A method for flavor balance optimization in a milk base concentration process

Abstract

This application provides a method for optimizing flavor balance during the concentration process of milk base material, comprising: acquiring initial concentration state indicators by collecting base material sample data during the concentration process; determining a list of key flavor small molecule components by processing the composition of lost substances using a permeate analysis algorithm based on the initial concentration state indicators; obtaining the mixed base material sample corresponding to the optimized flavor intensity equilibrium point and verifying the protein retention rate through secondary membrane filtration to determine the final protein flavor balance ratio; obtaining a stable quality output base material sequence by processing production batch data through continuous fluid simulation based on the final protein flavor balance ratio; if the level of perception parameter fluctuation in the stable quality output base material sequence exceeds a threshold, correcting the concentration parameter settings using a feedback adjustment algorithm to obtain a refining production process configuration; and extracting key operation node data from the refining production process configuration and verifying the overall quality indicators through batch simulation to determine an optimized milk base material generation path.

Description

Technical Field

[0001] This invention relates to the field of dairy product production technology, and in particular to a method for optimizing flavor balance during the concentration process of milk base materials. Background Technology

[0002] Yogurt, as a daily consumer fermented dairy product, is highly favored by consumers for its nutritional value and flavor characteristics. The development of high-protein yogurt has become an important direction in the industry. Concentrating protein to improve product texture, fullness, and health attributes has significant market implications. During the production process, the effective retention of protein directly affects the product's functionality and quality stability, making protein concentration technology a crucial step in yogurt processing.

[0003] While ultrafiltration technology, commonly used today, can effectively retain proteins for concentration, the membrane separation process often removes some small-molecule flavor compounds, resulting in a weakened flavor intensity of the concentrated base and a loss of the original fermented milk's characteristic sour aroma and aftertaste. Relying solely on ultrafiltration makes it difficult to achieve a balance between protein retention and flavor preservation. This flavor loss necessitates the addition of flavorings or acidulants in subsequent blending, increasing costs and hindering the expression of the product's natural properties.

[0004] During ultrafiltration concentration of proteins, small-molecule flavor compounds such as organic acids, aromatic compounds, and flavor precursors are easily lost with the permeate. These substances are the core components that constitute the fermentation characteristics of yogurt. Because membranes have a certain rigidity in separating molecular weights, when proteins are highly retained, some beneficial small-molecule flavor molecules are also removed, while harmful or excess small molecules are difficult to selectively remove. This lack of separation selectivity leads to a shift in the flavor profile of the concentrated base material as the protein content increases, disrupting the originally harmonious balance of acidity and aroma.

[0005] Specifically, in actual production, after ultrafiltration concentrates the protein to the target concentration, the base material often exhibits a prominent protein texture but a bland sour and aromatic flavor. The typical lactic acid and acetaldehyde aromas produced during fermentation are significantly weakened, and the flavor of the final packaged product lacks complexity compared to regular yogurt. Therefore, effectively controlling the balance between the retention and loss of flavor molecules during ultrafiltration protein concentration, while simultaneously achieving efficient protein retention, has become a key issue in improving the overall quality of concentrated yogurt base materials. Summary of the Invention

[0006] This invention provides a method for optimizing flavor balance during the concentration process of milk base materials, mainly comprising: Initial concentration state indicators are obtained by collecting base material sample data during the concentration process. Based on the initial concentration state indicators, the composition of lost substances is processed using a permeate analysis algorithm to determine the list of key flavor small molecules. If the concentration of a specific organic acid in the list of key flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered through an ultrafiltration module to obtain a supplementary flavor extract. Aromatic compound concentration data is extracted from the supplementary flavor extract, and a flavor compensation algorithm is used to adjust the mixing ratio of the extract and the concentrated base material to determine the optimized flavor intensity equilibrium point. The mixed base material sample corresponding to the optimized flavor intensity equilibrium point is obtained, and secondary membrane filtration is performed to verify the protein retention rate to determine the final protein flavor balance ratio. Based on the final protein flavor balance ratio, production batch data is processed through continuous fluid simulation to obtain a stable quality output base material sequence. If the level of variation parameter in the stable quality output base material sequence exceeds a threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration. Key operation node data is extracted from the refining production process configuration, and batch simulation verification is performed to determine the optimized milk base material generation path for the overall quality indicators.

[0007] Furthermore, the step of obtaining initial concentration state indicators by collecting base material sample data during the concentration process includes: collecting base material sample data during the ultrafiltration process to obtain protein content and small molecule flavor substance distribution parameters; obtaining initial concentration state indicators from the protein content and small molecule flavor substance distribution parameters; the initial concentration state indicators characterize the protein retention and small molecule flavor substance loss of the concentrated base material during the ultrafiltration stage; and the initial concentration state indicators are used for subsequent permeate analysis algorithms to process the composition of lost substances. Furthermore, the step of determining the list of key flavor small molecules by processing the composition of lost substances using a permeate analysis algorithm based on the initial concentration state index includes: processing the composition of lost substances during ultrafiltration using a permeate analysis algorithm based on the initial concentration state index to determine the list of key flavor small molecules from the composition of lost substances. The list of key flavor small molecules includes specific organic acids and other flavor-related small molecule compounds. The list of key flavor small molecules is used to determine whether the concentration of specific organic acids is lower than a preset threshold. Furthermore, the step of recovering the corresponding compound from the permeate through the ultrafiltration module to obtain a supplementary flavor extract if the concentration of the specific organic acid in the list of key flavor molecules is lower than a preset threshold includes: determining if the concentration of the specific organic acid in the list of key flavor molecules is lower than a preset threshold, then starting the ultrafiltration module to recover the corresponding compound from the permeate, and obtaining a supplementary flavor extract from the recovered permeate. The supplementary flavor extract is rich in the lost flavor molecules. The supplementary flavor extract is used for subsequent extraction of aromatic compound concentration data. Furthermore, the step of extracting aromatic compound concentration data from the supplementary flavor extract and adjusting the mixing ratio of the extract and concentrated base material using a flavor compensation algorithm to determine the optimized flavor intensity equilibrium point includes: extracting aromatic compound concentration data from the supplementary flavor extract; adjusting the mixing ratio of the supplementary flavor extract and concentrated base material using a flavor compensation algorithm based on the aromatic compound concentration data; and determining the optimized flavor intensity equilibrium point using the flavor compensation algorithm. The optimized flavor intensity equilibrium point represents that the flavor intensity of the mixed base material has reached a balanced state. Furthermore, the step of obtaining the mixed base sample corresponding to the optimized flavor intensity equilibrium point and performing secondary membrane filtration verification on the protein retention rate to determine the final protein flavor balance ratio includes: obtaining the mixed base sample corresponding to the optimized flavor intensity equilibrium point, performing secondary membrane filtration verification on the mixed base sample on the protein retention rate, and determining the final protein flavor balance ratio from the secondary membrane filtration verification result. The final protein flavor balance ratio simultaneously meets the protein retention rate requirement and the flavor intensity equilibrium requirement. Furthermore, the step of obtaining a stable quality output base material sequence by processing production batch data through continuous fluid simulation based on the final protein flavor balance ratio includes: processing production batch data using continuous fluid simulation based on the final protein flavor balance ratio to obtain a stable quality output base material sequence from the processed production batch data; the stable quality output base material sequence maintains consistent protein content and flavor substance distribution across multiple batches; and the stable quality output base material sequence is used for subsequent judgment of fluctuations in layering parameters. Furthermore, the step of using a feedback adjustment algorithm to correct the concentration parameter settings and obtain the refining production process configuration if the fluctuation of the hierarchical parameter in the stable quality output base material sequence exceeds a threshold includes: determining if the fluctuation of the hierarchical parameter in the stable quality output base material sequence exceeds a threshold, then initiating a feedback adjustment algorithm to correct the ultrafiltration parameter settings, obtaining the refining production process configuration from the corrected ultrafiltration parameter settings, wherein the refining production process configuration reduces the fluctuation of the hierarchical parameter, and the refining production process configuration is used for subsequent key operation node data extraction and batch simulation verification of overall quality indicators.

[0008] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a production method for optimizing flavor and protein balance during the milk base concentration process. Addressing the contradiction between protein retention and flavor loss in milk base processing, it proposes a logically linked solution. The method collects base sample data during the concentration process to obtain initial concentration state indicators. Combined with a permeate analysis algorithm, it determines a list of key flavor small molecules. When the concentration of a specific organic acid falls below a threshold, an ultrafiltration module is used to recover the permeate and prepare a supplementary flavor extract. Subsequently, a flavor compensation algorithm is used to adjust the mixing ratio, optimize the flavor intensity equilibrium point, and verify the final protein-flavor balance ratio through secondary membrane filtration. Furthermore, continuous fluid simulation and feedback regulation algorithms ensure batch stability and layered flavor parameter control, ultimately forming a refined production process configuration. This invention achieves a dual balance between protein retention and flavor intensity, significantly improving the consistency of milk base quality, optimizing production efficiency and product taste, and providing innovative technical support for dairy processing. Attached Figure Description

[0009] Figure 1 This is a flowchart of a flavor balance optimization method in the milk base concentration process of the present invention.

[0010] Figure 2 This is a schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0011] Figure 3 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0012] Figure 4 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0013] Figure 5 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0014] Figure 6 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0015] Figure 7 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0016] Figure 8 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0017] Figure 9 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0018] Figure 10 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0019] Figure 11 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0020] Figure 12 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0021] Figure 13 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0022] Figure 14 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention.

[0023] Figure 15 This is another schematic diagram of a flavor balance optimization method in the milk base concentration process of the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] like Figures 1-15 This embodiment of a method for optimizing flavor balance during the concentration process of a milk base material may specifically include: Initial concentration state indicators are obtained by collecting base material sample data during the concentration process; based on the initial concentration state indicators, the composition of lost substances is processed using a permeate analysis algorithm to determine a list of key flavor small molecule components.

[0026] Step S1: Obtain initial concentration state indicators by collecting base material sample data during the concentration process.

[0027] In one embodiment, step S1 specifically includes: step S11, collecting base material samples periodically during the ultrafiltration process, for example, collecting a 50 ml sample of base material every 10 minutes; step S12, performing protein content detection and small molecule flavor compound distribution analysis on the collected samples to obtain protein concentration values ​​and small molecule compound parameters, thereby forming initial concentration state indicators.

[0028] For example, this acquisition method ensures the real-time nature of the data, which is beneficial for the accurate assessment of the subsequent concentration state.

[0029] Step S2: Based on the initial concentration state index, the permeate analysis algorithm is used to process the composition of lost substances and determine the list of key flavor small molecules.

[0030] In one embodiment, step S2 specifically includes: Step S21, extracting compositional data of lost substances, such as protein retention rate and organic acid concentration, from initial concentration state indicators; Step S22, applying a permeate analysis algorithm based on high-performance liquid chromatography data processing to first separate compounds in the permeate and then quantify the concentration of each component, for example, calculating the content of a specific organic acid by peak area integration; Step S23, screening compounds with concentrations higher than 0.5 mg / L based on the quantification results to form a list of key flavor small molecule components.

[0031] For example, in the production of yogurt base, if the initial concentration index shows a protein concentration of 8%, the permeate analysis algorithm can identify key components such as lactic acid and citric acid in the list. This process improves the accuracy of flavor recovery and is beneficial to product quality stability.

[0032] In one embodiment, the extended application of the permeate analysis algorithm in step S2 includes: step S211, adjusting the algorithm parameters for different batches of base material, for example, setting a threshold of 1.0 mg / L in a high-concentration base material scenario; and step S212, verifying the algorithm output by combining it with mass spectrometry analysis to ensure the accuracy of the list.

[0033] For example, in determining the list of key flavor small molecules, if the algorithm shows that the concentration of a specific organic acid is 0.8 mg / L, the list will preferentially include that component. This expansion is beneficial for adapting to varying production environments and improving the overall flavor balance.

[0034] In one embodiment, for another scenario of step S2, step S221 involves using a permeate analysis algorithm to process the lost substances containing polysaccharide matrix and quantify the distribution of aromatic compounds. Step S222 involves excluding impurity components with concentrations below a threshold when generating the list.

[0035] For example, when optimizing yogurt base, this method ensures the accurate identification of key components, which helps reduce flavor loss and improve the sensory quality of the product.

[0036] If the concentration of a specific organic acid in the list of key components of flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered through the ultrafiltration module to obtain a supplementary flavor extract.

[0037] In one embodiment, if the concentration of a specific organic acid in the list of key components of flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered by the ultrafiltration module to obtain a supplementary flavor extract. Specifically, step S1 is to check whether the concentration of a specific organic acid in the list of key components of flavor small molecules is lower than a preset threshold, wherein the preset threshold is set to 0.5% according to the initial concentration state index to ensure flavor balance.

[0038] Step S11: Extract the concentration data of specific organic acids, such as lactic acid, from the list processed by the permeate analysis algorithm and compare it with the threshold.

[0039] Step S12: If the concentration is below the threshold, the ultrafiltration module is activated. This module uses a semi-permeable membrane material to separate small molecule compounds. The membrane pore size is controlled in the range of 0.1 nanometers to 1 nanometer to selectively retain the corresponding organic acids.

[0040] Step S13: The permeate is processed by the module to recover the corresponding compounds and form a supplementary flavor extract, which increases the concentration of organic acids to 1.2 times the initial level.

[0041] In one possible implementation, this step is applied to the ultrafiltration process of yogurt base material. The recovery efficiency of the ultrafiltration module is verified by continuous fluid simulation. The membrane material is a polyamide composite membrane. Its separation principle is based on the difference in solute molecule size and charge. For example, lactic acid molecules are concentrated and recovered because of their small size, which compensates for the flavor loss in the concentrated base material. The beneficial effect is to improve the flavor intensity balance of the final base material and avoid the product having a monotonous taste.

[0042] For example, when processing Greek yogurt base, if the citric acid concentration in the list is below 0.4%, the module recovers citric acid from the permeate, and the extract concentration reaches 0.6%. After mixing, the protein flavor balance ratio is optimized, the fluctuation of product layering parameters is reduced to within the threshold, and the stable quality output is improved.

[0043] In one embodiment, the recovery process in step S1 is extended to multi-stage reverse osmosis, firstly by pre-filtering the permeate to remove macromolecular impurities, then by separating organic acids in the main module, and finally by adjusting the pH to 4.5 in the concentration stage to enhance compound stability.

[0044] Step S111: Pre-filtration uses a microfiltration membrane to remove suspended solids, ensuring that the permeate purity reaches 95%.

[0045] In step S112, the main module applies a pressure of 2-5 bar to push the permeate through the membrane, achieving a recovery rate of 70%. The resulting extract is then used directly for subsequent mixing.

[0046] Specifically, this multi-stage approach is effective in high-concentration ultrafiltration scenarios, such as in the production of whole-fat yogurt. If the acetic acid concentration is below 0.3%, the recovered extract is used to replenish the flavor. The protein retention rate of the optimized mixed base material is verified to reach 98% through secondary membrane filtration. The beneficial effect is that the refining production process configuration is more stable and the overall quality indicators are improved.

[0047] For example, for low-fat yogurt, if the propionic acid concentration is below the threshold of 0.2%, the propionic acid in the extract is recovered in multiple stages to 0.4%, and the feedback adjustment algorithm corrects the ultrafiltration parameters, controlling the layering fluctuation within 5%, thus forming an optimized yogurt base material generation path.

[0048] In one embodiment, the threshold in step S1 is dynamically adjusted and calculated in real time based on the initial concentration state index, for example, threshold = initial concentration * 0.8, and the recovery module integrates sensors to monitor the compound distribution.

[0049] Step S121: The sensor collects real-time data of the permeate and inputs it into the comparison module to determine if it is below the threshold.

[0050] In step S122, during recycling, the module switches to a targeted membrane, such as a hydrophilic membrane for lactic acid, to improve selectivity to 85%.

[0051] It should be noted that this dynamic adjustment is applied in batch production, combined with continuous fluid simulation processing data, to ensure the optimization of the ratio of supplementary extract to concentrated base material. The beneficial effect is to reduce flavor fluctuations and achieve a stable quality output base material sequence.

[0052] For example, in fruit-flavored yogurt base, if the malic acid concentration is below 0.6%, the extract is dynamically recycled, the mixing ratio is adjusted to 1:10, the protein retention rate is verified twice to be 95%, the final formula supports the refining process, and the quality index simulation verification path is optimized.

[0053] The concentration data of aromatic compounds extracted from the supplementary flavor extract were used to adjust the mixing ratio of the extract and the concentrated base material using a flavor compensation algorithm to determine the optimal flavor intensity balance point.

[0054] In one embodiment, the concentration data of aromatic compounds extracted from the supplementary flavor extract is used to adjust the mixing ratio of the extract and the concentrated base material using a flavor compensation algorithm to determine the optimal flavor intensity equilibrium point. Specifically, this includes step S1, extracting the concentration data of aromatic compounds from the supplementary flavor extract, wherein the concentration data of aromatic compounds is obtained by chromatographic analysis, including the specific concentration values ​​of phenolic and ester compounds.

[0055] Step S2: The flavor compensation algorithm is used to adjust the mixing ratio of the extract and the concentrated base. The flavor compensation algorithm is based on the least squares method to fit the concentration data and the preset flavor curve to calculate the mixing ratio to balance acidity and sweetness.

[0056] Step S3: Determine the equilibrium point of flavor intensity after optimization. The equilibrium point is determined by comparing the flavor intensity index of the mixed sample with the threshold range to ensure the uniform distribution of protein and flavor substances.

[0057] In one embodiment, in step S1, aromatic compound concentration data is extracted from the supplemental flavor extract.

[0058] Specifically, the supplemental flavor extract is derived from the permeate recovered by the ultrafiltration module. It is then subjected to high-performance liquid chromatography analysis to separate and quantify the concentrations of aromatic compounds, such as benzoic acid and vanillin.

[0059] For example, in a yogurt production scenario, the volume of the extract sample is 50 ml, and the analysis results show that the concentration of benzoic acid is 0.2 mg / L. This can accurately capture key flavor components, which helps to accurately input subsequent compensation algorithms and avoid product quality degradation caused by flavor loss.

[0060] In one embodiment, step S2 uses a flavor compensation algorithm to adjust the mixing ratio of the extract and the concentrated base material. Specifically, step S21 involves constructing a concentration vector based on the extracted aromatic compound concentration data. The concentration vector includes numerical arrays of multiple compounds, such as phenol concentration arrays and ester concentration arrays. The similarity to the ideal flavor vector is calculated through the vector inner product.

[0061] Step S22: Optimize the mixing ratio using the least squares method, where the least squares method is solved by minimizing the sum of squared residuals between the concentration vector and the target flavor curve.

[0062] For example, the target flavor curve can be set as a linear function y = ax + b, where a is the slope representing the rate of concentration change and b is the intercept representing the baseline intensity. The ratio can be iteratively adjusted until the residual is less than 0.01. This can achieve precise flavor compensation and improve the consistency of the taste of yogurt base.

[0063] Step S23: Output the adjusted mixing ratio, such as 20% extract and 80% concentrated base, to ensure that flavor molecules are evenly integrated into the protein matrix.

[0064] For example: In one possible implementation, for strawberry-flavored yogurt, the extracted aromatic compound concentration data showed a low ester concentration. The mixing ratio was adjusted to 25% extract and 75% concentrated base through a flavor compensation algorithm. This enhanced the fruity aroma while maintaining a protein retention rate of over 90%, which is beneficial to product stability and market competitiveness.

[0065] In one embodiment, determining the optimized flavor intensity equilibrium point in step S3 specifically includes step S31, obtaining the flavor intensity index of the adjusted mixed base, wherein the flavor intensity index is obtained through sensory evaluation combined with instrument measurement, such as using an electronic nose to detect the peak value of volatile compounds.

[0066] Step S32: Compare the flavor intensity index with the preset balance threshold range. The balance threshold range is an intensity value between 8 and 12. If it is within the range, it is determined to be the balance point. Otherwise, it iterate back to step S2 to adjust the ratio.

[0067] Step S33: Confirm the equilibrium point corresponding to the mixed base sample for subsequent protein retention rate verification. This forms a complete chain from concentration extraction to equilibrium judgment, ensuring the flavor and protein balance of the final yogurt base.

[0068] For example: In one possible implementation, for plain yogurt, the optimized flavor intensity index is 10, which is within the threshold range and is determined to be the equilibrium point. This can bring about the technical effect of uniform flavor distribution, reduce production batch fluctuations, and improve the stability of the overall quality output base sequence.

[0069] The final protein flavor balance ratio was determined by verifying the protein retention rate through secondary membrane filtration of the mixed base sample corresponding to the optimized flavor intensity equilibrium point.

[0070] In one embodiment, step S5, obtaining the mixed base sample corresponding to the optimized flavor intensity equilibrium point, and performing secondary membrane filtration verification on the protein retention rate to determine the final protein flavor balance ratio, specifically includes step S51, collecting the mixed base sample corresponding to the optimized flavor intensity equilibrium point, wherein the mixed base sample is a mixture of extract and concentrated base adjusted by the aforementioned flavor compensation algorithm, containing a specific proportion of protein and small molecule flavor substances.

[0071] Step S52: Based on the collected mixed base material sample, perform a secondary membrane filtration operation to verify the protein retention rate. The secondary membrane filtration uses ultrafiltration membrane technology, with the membrane pore size controlled within the range of 0.01 to 0.1 micrometers to ensure that large molecular proteins are retained while small molecular substances partially permeate.

[0072] Step S53: Based on the data of the retained liquid and permeate after secondary membrane filtration, calculate the protein retention rate and compare it with a preset threshold to determine the final protein flavor balance ratio.

[0073] Specifically, this implementation method is applied to the yogurt base production process. First, in step S51, a sample of 10 to 50 ml is extracted from the optimized mixed base to ensure that the sample represents the entire batch. Through this collection, the flavor balance point data from the previous step can be directly linked to form a continuous chain of production parameters.

[0074] In one possible implementation, the secondary membrane filtration verification process in step S52 needs to be elaborated in detail because it involves precise control of protein retention. The protein retention rate is calculated by dividing the protein concentration in the retention solution by the initial sample protein concentration multiplied by 100%, where the protein concentration is determined by the Kjeldahl method, which involves digesting the sample, distilling ammonia, titrating to calculate the nitrogen content, and then multiplying by 6.38 to convert the protein value.

[0075] For example, during filtration, a pressure of 0.5 to 2 bar is applied to push the base material through the membrane. The retainer collects large protein molecules, while some small flavor molecules, such as organic acids, may be lost. However, the validation aims to confirm that the retention rate is not less than 85% to maintain the texture stability of the yogurt. This approach yields beneficial results, such as improved product consistency and flavor persistence, and avoids protein loss due to over-filtration.

[0076] For example, the membrane filtration parameters in step S52.

[0077] In one embodiment, if the initial protein concentration is 3.5% and the concentration of the retained solution after secondary filtration is 3.2%, the retention rate is 91.4%, which is higher than the threshold of 80%, indicating that the ratio is appropriate. If it is lower than the threshold, the mixing ratio is adjusted and re-verified. This multi-round verification logic ensures the balance between protein and flavor and reduces production fluctuations.

[0078] It should be noted that in step S53, when determining the final protein flavor balance ratio, the aforementioned protein retention rate data is used. If the retention rate is between 85% and 95%, the current mixing ratio is directly output as the balance ratio, such as a 1:1 ratio of extract to base material; otherwise, iterative adjustments are made until the conditions are met. The beneficial effect of this method is to optimize the overall quality of the yogurt base material and enhance its market competitiveness.

[0079] In one embodiment, the verification scenario of extended step S52 is considered for different types of yogurt, such as high-protein Greek yogurt, where the initial sample protein concentration is set at 5%, a 0.05-micron membrane is used for secondary filtration at a pressure of 1 bar, and the retention rate after verification is 92%. Based on the flavor intensity data, the ratio is determined to be 20% of the extract. Another scenario is low-fat yogurt with a protein concentration of 2.8% and a retention rate of 88%. The ratio is adjusted to 15% of the extract. These examples support the flexibility of verification from the perspective of protein concentration, ensuring applicability in the field of yogurt production.

[0080] For example, regarding the determination of the balanced ratio in step S53, in scenarios with high flavor requirements, if the retention rate is 90%, the balanced ratio emphasizes flavor enhancement, resulting in a richer yogurt taste. In standard production, however, a balanced ratio is achieved when the retention rate is 85%, reducing the need for subsequent adjustments and improving efficiency. These examples highlight the logical chain of the steps, directly leading from validation data to the ratio output, closely adhering to the optimized yogurt base material generation path.

[0081] S105. Based on the final protein flavor balance ratio, a stable quality output base material sequence is obtained by processing the production batch data through continuous fluid simulation.

[0082] Based on the final protein flavor balance ratio, a stable quality output base sequence is obtained by processing production batch data through continuous fluid simulation.

[0083] In one embodiment, a stable quality output base material sequence is obtained by processing production batch data through continuous fluid simulation based on the final protein flavor balance ratio. Specifically, step S61 involves collecting production batch data, including base material flow rate, temperature distribution, and mixing ratio parameters. These data are derived from the measured values ​​of the aforementioned optimized mixed base material samples.

[0084] Step S62: The collected data is divided into grids using the finite volume method to establish a fluid flow model. The finite volume method divides the fluid domain into a finite number of control volumes and solves for the variable changes within each volume using integral conservation equations.

[0085] Step S63: Input the final protein flavor balance ratio into the model as boundary conditions, simulate the continuous flow process, and calculate the uniformity of protein and small molecule flavor substances distribution in the batch. The boundary conditions include the concentration gradient at the inlet and the pressure setting at the outlet.

[0086] Step S64: Iteratively solve the Navier-Stokes equations to obtain the velocity field and concentration field. The Navier-Stokes equations describe the conservation of momentum in fluid motion. The steady-state solution is approximated step by step using numerical discretization methods such as finite difference.

[0087] Step S65: Extract stable quality output base material sequence based on simulation results, including time series concentration values ​​for each batch, to ensure that the fluctuation of flavor intensity and protein retention rate is less than the preset threshold of 0.5%.

[0088] Specifically, this implementation method can be applied to small and medium-sized yogurt production lines, reducing quality fluctuations and improving product consistency by simulating and optimizing batch processing time.

[0089] In one possible implementation, the production batch data collected in step S61 can be adjusted for different fermentation times. For example, in a 24-hour fermentation cycle, the base material flow rate is set to 5 liters per minute, and the temperature distribution is controlled at 40-42 degrees Celsius, so that subsequent simulations can more accurately reflect actual production.

[0090] For example, the mesh generation in step S62 can use an unstructured mesh to adapt to complex pipe geometry and ensure that the simulation accuracy converges when the error is less than 1%.

[0091] Understandably, when inputting boundary conditions in step S63, if the protein balance ratio is 15% protein concentration and 0.8% flavor substances, the simulation shows that the batch end uniformity is improved by 20%, which is beneficial to stabilizing the taste of yogurt.

[0092] Preferably, in the iterative solution process of step S64, a relaxation factor of 0.7 is used to accelerate convergence, reduce the number of calculation iterations to less than 500, and improve efficiency.

[0093] In one embodiment, after the sequence is extracted in step S65, if the fluctuation exceeds the threshold, the input parameters can be adjusted to form a closed-loop optimization, ensuring that the final base material sequence has stable quality when used for large-scale production.

[0094] For example, in a protein retention rate verification scenario, simulation results show that the batch yield corresponding to the flavor intensity equilibrium point in the sequence increases by 15%, reducing waste generation.

[0095] Specifically, for innovative applications of continuous fluid simulation in the field of yogurt base materials, this method can achieve smooth batch-to-batch transitions, avoid flavor loss, and improve overall production stability.

[0096] S106. If the fluctuation of the hierarchical parameter in the stable quality output base material sequence exceeds the threshold, a feedback adjustment algorithm is used to correct the concentration parameter setting to obtain the refining production process configuration.

[0097] In one embodiment, in step S1, if the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds a threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration.

[0098] Step S11: Obtain the hierarchical parameter data in the stable quality output base material sequence, and calculate the comparison result of its fluctuation value with the preset threshold.

[0099] Specifically, by statistically calculating the hierarchical parameters of each batch of base material in the sequence, such as calculating the standard deviation as a fluctuation value, if it exceeds a threshold such as 0.5, subsequent corrections are triggered. This comparison ensures timely intervention in the production process, which is beneficial to maintaining the stability of base material quality.

[0100] In one embodiment, step S2 involves using a feedback adjustment algorithm to correct the concentration parameter settings.

[0101] Step S21: Based on the proportional-integral-derivative controller as the feedback adjustment algorithm, input the current level sensing parameter fluctuation value and the target value, and calculate the correction amount.

[0102] The proportional-integral-derivative (PID) controller is a classic control algorithm. Its principle is to generate a control output by using a proportional term to respond to the current error, an integral term to eliminate the steady-state error, and a derivative term to predict error changes. Here, the fluctuation of the hierarchical sensing parameter is used as the error signal, and the target value is a stable range below a threshold. The controller output is used to adjust condensed parameters such as pressure or temperature settings. Applying this algorithm can effectively reduce fluctuations and result in more stable production output.

[0103] Step S22: Update the concentration parameter settings according to the correction amount, for example, adjust the membrane pressure in the ultrafiltration process from the initial 15 bar to the corrected 12 bar.

[0104] Specifically, in yogurt base production, if fluctuations in the layering parameters stem from uneven concentration, the feedback adjustment algorithm will gradually optimize the parameters based on the integral portion of historical data to ensure a balanced distribution of proteins and flavor compounds. This adjustment process forms a closed-loop control, improving the robustness of the overall process.

[0105] In one embodiment, the extension of step S2 considers parameter adjustments under different batch scenarios.

[0106] For example, in small-batch production, the fluctuation threshold is set to 0.3, the controller proportional coefficient is 0.8, the integral time is 2 minutes, and the derivative time is 0.5 minutes. After the correction, the concentration parameters reduce the layering parameter from fluctuation of 0.4 to 0.2, thereby improving the consistency of the base material's taste.

[0107] In another embodiment, in large-scale continuous production, the threshold is set to 0.6, and the controller parameters are adjusted to a proportional coefficient of 1.2, an integral time of 3 minutes, and a derivative time of 1 minute. Based on the feedback of flavor molecule distribution, the concentration temperature is corrected from 40°C to 38°C, which reduces the quality deviation caused by the fluctuation of layering and is beneficial to the stable output of large-scale yogurt base.

[0108] In one embodiment, step S3 involves generating a refining production process configuration based on the modified concentration parameter settings.

[0109] Specifically, the updated parameters are integrated into the process to form a configuration that includes operation nodes and batch verification, ensuring the optimization of the final yogurt base material generation path.

[0110] For example, in validation, the configuration is applied to simulated batches to confirm that the layering parameters are stable within the threshold, such as the fluctuation value being kept below 0.4, thereby supporting the continuous production of protein flavor balance. Obtaining this configuration directly improves the reliability of the overall quality indicators.

[0111] Key operational node data were extracted from the refining production process configuration to conduct batch simulation verification of overall quality indicators and determine the optimal emulsion base material generation path.

[0112] In one embodiment, step S1 involves extracting key operation node data from the refining production process configuration to perform batch simulation verification for overall quality indicators and determine the optimized milk base material generation path. Specifically, step S11 involves identifying core parameters in the refining production process configuration, such as ultrafiltration parameter settings and feedback adjustment results, and extracting key operation node data from them. These data include protein retention rate, flavor intensity equilibrium point, and layering parameter fluctuation value for each node.

[0113] Step S12: Based on the extracted key operation node data, a batch simulation model is constructed. The Monte Carlo simulation method is used to handle the data variation of multiple production batches. The Monte Carlo simulation method is a statistical method based on random sampling. It generates a large number of random samples to simulate the impact of uncertainty factors, thereby evaluating the distribution of overall quality indicators. For example, multiple iterative calculations are performed on the protein flavor balance ratio to verify quality stability.

[0114] Step S13: For the overall quality indicators in the simulation results, such as protein content uniformity and flavor small molecule distribution parameters, threshold comparison is performed. If the fluctuation exceeds the preset threshold, the node parameters are adjusted to determine the optimized milk base material generation path. This path includes a continuous sequence from initial concentration to final output.

[0115] Specifically, the path determination method provided by this implementation can be applied to the production process of yogurt base materials. It uses a stable quality output base material sequence obtained from batch data processing via continuous fluid simulation as a foundation, and data extracted from the refining production process configuration ensures the accuracy of the simulation. The Monte Carlo simulation method involves defining the probability distribution of input variables, such as the normal distribution of protein retention rate, and then generating random samples for calculation. For example, assuming a batch has 100 nodes, 1000 iterations are simulated to evaluate the confidence interval of quality indicators, thereby identifying the optimal path. The beneficial effects of this method are improved production efficiency, reduced resource consumption in actual batch trials, and ensured flavor and protein balance in the milk base material.

[0116] In one possible implementation, the batch simulation verification in step S12 is further refined. In step S121, the extracted key operation node data is input into the simulation environment, and initial conditions such as a preset threshold for a specific organic acid concentration are set.

[0117] Step S122: Use the Monte Carlo simulation method to generate random batch scenarios, such as simulating ultrafiltration processes at different temperatures, and calculate the overall quality index value for each scenario.

[0118] Step S123: Summarize the simulation results and calculate the average quality score. If the score is lower than the threshold, it is marked as a node that needs optimization, thereby supporting the path determination in step S13.

[0119] For example, in the optimization of the milk base material production path, assuming that a production batch involves 10 key nodes, Monte Carlo simulation generates 500 random samples. Each sample considers the variation of key flavor small molecule components, such as the uniform distribution of organic acid concentration from 0.5% to 1.5%. Through iterative calculation, the fluctuation range of quality indicators is found to be ±0.2. The beneficial effect of this simulation is to discover potential unstable factors early and ensure the stability of the final path.

[0120] In one embodiment, for the overall process of step S1, different batch-scale scenarios are considered, such as small laboratory batches and large-scale industrial batches. In small batches, the extracted key operational node data focuses on a small number of samples, such as the protein retention rate of 5 nodes, and the quality indicators are validated through Monte Carlo simulation to determine the adjustment ratio of the path-emphasized flavor compensation algorithm. In large-scale batches, the node data is expanded to 50, and the number of simulation iterations is increased to 2000 to cover more variations, thereby obtaining a more robust optimized milk base material generation path. The beneficial effect of this diversity implementation is that it adapts to different production environments and improves the versatility of the path.

[0121] For example, in practical applications, the extraction process in step S11 first selects nodes from the refining production process configuration, such as the recovery data of the ultrafiltration module, and then calculates the average value as the key data. The beneficial effect of this extraction is that it simplifies the input preparation for subsequent simulations and ensures the consistency of the data.

[0122] In one possible implementation, the path determination in step S13 is based on the results of simulation verification, generating a path sequence diagram where each node is labeled with optimization parameters, such as the adjusted mixing ratio, and the final path aims to minimize quality fluctuations. The advantage of this approach is that it visualizes the optimization process, facilitates production implementation, and closely adheres to the core requirement of optimizing the emulsion base material generation path.

[0123] The initial concentration state index is obtained by collecting base material sample data during the concentration process, including: collecting base material sample data during the ultrafiltration process to obtain protein content and small molecule flavor substance distribution parameters; obtaining the initial concentration state index from the protein content and small molecule flavor substance distribution parameters; the initial concentration state index characterizes the protein retention and small molecule flavor substance loss of the concentrated base material during the ultrafiltration stage; the initial concentration state index is used for subsequent permeate analysis algorithm to process the composition of lost substances.

[0124] In one embodiment, step S1 obtains the initial concentration state index by collecting base material sample data during the concentration process, including the following sub-steps.

[0125] Step S11: Collect data from the base material sample during the ultrafiltration process to obtain protein content and small molecule flavor substance distribution parameters.

[0126] Specifically, during the operation of the ultrafiltration unit, samples are periodically extracted from the substrate stream, and a spectrometer is used to measure the protein content and distribution parameters of small molecule flavor substances such as organic acids. These parameters reflect the compositional changes of the substrate under membrane filtration.

[0127] Step S12: Obtain the initial concentration state index from the protein content and small molecule flavor substance distribution parameters.

[0128] In one possible implementation, the collected protein content data and small molecule flavor substance distribution parameters are input into a preset weighted average calculation formula to calculate the value of the initial concentration state index. The weighted average calculation formula is based on the ratio of protein retention rate to small molecule loss rate. For example, by dividing the protein content by the ratio of the initial base protein concentration and subtracting the distribution ratio of small molecule flavor substances in the permeate, a quantitative index value is obtained. This process ensures that the index accurately reflects the concentration effect.

[0129] Step S13, the initial concentration state index characterizes the protein retention and small molecule flavor substance loss of the concentrated base material during the ultrafiltration stage.

[0130] For example, the indicator is expressed in numerical form. For instance, an indicator value greater than 0.8 indicates high protein retention and low loss of flavor substances, while a value lower than 0.8 indicates that the ultrafiltration parameters need to be adjusted.

[0131] Step S14: The initial concentration state index is used in subsequent permeate analysis algorithms to process the composition of lost substances.

[0132] In one embodiment, the initial concentration state index is used as input data to the permeate analysis algorithm. The algorithm uses principal component analysis to decompose the composition of lost substances. Principal component analysis is a statistical method that identifies key lost components by calculating the covariance matrix of variables and extracting principal components, thereby determining a list of key flavor small molecule components.

[0133] In one embodiment, the process of obtaining the initial concentration state index from protein content and small molecule flavor compound distribution parameters in step S12 is extended to the production of yogurt base materials. The collected protein content is the amount of protein per liter of base material, for example, 25 g / L measured in milk ultrafiltration. The small molecule flavor compound distribution parameters include the concentration distribution of lactic acid and citric acid, such as 0.5% lactic acid and 0.3% citric acid. Using a weighted average calculation formula, the protein content is multiplied by the retention coefficient of 0.9 and the flavor compound loss coefficient of 0.2 is subtracted to obtain an initial concentration state index value of 0.75. This index value is used to evaluate the quality stability of the concentrated base material. The beneficial effect is improved flavor retention efficiency and avoidance of uneven taste caused by protein loss in traditional methods.

[0134] For example, in another scenario, for the production of high-concentration Greek yogurt, the protein content was measured at 35 g / L, and the distribution parameters of small molecule flavor compounds showed uneven distribution of organic acids. Through the calculation in step S12, the weighting coefficients were adjusted to 0.8 for protein and 0.2 for flavor, resulting in an index value of 0.85. This has the beneficial effect of optimizing the protein-flavor balance in the concentration process and reducing the resource consumption of subsequent compensation steps.

[0135] In one embodiment, the initial concentration state index in step S14 is used in the subsequent permeate analysis algorithm, which can be extended in principle. The principal component analysis method first constructs a data matrix, including index values ​​and component concentrations of the permeate sample, and then calculates feature vectors to extract principal components; for example, the first principal component corresponds to the organic acid component with high loss. This has the advantage of accurately identifying lost substances and improving recovery efficiency. In a specific application, if the index value is 0.7, the analysis algorithm outputs that citric acid accounts for 40% of the lost substance composition, thereby guiding the ultrafiltration module to recover the corresponding compound, forming a tight logical chain from data acquisition to substance recovery.

[0136] For example, in the data collection process of step S11, on a continuous production line, an online sensor is used to collect a sample of the base material every 5 minutes to obtain the protein content data as 28 g / L and the small molecule distribution parameter as lactic acid 0.4%. This ensures real-time monitoring and has the beneficial effect of reducing human error and improving production consistency.

[0137] In one embodiment, combined with the characterization effect of step S13, in the case of low-fat yogurt, an initial concentration state index value of 0.6 indicates that the protein retention is good but the loss of small molecule flavor substances is high. This can bring beneficial effects, namely, timely triggering of the flavor compensation algorithm to maintain the product's balanced taste.

[0138] The process involves using a permeate analysis algorithm to process the composition of lost substances based on the initial concentration state index to determine a list of key flavor small molecules. This includes: processing the composition of lost substances during ultrafiltration using a permeate analysis algorithm based on the initial concentration state index to determine a list of key flavor small molecules from the composition of lost substances. The list of key flavor small molecules includes specific organic acids and other flavor-related small molecule compounds. The list of key flavor small molecules is used to determine whether the concentration of specific organic acids is below a preset threshold.

[0139] Step S1: Based on the initial concentration state indicators, the permeate analysis algorithm is used to process the composition of substances lost during ultrafiltration.

[0140] In one embodiment, step S1 specifically includes step S11, collecting a permeate sample during the ultrafiltration process and extracting compositional data of the lost substances from it, such as the initial concentration of specific organic acids and the distribution parameters of small molecule compounds. These data are derived from the initial concentration state index, which is the protein content and the distribution parameters of small molecule flavor substances obtained by collecting the base material sample.

[0141] Step S12: The extracted composition data is quantified using a permeate analysis algorithm. The permeate analysis algorithm is a calculation method based on component separation. It identifies key flavor components by comparing the molecular weight and concentration distribution of the lost substances. For example, the algorithm first inputs the composition data into a threshold filtering model, which classifies compounds with molecular weights below a certain value as small molecule components, and then calculates the relative abundance of each component.

[0142] Step S13: Generate a quantitative characterization of the lost substances based on the processing results, which will be used to determine the list of key flavor small molecules in the subsequent process.

[0143] For example, in the production of yogurt base, if the initial concentration indicators show a high protein content but uneven distribution of small molecule flavor substances, the permeate analysis algorithm can effectively identify the lost organic acids, thereby helping to optimize the concentration process. The beneficial effect of this is to improve the flavor stability of the final base and avoid taste loss caused by ultrafiltration.

[0144] Step S2: Determine a list of key flavor small molecules from the composition of the lost substances.

[0145] In one embodiment, step S2 specifically includes screening flavor-related compounds from the quantitative characterization in step S1 to form a list.

[0146] Step S3, the list of key flavor small molecules includes specific organic acids and other flavor-related small molecule compounds.

[0147] For example, specific organic acids can be lactic acid or citric acid, and other small molecule compounds include volatile aldehydes. These components directly affect the acidity and aroma of yogurt. By including these, the list can be made to fully cover the flavor elements, which is beneficial for accurately judging the quality balance after concentration.

[0148] Step S4: The list of key components of flavor small molecules is used to determine whether the concentration of a specific organic acid is below a preset threshold.

[0149] In one embodiment, step S4 specifically includes step S41, comparing the concentration of a specific organic acid in the list with a preset threshold, for example, setting the threshold to 0.5%. If it is lower than this value, the subsequent recovery process is triggered.

[0150] Step S42: Based on the judgment result, decide whether to use the ultrafiltration module to recover the corresponding compound in the permeate.

[0151] For example: In one possible implementation, for Greek yogurt base, if the list shows a lactic acid concentration of less than 0.5%, the result will guide the generation of supplementary flavor extract, which can enhance the overall flavor intensity. The beneficial effect is to achieve a balanced ratio of protein and flavor, and ultimately optimize the yogurt base generation path.

[0152] In one embodiment, the permeate analysis algorithm in step S1 is extended to different base material scenarios. For example, for high-protein yogurt, the algorithm is adjusted to prioritize compounds with molecular weights in the range of 100-500. In step S11, when collecting samples, the temperature parameter is increased to control the temperature at 4 degrees Celsius to maintain the stability of the substances. In step S12, the algorithm uses a weighted average method to calculate the relative abundance, that is, multiplying the concentration of each component by its flavor contribution coefficient, with the coefficient based on experimental data such as 0.8 for organic acids. After generating the characterization in step S13, it is directly linked to the screening in step S2 to ensure logical continuity. This extension can cover low-temperature processing scenarios, and its beneficial effect is to reduce the further loss of heat-sensitive flavor substances.

[0153] For example, in another scenario, for fruit-flavored yogurt base, the algorithm is extended to include specific identification of fruit acids. When extracting data in step S11, pH monitoring is added, such as pH 4.2. In step S12, the algorithm separates citric acid and other components and calculates their distribution density. This complements the aforementioned high-protein scenario, emphasizing the adaptability of the algorithm. Ultimately, it serves to determine the list of key flavor small molecules for threshold judgment, and the beneficial effect is to improve the quality consistency of different types of yogurt.

[0154] In one embodiment, the list content in step S3 is being expanded to determine the concentration of specific organic acids. For example, the principle is illustrated by the following: organic acids such as lactic acid are key to yogurt fermentation, affecting pH and taste. If the concentration is below a threshold such as 0.4%, it will result in a weak flavor. Including it in the list allows for early identification of the problem. Other small molecules such as acetaldehyde provide aroma. The list integrates these to form a complete flavor profile, which is beneficial because it facilitates subsequent compensation algorithms to adjust the mixing ratio.

[0155] For example, consider the judgment process in step S4.

[0156] In one possible implementation, if the concentration is 0.3% which is below the 0.5% threshold, the system automatically triggers recovery. The analysis process includes comparing the list value with the threshold. The cause is ultrafiltration loss, and the consequence is uneven flavor. By judging, quality fluctuations are avoided. The beneficial effect is to obtain a stable quality output base material sequence, which is closely linked to the yogurt base material generation path.

[0157] If the concentration of a specific organic acid in the list of key flavor molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered through the ultrafiltration module to obtain a supplementary flavor extract. This includes: determining if the concentration of a specific organic acid in the list of key flavor molecules is lower than a preset threshold, then starting the ultrafiltration module to recover the corresponding compound in the permeate, and obtaining a supplementary flavor extract from the recovered permeate. The supplementary flavor extract is rich in the lost flavor molecules. The supplementary flavor extract is used for subsequent extraction of aromatic compound concentration data.

[0158] Step S3: If the concentration of a specific organic acid in the list of key components of flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered through the ultrafiltration module to obtain a supplementary flavor extract.

[0159] In one embodiment, step S3 specifically includes step S31, comparing the concentration value of a specific organic acid in the list of key components of flavor small molecules with a preset threshold, wherein the preset threshold is set to 0.5% to 1.0% according to the flavor standard of yogurt base material, and if the concentration is lower than the threshold, subsequent recycling operation is triggered.

[0160] Step S32: Start the ultrafiltration module. This module uses a semi-permeable membrane material such as a polyamide composite membrane. Apply a pressure of 2-4 MPa to the permeate, allowing the solvent to pass through the membrane while retaining the solute, thereby selectively recovering specific organic acid compounds. The recovery rate of the reverse osmosis process is controlled at 70%-90% to ensure that the concentration of compounds in the extract is increased.

[0161] Step S33: Separate the organic acid-rich component from the recovered permeate to form a supplementary flavor extract. The pH of this extract is adjusted to 3.5-4.5 to maintain flavor stability.

[0162] For example, the supplemental flavor extract is rich in small flavor molecules such as lactic acid or citric acid that are lost during ultrafiltration. These compounds are easily lost with the permeate during ultrafiltration and can be effectively replenished to balance the acidity and aroma in the base material through recycling.

[0163] In one possible implementation, the supplementary flavor extract is used for subsequent extraction of aromatic compound concentration data. The content of compounds such as benzaldehyde or vanillin is determined by high performance liquid chromatography with a data accuracy of 0.01 mg / L. This is beneficial for accurately adjusting the mixing ratio and improving the flavor consistency of the final yogurt base.

[0164] It should be noted that the ultrafiltration module is started based on an automated signal transmission for concentration judgment, avoiding human intervention and ensuring production continuity. This process can reduce flavor loss by 15%-20% and improve product quality stability.

[0165] Specifically, in the production of yogurt base material, if the specific organic acid is lactic acid and the preset threshold is 0.8%, when the concentration in the list is 0.6%, the lactic acid concentration in the recovered extract rises to 1.2%, which is used to compensate for the lack of flavor in the concentrated base material, form a balanced acidity distribution, and help extend the shelf life of the product.

[0166] In one embodiment, citric acid is used as a specific organic acid with a threshold of 0.7%. The recovery pressure is adjusted to 3 MPa, and the recovery rate of citric acid in the extract reaches 85%. When extracting aromatic data, gas chromatography is used to verify the purity of the compound to ensure the accuracy of the input to the flavor compensation algorithm. This can optimize the layering parameters of the base material and reduce the risk of fluctuations exceeding the threshold.

[0167] For example, through this recycling step, the protein flavor balance ratio of the overall production process is more accurate, the fluctuation of the hierarchical parameters of the stable quality output base sequence is controlled within 5%, which is beneficial to the batch simulation verification of the refining production process configuration, and finally determines the optimized yogurt base generation path.

[0168] The process involves extracting aromatic compound concentration data from the supplementary flavor extract, adjusting the mixing ratio of the extract and concentrated base material using a flavor compensation algorithm, and determining the optimized flavor intensity equilibrium point. This includes: extracting aromatic compound concentration data from the supplementary flavor extract, adjusting the mixing ratio of the supplementary flavor extract and concentrated base material using a flavor compensation algorithm based on the aromatic compound concentration data, and determining the optimized flavor intensity equilibrium point using the flavor compensation algorithm. The optimized flavor intensity equilibrium point represents the balanced state of the flavor intensity of the mixed base material.

[0169] Step S1: Extract aromatic compound concentration data from the supplemental flavor extract.

[0170] In one embodiment, step S1 specifically includes sampling the supplementary flavor extract using a chromatographic analysis device, separating aromatic compounds, and measuring their concentration values.

[0171] Step S11 involves pre-treating the supplemental flavor extract, including filtering impurities and temperature control, to ensure the accuracy of the extraction data.

[0172] Step S12: Use a high-performance liquid chromatograph to detect the peak area of ​​aromatic compounds and calculate the concentration data based on the standard curve.

[0173] For example, in a yogurt production scenario, the flavor extract is derived from the permeate recovered from reverse osmosis. If the concentration of aromatic compounds such as phenylethanol needs to be extracted, the extract can be stabilized at 25 degrees Celsius and then injected into the chromatograph. The peak area corresponds to a concentration of 0.5 mg / L. The extracted data can be directly used for subsequent adjustments to ensure the accuracy of flavor recovery.

[0174] Step S2: Based on the aromatic compound concentration data, the mixing ratio of the supplementary flavor extract and the concentrated base is adjusted using a flavor compensation algorithm.

[0175] In one embodiment, step S2 specifically includes inputting the extracted aromatic compound concentration data into a linear regression model, which is a statistical algorithm based on least squares fitting of the relationship between historical concentration and flavor intensity, used to predict the optimal mixing ratio.

[0176] Step S21: Collect historical datasets, including flavor scores and mixing ratio records at different concentrations.

[0177] Step S22: Apply a linear regression model to calculate the regression coefficients, where the model formula is Y = aX + b, Y represents flavor intensity, X represents aromatic concentration, and a and b are obtained by the least squares method.

[0178] Step S23: Adjust the ratio of supplementary flavor extract to concentrated base material according to the calculated coefficient. For example, if the concentration data is 0.5 mg / L, the model output ratio is 1:10.

[0179] Specifically, a linear regression model here refers to an algorithm that describes the relationship between variables by fitting a straight line. The process involves calculating the mean, covariance, and variance of the variables, and then solving for the slope 'a' and intercept 'b'. For example, after fitting 10 sets of historical data, we obtain a=0.2 and b=1.0, thereby adjusting the ratios to compensate for flavor loss. This adjustment can bring beneficial effects, such as improving the flavor consistency of yogurt base and reducing production fluctuations.

[0180] For example, in one possible implementation, for extracts of low-concentration aromatic compounds, the model adjusts the ratio to 1:15 to ensure that the flavor intensity after mixing is not lower than the threshold of 8 points; while in high-concentration scenarios, the ratio is adjusted to 1:5 to avoid excessive flavor superposition. This supports the robustness of the algorithm from multiple parameter perspectives.

[0181] In one embodiment, considering different types of yogurt, such as Greek yogurt, with a concentration of 0.3 mg / L, the model calculated a ratio of 1:12, optimizing the balance of acidity and aroma.

[0182] For example, extending from the core solution, if the initial flavor of the base material is weak, the algorithm prioritizes increasing the ratio of the extract to 1:8, resulting in a flavor enhancement effect; in an optional embodiment, the ratio is finely adjusted by 0.1 units in conjunction with the temperature parameter, further enriching the diversity of the solution.

[0183] Step S3: Determine the equilibrium point of flavor intensity after optimization using a flavor compensation algorithm.

[0184] In one embodiment, step S3 specifically includes using the same linear regression model to evaluate the flavor intensity score of the adjusted mixture, and determining the equilibrium point when the score reaches a preset equilibrium threshold.

[0185] Step S31: Simulate the interaction of the mixed compounds and calculate the overall flavor intensity score.

[0186] Step S32: Compare the score with the threshold. If the threshold is 7.5 points, the equilibrium point is confirmed.

[0187] Step S33: Record the proportion and concentration data corresponding to the equilibrium point for subsequent verification.

[0188] Specifically, the judgment process involves iterative calculations. For example, if the initial intensity score is 6.0, it can be adjusted to 7.8 through model iteration to achieve equilibrium. Here, the equilibrium point refers to a state where the flavor intensity is evenly distributed across multiple dimensions such as acidity, sweetness, and aroma. This judgment can bring beneficial effects, such as improving product quality stability and avoiding waste caused by flavor deviations.

[0189] For example, in the optimization of yogurt base, if the strength score is 7.2 at a mixing ratio of 1:10, which does not reach the threshold, it is iterated to 1:9, and the strength score rises to 7.6, confirming the equilibrium point; on the other hand, if the initial concentration is high, it quickly reaches 8.0, supporting the efficiency of the algorithm.

[0190] In one embodiment, for fruit-flavored yogurt, the balance point judgment considers an additional fruit flavor factor, with a threshold set at 8.0 points, and confirmation is made after 3 iterations, thus expanding the application scenarios.

[0191] For example, in the core solution, the equilibrium point ensures the balance between protein and flavor; in the optional embodiment, sensory simulation is added to enrich the judgment dimensions and support the integrity of flavor optimization from multiple directions.

[0192] Step S4, the optimized flavor intensity equilibrium point represents the balanced state of the flavor intensity of the mixed base material.

[0193] In one embodiment, step S4 characterizes the flavor intensity balance of the mixed base by recording equilibrium point data, which is used to guide the final protein flavor balance ratio.

[0194] The process of obtaining the mixed base sample corresponding to the optimized flavor intensity balance point and performing secondary membrane filtration verification on the protein retention rate to determine the final protein flavor balance ratio includes: obtaining the mixed base sample corresponding to the optimized flavor intensity balance point and performing secondary membrane filtration verification on the mixed base sample on the protein retention rate; determining the final protein flavor balance ratio from the secondary membrane filtration verification results; and the final protein flavor balance ratio simultaneously meeting the protein retention rate requirement and the flavor intensity balance requirement.

[0195] In one embodiment, the process of obtaining the mixed base sample corresponding to the optimized flavor intensity equilibrium point and performing secondary membrane filtration verification on the protein retention rate to determine the final protein flavor balance ratio includes the following steps.

[0196] Step S1: Obtain the mixed base sample corresponding to the optimized flavor intensity equilibrium point.

[0197] This step ensures that the samples represent the flavor-compensated base state by taking samples directly from the previously adjusted mixing ratio.

[0198] Step S2: Perform secondary membrane filtration verification on the protein retention rate of the mixed base sample.

[0199] Step S21: Prepare a secondary membrane filtration device, use an ultrafiltration membrane with a pore size of 0.1 micrometers to filter the mixed substrate sample, and monitor the changes in protein concentration before and after filtration.

[0200] Step S22: Calculate the protein retention rate by comparing the ratio of protein content in the residue after filtration to the initial content. For example, the retention rate should reach more than 85%.

[0201] In this embodiment, secondary membrane filtration verification can effectively separate small molecules while retaining large protein molecules, ensuring that no additional contaminants are introduced during the verification process.

[0202] Step S3: Determine the final protein flavor balance ratio based on the secondary membrane filtration verification results. The final protein flavor balance ratio simultaneously meets the requirements for protein retention rate and flavor intensity balance.

[0203] Step S31: Analyze the protein retention rate data in the verification results. If the retention rate meets the preset requirements, such as being greater than 90%, adjust the ratio parameters in conjunction with the flavor intensity data.

[0204] Step S32: Verify whether the ratio meets both the protein retention rate requirement and the flavor intensity balance requirement. Ensure the balance between the two through iterative calculation, for example, the protein retention rate is not less than 88% and the flavor intensity deviation is less than 5%.

[0205] For example, in the production of yogurt base material, assuming that the protein content of the initial mixed base material sample is 4.5%, the retention rate is 92% after secondary membrane filtration verification, and the flavor intensity equilibrium point is 7.2, if both meet the requirements, the final ratio is determined to be 1:3 between the extract and the base material, which can improve product stability.

[0206] In one possible implementation, for the secondary membrane filtration validation in step S2, the application of different temperature parameters is considered. For example, filtration at 25 degrees Celsius can reduce the risk of protein denaturation and increase the retention rate to 95%, which helps maintain quality in production during high-temperature seasons.

[0207] Specifically, in step S3, when determining the final ratio, if the verification results show that the protein retention rate is slightly low, such as 87%, the membrane pressure is finely adjusted to 2 bar, and the verification is repeated to meet the requirements, while maintaining a balanced flavor intensity. This will result in higher product consistency and improved taste.

[0208] In one embodiment, regarding the combination of steps S2 and S3, assuming the sample flavor intensity equilibrium point is 6.8, the protein retention rate after secondary filtration is 91%, and the final ratio is adjusted to 15% of the extract, which meets the protein requirements and balances the flavor. This can reduce fluctuations and improve overall efficiency in continuous production batches.

[0209] For example, in another scenario, if the initial organic acid concentration of the base sample is low and the retention rate is 89% after secondary verification, the ratio is set to 1:4 through step S3 iteration, which can simultaneously meet the requirements, avoid sensory defects caused by flavor loss, and thus enhance the layering of yogurt.

[0210] In one possible implementation, after obtaining the sample in step S1, it is directly input into step S2 for verification, which can form a chain to ensure the logical continuity from the equilibrium point to the final ratio. For example, the sample protein data is used as the filtering input, and the output retention rate is used for adjustment in S3.

[0211] Specifically, for innovative aspects such as the detailed process of secondary membrane filtration, membrane filtration involves pressure-driven separation, where large protein molecules are retained by the membrane while small molecules permeate. The retention rate is calculated using the formula (residual protein / initial protein) * 100%. This explanation clarifies the process and allows for precise control of protein content.

[0212] In one embodiment, if the flavor intensity deviation is large when step S3 ensures that the requirements are met simultaneously, step S2 can be back to adjust the membrane pore size to 0.05 micrometers and retested for optimization. This provides multi-faceted support for the reliability of the formulation and reduces batch variation in the yogurt industry.

[0213] For example, consider this parameter adjustment: the sample equilibrium point is 7.0, the filter validation retention rate is 90%, and the final ratio is 1:2.5, which meets the protein requirement of 88% and the flavor balance deviation of 3%. This can improve the product's shelf life and enhance the flavor balance.

[0214] In one possible implementation, for business-specific extensions, such as application in low-acidity base materials, step S2 verifies that the retention bottleneck can be identified, and S3 solves the problem by compensating for the proportion of extract, resulting in the technical effect of flavor restoration.

[0215] Specifically, the entire process closely follows the optimization of the yogurt base material production path to ensure the application of a balanced protein flavor ratio in the final product.

[0216] Based on the final protein flavor balance ratio, a stable quality output base material sequence is obtained by processing production batch data through continuous fluid simulation, including: processing production batch data using continuous fluid simulation based on the final protein flavor balance ratio, and obtaining a stable quality output base material sequence from the processed production batch data. The stable quality output base material sequence maintains consistent protein content and flavor substance distribution across multiple batches. The stable quality output base material sequence is used for subsequent judgment of fluctuations in layering parameters.

[0217] In one embodiment, step S1, processing the production batch data using continuous fluid simulation based on the final protein flavor balance ratio, specifically includes step S11, collecting production batch data, including protein content and small molecule flavor substance distribution parameters of multiple batches. These parameters are derived from the base sample data in the ultrafiltration process and are processed by a continuous fluid simulation algorithm. This algorithm simulates the flow and mixing behavior of fluid in the production pipeline based on the principles of fluid dynamics to ensure the application of the ratio in continuous production.

[0218] Step S12: Adjust the simulation parameters according to the final protein flavor balance ratio, for example, set the protein retention rate and flavor compensation ratio as inputs, and run the simulation to predict batch-to-batch changes.

[0219] Step S13: Output the processed production batch data, including optimized protein and flavor distribution sequences.

[0220] For example, in the production of yogurt base materials, continuous fluid simulation can simulate the mixing effect at different flow rates, which is beneficial to reduce batch fluctuations and improve production efficiency.

[0221] In one embodiment, step S2 involves obtaining a stable quality output base material sequence from the processed production batch data.

[0222] Specifically, stable sequence data are extracted from the simulation results to form a base material sequence, which is used to ensure consistent quality.

[0223] For example, the resulting sequence can be directly applied to subsequent verification.

[0224] In one embodiment, step S3, the stable quality output base material sequence maintains consistent protein content and flavor substance distribution across multiple batches, specifically includes step S31, monitoring protein content parameters in the sequence, for example, in batches 1 to 5, the protein content is maintained within the range of 15% to 16%, and by comparing the distribution consistency, ensuring that the concentration fluctuation of flavor substances such as organic acids does not exceed 2%.

[0225] Step S32: Verify consistency. If the deviation exceeds the threshold, iterate the simulation parameters to maintain balance.

[0226] Step S33: Generate the final consistent sequence, which is then applied to yogurt production.

[0227] For example, in small-scale batch production, this consistency can prevent flavor deviations and improve product stability; in medium-scale scenarios, by adjusting flow rate parameters, such as increasing the flow rate from 1.5 m / s to 2.0 m / s, maintaining sequence consistency is beneficial for quality control in large-scale applications; in another scenario, for high protein ratios, sequence consistency ensures balanced flavor and avoids loss of texture. These aspects support the reliability of stable output and emphasize the beneficial effects in the yogurt base production path, such as improving the overall uniformity of taste.

[0228] In one embodiment, step S4, the stable quality output base material sequence is used for subsequent layer sense parameter fluctuation judgment.

[0229] Specifically, the sequence input feedback adjustment algorithm is used to determine whether the fluctuation exceeds the threshold in order to optimize the production process.

[0230] For example, this application is directly connected to the refining production process configuration to ensure the final optimization of the yogurt base.

[0231] If the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds a threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration. This includes: determining if the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds a threshold, then initiating a feedback adjustment algorithm to correct the ultrafiltration parameter settings, obtaining the refining production process configuration from the corrected ultrafiltration parameter settings, reducing the fluctuation of the hierarchical sensitivity parameter, and using the refining production process configuration for subsequent key operation node data extraction and batch simulation verification of overall quality indicators.

[0232] In one embodiment, if the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds a threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration, including the following steps.

[0233] Step S1: If the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds the threshold, the feedback adjustment algorithm is activated to correct the ultrafiltration parameter settings.

[0234] Specifically, by collecting production batch data after continuous fluid simulation processing, the fluctuation value of the hierarchical sensing parameter, such as standard deviation or peak-to-valley difference, is calculated. If it exceeds a preset threshold such as 0.5, the feedback adjustment algorithm is activated. This algorithm is based on the proportional-integral-derivative control principle and iteratively adjusts the pressure and flow rate parameters in the ultrafiltration process to reduce fluctuations.

[0235] Step S11: Collect the hierarchical parameter data sequence from the stable quality output base material sequence.

[0236] For example, in the production of yogurt base materials, layering parameters, such as the gradient values ​​of flavor molecule distribution, are extracted from multiple batches of samples to form a data sequence for subsequent judgment.

[0237] Step S12: Calculate the volatility index of the data sequence and compare it with the threshold.

[0238] For example, volatility indicators can be obtained by calculating the standard deviation of the sequence. If the standard deviation is greater than a threshold, the correction phase begins. This approach can identify quality instability early and improve overall stability.

[0239] Step S13: Start the feedback adjustment algorithm to correct the ultrafiltration parameters based on proportional-integral-derivative control.

[0240] In one possible implementation, the feedback adjustment algorithm employs proportional-integral-derivative (PID) control, where the proportional part responds quickly to the current fluctuation error, the integral part accumulates historical errors to eliminate steady-state deviations, and the derivative part predicts fluctuation trends to suppress overshoot. Through these control links, the pore size selection and temperature setting of the ultrafiltration membrane are iteratively adjusted, for example, adjusting the temperature from an initial 40 degrees Celsius to 38 degrees Celsius to optimize protein retention and flavor balance.

[0241] Step S2: Obtain the refining process configuration from the corrected ultrafiltration parameter settings. The refining process configuration reduces the fluctuation of the hierarchical parameters.

[0242] Specifically, the corrected parameters, such as the adjusted pressure value and flow rate combination, form a refining production process configuration. This configuration, when applied in batch simulation, can reduce the fluctuation of the layering parameter from the initial 0.7 to below 0.3, resulting in a more stable base material output.

[0243] Step S21: Integrate the corrected ultrafiltration parameters to form the production process configuration.

[0244] For example, inputting the corrected parameters into a continuous fluid simulation generates a configuration sequence that ensures the configuration covers all nodes from initial enrichment to final equilibrium.

[0245] Step S22: Verify the effect of the configuration on reducing the fluctuation of the hierarchy parameter.

[0246] In one embodiment, simulation verification showed that the fluctuation was reduced by 20% after the configuration application, thanks to the accuracy of parameter correction, which is beneficial to maintaining the flavor consistency of the yogurt base.

[0247] Step S3, the refining production process configuration is used for subsequent key operation node data extraction and batch simulation verification of overall quality indicators.

[0248] Specifically, this configuration is used to extract node data from the process, such as protein content and organic acid concentration, and then batch simulations are performed to confirm the stability of quality indicators such as the flavor intensity equilibrium point.

[0249] Step S31: Extract key operation node data.

[0250] For example, select node data for ultrafiltration and ultrafiltration modules from the configuration for further analysis.

[0251] Step S32: Perform batch simulation verification of overall quality indicators.

[0252] In one embodiment, multiple batches are simulated to verify whether the indicators are balanced. This optimizes the yogurt base production path and brings the beneficial effect of improving production efficiency.

[0253] In one embodiment, the correction process of the feedback adjustment algorithm considers different threshold scenarios. For example, when the threshold is 0.4, the algorithm prioritizes adjusting the flow rate, reducing the fluctuation from 0.6 to 0.2; when the threshold is 0.6, the algorithm focuses on correcting the pressure, reducing the fluctuation from 0.8 to 0.3. These adjustments are based on batch data from continuous fluid simulation in the technical disclosure, and after expansion, ensure the stability of the protein flavor balance ratio.

[0254] For example, in high-fluctuation scenarios, the algorithm accumulates errors through the integral link of proportional-integral-derivative control, gradually adjusting the temperature parameter from 42 degrees Celsius to 39 degrees Celsius, reducing the fluctuation of the layering parameter by 15%, which is beneficial to the formation of the refining production process configuration.

[0255] In one embodiment, the effect of reducing fluctuations is extended from the mixing ratio adjustment of the flavor compensation algorithm in the technical disclosure. After configuration and application, not only is fluctuation reduced, but the retention rate of small molecule flavor substances is also indirectly improved. For example, the concentration of organic acids recovers from below the threshold to the equilibrium point, resulting in a richer yogurt flavor profile.

[0256] For example, in the simulated batch, after configuring the extraction node data, the verification showed that the fluctuation rate of quality indicators dropped to within 5%. This was achieved through precise correction of feedback adjustment, which is beneficial to the optimization of the overall yogurt base production path.

[0257] The above embodiments are merely one of the preferred embodiments of the present invention and should not be used to limit the scope of protection of the present invention. Any modifications or refinements made to the main design concept and spirit of the present invention that are not of substantial significance, but solve the same technical problem as the present invention, should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing flavor balance during the concentration process of milk base, characterized in that, include: Initial concentration state indicators were obtained by collecting base material sample data during the concentration process; Based on the initial concentration state indicators, the permeate analysis algorithm was used to process the composition of lost substances and determine the list of key flavor small molecules; If the concentration of a specific organic acid in the list of key components of flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered by the ultrafiltration module to obtain a supplementary flavor extract. The concentration data of aromatic compounds extracted from the supplementary flavor extract were used to adjust the mixing ratio of the extract and the concentrated base material using a flavor compensation algorithm to determine the balanced point of flavor intensity after optimization. The final protein flavor balance ratio was determined by verifying the protein retention rate through secondary membrane filtration of the mixed base sample corresponding to the optimized flavor intensity equilibrium point. Based on the final protein flavor balance ratio, a stable quality output base material sequence was obtained by processing production batch data through continuous fluid simulation. If the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds the threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration. Key operational node data were extracted from the refining production process configuration to conduct batch simulation verification of overall quality indicators and determine the optimal emulsion base material generation path.

2. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, The method of obtaining initial concentration state indicators by collecting base material sample data during the concentration process includes: During the ultrafiltration process, data from the substrate sample are collected to obtain protein content and small molecule flavor compound distribution parameters. From these parameters, an initial concentration state index is obtained. This initial concentration state index characterizes the protein retention and small molecule flavor compound loss of the concentrated substrate during the ultrafiltration stage. The initial concentration state index is used in subsequent permeate analysis algorithms to process the composition of lost substances.

3. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, The process of determining the list of key flavor small molecules by processing the composition of lost substances using a permeate analysis algorithm based on the initial concentration state index includes: Based on the initial concentration state indicators, the permeate analysis algorithm is used to process the composition of substances lost during ultrafiltration. From the composition of the lost substances, a list of key flavor small molecules is determined. The list of key flavor small molecules includes specific organic acids and other flavor-related small molecule compounds. The list of key flavor small molecules is used to determine whether the concentration of specific organic acids is lower than a preset threshold.

4. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, If the concentration of a specific organic acid in the list of key flavor small molecules is lower than a preset threshold, the corresponding compound in the permeate is recovered through an ultrafiltration module to obtain a supplementary flavor extract, including: If the concentration of a specific organic acid in the list of key flavor small molecules is lower than a preset threshold, the ultrafiltration module is activated to recover the corresponding compound in the permeate. A supplementary flavor extract is obtained from the recovered permeate. The supplementary flavor extract is rich in the lost flavor small molecule compounds. The supplementary flavor extract is used for subsequent extraction of aromatic compound concentration data.

5. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, The extraction of aromatic compound concentration data from the supplementary flavor extract is used to determine the optimal flavor intensity balance point by adjusting the mixing ratio of the extract and the concentrated base using a flavor compensation algorithm. This includes: Aromatic compound concentration data were extracted from the supplementary flavor extract. Based on the aromatic compound concentration data, a flavor compensation algorithm was used to adjust the mixing ratio of the supplementary flavor extract and the concentrated base. The flavor compensation algorithm was used to determine the optimized flavor intensity equilibrium point. The optimized flavor intensity equilibrium point represents the balanced state of the flavor intensity of the mixed base.

6. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, The process of obtaining the optimized flavor intensity equilibrium point corresponding to the mixed base sample and performing secondary membrane filtration verification on the protein retention rate to determine the final protein flavor balance ratio includes: Obtain the mixed base sample corresponding to the optimized flavor intensity balance point. Perform secondary membrane filtration verification on the protein retention rate of the mixed base sample. Determine the final protein flavor balance ratio from the secondary membrane filtration verification results. The final protein flavor balance ratio simultaneously meets the requirements of protein retention rate and flavor intensity balance.

7. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, The process of obtaining a stable quality output base material sequence by processing production batch data through continuous fluid simulation based on the final protein flavor balance ratio includes: Based on the final protein flavor balance ratio, continuous fluid simulation was used to process the production batch data to obtain a stable quality output base sequence. The stable quality output base sequence maintains consistent protein content and flavor substance distribution across multiple batches. The stable quality output base sequence is used for subsequent judgment of layering parameter fluctuations.

8. The method for flavor balance optimization during the concentration process of a milk base material according to claim 1, characterized in that, If the fluctuation of the hierarchical sensitivity parameter in the stable quality output base material sequence exceeds a threshold, a feedback adjustment algorithm is used to correct the concentration parameter settings to obtain the refining production process configuration, including: If the fluctuation of the hierarchical sense parameter in the stable quality output base material sequence exceeds the threshold, a feedback adjustment algorithm is activated to correct the ultrafiltration parameter settings. The refined production process configuration is obtained from the corrected ultrafiltration parameter settings. The refined production process configuration reduces the fluctuation of the hierarchical sense parameter. The refined production process configuration is used for subsequent key operation node data extraction and batch simulation verification of overall quality indicators.

Images