / output component, a hard disk, etc. A storage device in the electronic device, such as a ROM or a hard disk, can store the intelligent control method for the enzymatic animal protein production process provided in this application. The intelligent control method for the enzymatic animal protein production process may, for example, include: obtaining key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed, establishing a raw material grading model, inputting the collected key attribute parameters into the raw material grading model to obtain an initial grading of the raw material; according to the initial grading of the raw material, matching a preset grading correspondence library to obtain the optimal production parameter range for the initial stage of different batches of raw materials; collecting the first production parameters before enzymatic hydrolysis, and obtaining the adjusted production parameters based on the first production parameters and the optimal production parameter range; starting the enzymatic hydrolysis reaction according to the adjusted production parameters, collecting the key parameters of the enzymatic hydrolysis process in real time, and forming a dynamically updated process parameter sequence; inputting the real-time collected process parameter sequence into a pre-trained optimal production parameter prediction model, and dynamically predicting the optimal production parameter combination at the next moment t+Δt; comparing the optimal production parameter in the optimal production parameter combination with a preset process safety threshold, and when it exceeds the safety range, performing safety correction on the optimal production parameter to obtain a corrected production parameter combination; continuing the enzymatic hydrolysis process according to the corrected production parameter combination; and continuously monitoring and collecting the key parameters of the enzymatic hydrolysis process during the execution process, predicting and correcting the production parameters until the end of the enzymatic hydrolysis process, t is the current moment, and Δt is the prediction time step.

[0176] Furthermore, the electronic device may further include a user interface. Of course, the architecture disclosed in the present invention is only exemplary, and when implementing different devices, one or more components in the electronic device disclosed in the present invention may be omitted according to actual needs.

[0177] Example 5

[0178] This embodiment discloses a computer-readable storage medium having computer-readable instructions stored thereon. When the computer-readable instructions are executed by a processor, the intelligent control method for the enzymatic animal protein production process according to the embodiment of the present application described with reference to the above drawings can be executed. The storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may, for example, include random access memory (RAM) and cache memory (cache). Non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0179] In addition, according to the embodiment of the present application, the process described in the above reference flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be run by a processor to execute instructions corresponding to the method steps provided in the present application, such as: obtaining key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed, establishing a raw material grading model, inputting the collected key attribute parameters into the raw material grading model, and obtaining the initial grading of the raw material; according to the initial grading of the raw material, matching the preset grading correspondence library, and obtaining the optimal production parameter range for the initial stage of different batches of raw materials; collecting the first production parameter before enzymatic hydrolysis, and obtaining the adjusted production parameter according to the first production parameter and the optimal production parameter range; according to the adjusted production parameter, Production parameters start the enzymolysis reaction, collect key parameters of the enzymolysis process in real time, and form a dynamically updated process parameter sequence; the process parameter sequence collected in real time is input into the pre-trained optimal production parameter prediction model, and the optimal production parameter combination at the next moment t+Δt is dynamically predicted; the optimal production parameters in the optimal production parameter combination are compared with the preset process safety threshold value. When the safety range is exceeded, the optimal production parameters are corrected for safety to obtain the corrected production parameter combination; the enzymolysis process is continued according to the corrected production parameter combination; and the key parameters of the enzymolysis process are continuously monitored and collected during the execution process, and the production parameters are predicted and corrected until the end of the enzymolysis process, t is the current moment, and Δt is the prediction time step. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the method of the present application are executed.

[0180] The methods, systems, and devices of the present application may be implemented in many ways. For example, the methods, systems, and devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified. In addition, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, which include machine-readable instructions for implementing the methods according to the present application. Therefore, the present application also covers recording media that store programs for executing the methods according to the present application.

[0181] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0182] The above-described specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. An intelligent control method for enzymatic animal protein production process, characterized in that: The method comprises: Acquiring key attribute parameters of the animal protein raw material to be enzymatically hydrolyzed, establishing a raw material grading model, and inputting the collected key attribute parameters into the raw material grading model to obtain an initial grading of the raw material; the key attribute parameters of the raw material include freshness, fat content, and tissue structure of the raw material; matching a preset grading correspondence library based on the initial grading of the raw material to obtain an optimal production parameter range for the initial stage of the batch of raw materials; collecting first production parameters before enzymatic hydrolysis, and obtaining adjusted production parameters based on the first production parameters and the optimal production parameter range; The method for establishing a raw material grading model comprises: obtaining raw material historical data, wherein the raw material historical data includes key attribute parameters of different batches of raw materials in the historical enzymatic hydrolysis process; performing cluster analysis on the raw material historical data using a K-means clustering algorithm, setting the number of clusters to 3, and dividing the raw materials into three grades: high-quality material A, medium-quality material B, and low-quality material C, to obtain a raw material grading model; The method of obtaining the adjusted production parameters according to the first production parameter and the optimal production parameter range includes: collecting the first production parameter P1=(pH1, T1, E1) in real time after the raw material is fed and before the enzymatic hydrolysis begins, where pH1 is the measured pH value of the enzymatic hydrolysis system, T1 is the measured temperature of the enzymatic hydrolysis system, and E1 is the measured enzyme concentration of the enzymatic hydrolysis system; using the median of the optimal range of each parameter in the optimal production parameter range as a reference value, respectively calculating the deviations of pH1, T1, and E1 from the reference values, and calculating a first deviation value Δδ based on the deviations of pH1, T1, and E1 from the reference values; comparing the first deviation value Δδ with a preset first deviation threshold ε1, and when Δδ>ε1, triggering a first adjustment of the production parameters to obtain the adjusted production parameters; The enzymatic hydrolysis reaction is started according to the adjusted production parameters, and the key parameters of the enzymatic hydrolysis process are collected in real time to form a dynamically updated process parameter sequence, wherein the key parameters of the enzymatic hydrolysis process include temperature, pH value and enzyme concentration; the process parameter sequence collected in real time is input into a pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment t+Δt; the optimal production parameter prediction model is a long short-term memory network; the optimal production parameters in the optimal production parameter combination are compared with the preset process safety threshold, and when they exceed the safety range, the optimal production parameters are corrected for safety to obtain a corrected production parameter combination; the optimal production parameters in the optimal production parameter combination include optimal temperature, optimal pH value and optimal enzyme concentration; the enzymatic hydrolysis process is continued according to the corrected production parameter combination; and during the execution process, the key parameters of the enzymatic hydrolysis are continuously monitored and collected, and the production parameters are predicted and corrected until the end of the enzymatic hydrolysis process, t is the current moment, and Δt is the prediction time step.

2. The intelligent control method for the production process of enzymatic animal protein according to claim 1, characterized in that: The freshness of the raw material is calculated by weighting the pH value and volatile basic nitrogen content of the raw material.

3. The intelligent control method for the production process of enzymatic animal protein according to claim 1, characterized in that: Comparing the optimal production parameter in the optimal production parameter combination with a preset process safety threshold includes: When T' <T min or T'>T max When , it is determined that T' is beyond the safety range; When pH <pH min or pH'>pH max When pH' is beyond the safe range; When E' <E min or E'>E max When , it is determined that E' is beyond the safety range; Among them, T' is the optimal temperature, pH' is the optimal pH value, E' is the optimal enzyme concentration, T min， is the lower limit of the temperature safety threshold range, T max is the upper limit of the temperature safety threshold range, pH min The lower limit of the pH safety threshold range, pH max is the upper limit of the pH safety threshold range, E min is the lower limit of the enzyme concentration safety threshold, E max It is the upper limit of the enzyme concentration safety threshold range.

4. The intelligent control method for the production process of enzymatic animal protein according to claim 1, characterized in that: The safety correction of the optimal production parameters includes safety correction of the optimal temperature T', safety correction of the optimal pH value pH', and safety correction of the optimal enzyme concentration E'; The safety correction of the optimal temperature T' includes: T*=min(max(T',T min ),T max ); Where T* is the corrected temperature.

5. The intelligent control method for the production process of enzymatic animal protein according to claim 4, characterized in that: The safety correction of the optimal pH value pH' comprises: pH*=min(max(pH',pH min ),pH max ); Wherein, pH* is the corrected pH value.

6. The intelligent control method for the production process of enzymatic animal protein according to claim 4, characterized in that: The safety correction of the optimal enzyme concentration E' comprises: E*=min(max(E',E min ),AND max ); Where E* is the corrected enzyme concentration.

7. An intelligent control system for an enzymatically hydrolyzed animal protein production process, which is used to implement the intelligent control method for an enzymatically hydrolyzed animal protein production process according to any one of claims 1 to 6, characterized in that: The system comprises: Raw material grading module: used to obtain key attribute parameters of animal protein raw materials to be enzymatically hydrolyzed, establish a raw material grading model, input the collected key attribute parameters into the raw material grading model, and obtain the initial classification of the raw materials; Parameter adjustment module: used to match the preset classification correspondence library according to the initial classification of raw materials to obtain the optimal production parameter range for the initial stage of the batch of raw materials; collect the first production parameters before enzymatic hydrolysis, and obtain the adjusted production parameters based on the first production parameters and the optimal production parameter range; Parameter prediction module: used to start the enzymatic hydrolysis reaction according to the adjusted production parameters, collect key parameters of the enzymatic hydrolysis process in real time, and form a dynamically updated process parameter sequence; the real-time collected process parameter sequence is input into the pre-trained optimal production parameter prediction model to dynamically predict the optimal production parameter combination at the next moment t+Δt, where t is the current moment and Δt is the prediction time step; Parameter correction module: used to compare the optimal production parameters in the optimal production parameter combination with the preset process safety threshold. When the optimal production parameters exceed the safety range, the optimal production parameters are corrected for safety to obtain the corrected production parameter combination; Loop module: used to continue the enzymatic hydrolysis process according to the revised production parameter combination; and continuously monitor and collect key parameters of enzymatic hydrolysis during the execution process, predict and correct production parameters until the enzymatic hydrolysis process is completed.

8. An electronic device comprising a memory, a central processing unit, and a computer program stored in the memory and executable on the central processing unit, wherein: When the central processing unit executes the computer program, the intelligent control method for the enzymatic animal protein production process according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed, implements the intelligent control method for the enzymatic animal protein production process according to any one of claims 1 to 6.

Images