Yogurt smoothie fermentation synergistic regulation method based on multi-source data fusion

Abstract

The application discloses a yogurt smoothie fermentation synergistic regulation method based on multi-source data fusion, relates to the technical field of fermentation synergistic regulation, and is used for solving the problem of unstable fermentation product quality caused by the fermentation termination control method failing to simultaneously consider pH flavor targets and viscosity texture targets, collecting a pH drop rate in a fermentation tank and calculating a pH target final value deviation, collecting a material viscosity growth slope and analyzing a viscosity overrun risk coefficient, generating a double-target conflict intensity index and judging whether to enter a termination decision window, performing balance adjustment on a fermentation temperature compensation amount according to a double-target imbalance direction to generate a temperature intervention instruction, mapping a double-target synchronous convergence state into a termination timing evaluation amount and issuing a fermentation termination instruction, observing a gel network stable state after fermentation termination and performing remedial correction on the temperature intervention instruction, and improving the quality stability and termination timing accuracy of a yogurt smoothie fermentation product.

Description

Technical Field

[0001] This invention relates to the field of fermentation synergistic regulation technology, and more specifically, to a method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion. Background Technology

[0002] Yogurt smoothies are a type of fermented dairy product that balances flavor and texture. In the industrial production process, the acid metabolism of lactic acid bacteria continuously promotes the decrease of pH, bringing the flavor target close to the preset final value. The protein gel network continues to develop in a low pH environment. The choice of when to terminate fermentation directly determines the taste and texture of the final product. Therefore, effective online monitoring of the fermentation process and determination of the termination time are of great industrial significance. Existing fermentation control systems are usually equipped with pH sensors and online viscometers, which can realize the real-time acquisition of single indicators and threshold comparison.

[0003] The existing technology has the following shortcomings:

[0004] Currently, existing technologies cannot simultaneously quantify the antagonistic relationship between the deviation of the final pH target and the risk coefficient of viscosity exceeding the limit. The accuracy of judging the timing of fermentation termination is low, the risk of loss of control of the dual target quality of product pH and viscosity is high, and there is a lack of observation and temperature intervention remediation mechanisms for changes in the continuity of gel network after fermentation termination.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a synergistic regulation method for yogurt smoothie fermentation based on multi-source data fusion. By employing multi-source sensor data fusion and a dual-objective conflict quantification mechanism, the deviation of the final pH target and the viscosity exceedance risk coefficient are transformed into a unified conflict intensity index, driving the directional generation of temperature intervention commands and the synchronous convergence judgment of termination timing, thereby solving the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion, the specific operation process of which is as follows:

[0008] Step S1: Collect the pH decrease rate in the fermenter, calculate the deviation of the final pH value based on the pH decrease rate in the fermenter, collect the viscosity increase slope of the material and analyze the viscosity exceedance risk coefficient.

[0009] Step S2: Generate a dual-objective conflict intensity index by combining the deviation of the final pH target value and the viscosity over-limit risk coefficient. Determine whether to enter the termination decision window based on the dual-objective conflict intensity index. If the termination decision window is entered, read the fermentation temperature compensation amount. Perform a balance adjustment on the fermentation temperature compensation amount according to the dual-objective conflict intensity index and set the temperature intervention command.

[0010] Step S3: Drive the fermenter to perform temperature callback according to the temperature intervention command, continuously monitor the deviation of the final pH value and the viscosity over-limit risk coefficient within the termination decision window, generate the termination timing assessment quantity based on the synchronous convergence state of the two, and issue the fermentation termination command based on the termination timing assessment quantity.

[0011] Step S4: After fermentation is terminated, continuously collect the slope of material viscosity growth. Based on the slope of material viscosity growth, determine whether to perform remedial correction on the temperature intervention command. After performing remedial correction, output the corrected temperature intervention command.

[0012] In a preferred embodiment, in step S1, the pH inside the fermenter is collected in real time by a pH sensor installed on the top of the fermenter, and a pH time series is obtained according to a preset sampling period. The ratio of the pH difference between two adjacent samples to the sampling time interval is used as the pH decrease rate inside the fermenter.

[0013] The viscosity of the material is collected in real time by an online viscometer installed on the side wall of the fermenter. The material viscosity time series is obtained by sampling at the same period. The ratio of the difference in material viscosity between two adjacent samples to the sampling time interval is taken as the slope of the material viscosity increase.

[0014] In a preferred embodiment, in step S1, the difference between the current real-time value of the pH sensor and the preset target final value is obtained to get the current pH remaining deviation, and the pH target final value deviation is calculated in combination with the pH decrease rate in the fermenter.

[0015] The viscosity exceedance risk coefficient is obtained by calculating the ratio of the material viscosity growth slope to the difference between the preset viscosity upper limit threshold and the current value of the online viscometer.

[0016] In a preferred embodiment, in step S2, the deviation of the final pH target and the viscosity exceedance risk coefficient are respectively subjected to interval standardization.

[0017] The standardized results of the pH target final value deviation and the standardized results of the viscosity exceedance risk coefficient are fused using the geometric mean method to obtain the dual-objective conflict intensity index.

[0018] In a preferred embodiment, in step S2, the bi-target conflict intensity index is compared with the disturbance warning entry boundary:

[0019] When the dual-target conflict intensity index is greater than or equal to the disturbance warning entry boundary, it is determined to enter the termination decision window;

[0020] When the dual-target conflict intensity index is less than the disturbance warning entry boundary, it is determined that the termination decision window will not be entered.

[0021] The fermentation temperature compensation amount is a temperature adjustment benchmark amount that has been dynamically calculated by the temperature control unit based on the current operating conditions of the fermenter.

[0022] The temperature intervention command is obtained by performing a balance adjustment on the fermentation temperature compensation amount based on the bi-objective conflict intensity index.

[0023] In a preferred embodiment, in step S3, the fermenter temperature control unit is driven to perform a temperature callback action according to the temperature intervention command;

[0024] Within the termination decision window, the pH time series is continuously collected and the deviation of the pH target final value is recalculated at a preset sampling period. The material viscosity time series is collected and the viscosity exceedance risk coefficient is recalculated.

[0025] The deviation of the final pH target value and the viscosity exceedance risk coefficient reacquired within the termination decision window were respectively processed by interval standardization.

[0026] The sum of the absolute values ​​of the two standardized results is mapped to the termination timing assessment quantity using the inverse decay form.

[0027] In a preferred embodiment, in step S3, the termination timing evaluation quantity is compared with the termination trigger boundary:

[0028] When the evaluation value of the termination timing is greater than or equal to the termination trigger boundary, a fermentation termination command is issued.

[0029] When the termination timing assessment is less than the termination trigger boundary, the temperature callback state driven by the current temperature intervention command is maintained.

[0030] In a preferred embodiment, in step S4, the time when the fermentation termination command is issued is taken as the starting point of the observation period. The material viscosity time series is re-acquired within the preset observation period, and the material viscosity growth slope is recalculated.

[0031] Compare the viscosity growth slope with the viscosity stability fluctuation band:

[0032] When the viscosity increase slope of the material exceeds the viscosity stability fluctuation zone, a remedial and corrective action is initiated to respond to the temperature intervention command.

[0033] Conversely, no remedial action is taken, and the temperature intervention command remains unchanged.

[0034] In a preferred embodiment, in step S4, when the remedial correction action is initiated, the material viscosity growth slope is multiplied by the remedial correction sensitivity factor and the negative value is used as the remedial correction increment of the temperature intervention command.

[0035] The temperature intervention command is superimposed with the remedial correction increment to obtain the corrected temperature intervention command.

[0036] The technical effects and advantages of this invention are as follows:

[0037] This invention calculates the deviation of the final pH target by collecting the pH decrease rate in the fermenter and analyzes the viscosity exceedance risk coefficient by collecting the viscosity increase slope of the material. These are then fused to generate a dual-objective conflict intensity index. Based on the conflict intensity, the fermentation temperature compensation is adjusted to set a temperature intervention command, thus achieving coordinated intervention for the two opposing objectives. The synchronous convergence state of the two objectives is mapped to a termination timing assessment quantity, and a fermentation termination command is issued. After fermentation termination, the viscosity increase slope of the material is continuously collected, and the temperature intervention command is corrected accordingly. This forms a post-event correction for the termination decision deviation, improving the quality consistency of the same batch of fermented products. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the implementation of the synergistic regulation method for yogurt smoothie fermentation based on multi-source data fusion, as described in this invention.

[0039] Figure 2 This is a schematic diagram illustrating the steps of the synergistic regulation method for yogurt smoothie fermentation based on multi-source data fusion according to the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0041] This invention generates a dual-objective conflict intensity index by fusing the deviation of the final pH target value and the viscosity over-limit risk coefficient. This index drives the directional output of temperature intervention commands and the synchronous convergence evaluation of termination timing. Furthermore, it continuously observes the stable state of the gel network after fermentation termination to perform remedial corrections, thereby achieving the coordinated hit of the two opposing targets of pH and viscosity during the fermentation of yogurt smoothies.

[0042] Example 1

[0043] Please see Figures 1 to 2 The specific operation process of the synergistic regulation method for yogurt smoothie fermentation based on multi-source data fusion is as follows:

[0044] S1: Collect the pH decrease rate in the fermenter, calculate the deviation of the final pH target value based on the pH decrease rate in the fermenter, collect the viscosity increase slope of the material and analyze the viscosity exceedance risk coefficient.

[0045] S2: Generate a dual-objective conflict intensity index by combining the deviation of the final pH target value and the viscosity over-limit risk coefficient. Determine whether to enter the termination decision window based on the dual-objective conflict intensity index. After entering the termination decision window, read the fermentation temperature compensation amount, perform a balance adjustment on the fermentation temperature compensation amount according to the dual-objective conflict intensity index, and set the temperature intervention command.

[0046] S3: Drive the fermenter to perform temperature callback according to the temperature intervention command, continuously monitor the deviation of the final pH value and the viscosity over-limit risk coefficient within the termination decision window, generate the termination timing assessment quantity based on the synchronous convergence state of the two, and issue the fermentation termination command based on the termination timing assessment quantity.

[0047] S4: After fermentation is terminated, continuously collect the slope of material viscosity growth. Based on the slope of material viscosity growth, determine whether to perform remedial correction on the temperature intervention command. After performing remedial correction, output the corrected temperature intervention command.

[0048] The specific implementation is as follows:

[0049] In step S1, for the late stage of yogurt smoothie fermentation, the two core physical quantities reflecting the progress of the two objectives in the fermentation tank are collected in real time and corresponding target deviation representations are generated.

[0050] The late fermentation stage refers to the phase in yogurt smoothie fermentation where the pH approaches the target final value, the gel network is basically formed, and flavor and texture develop simultaneously. This is the time window where the conflict between the two objectives is most concentrated. During the late fermentation stage, the continued acid production and metabolism of lactic acid bacteria drive a continuous decrease in pH, thus pushing the flavor target closer to reality. Simultaneously, the protein gel network further develops in the low pH environment, causing a continuous increase in viscosity, thus approaching the texture target. These two objectives exhibit a conflicting direction at the late fermentation stage: continuing fermentation to approach the pH target leads to gel aging and viscosity exceeding the upper limit; prematurely terminating fermentation to preserve viscosity prevents the pH from reaching the flavor target.

[0051] Specifically, the pH inside the fermenter is collected in real time by a pH sensor installed on the top of the fermenter. The pH time series is obtained according to a preset sampling period. The ratio of the pH difference between two adjacent samples to the sampling time interval is used as the pH decrease rate in the fermenter to characterize the activity intensity of acid production metabolism at the current moment.

[0052] The viscosity of the material is collected in real time by an online viscometer installed on the side wall of the fermenter. The viscosity time series of the material is obtained by the same sampling period. The ratio of the difference in viscosity between two adjacent samples to the sampling time interval is used as the viscosity growth slope to characterize the development rate of the gel network at the current moment.

[0053] After obtaining the rate of pH decrease and the slope of material viscosity increase in the fermenter, corresponding deviation values ​​are generated to characterize the gap in flavor target and the risk of exceeding the quality target.

[0054] Specifically, the difference between the current real-time pH sensor value and the preset target final value is used to obtain the current remaining pH deviation; the pH decrease rate in the fermenter is multiplied by the preset extrapolation time, and then multiplied by the extrapolation weighting coefficient to obtain the direction correction term for the pH decrease rate in the fermenter on the deviation; the current remaining pH deviation is subtracted from the direction correction term to obtain the pH target final value deviation.

[0055]

[0056] in, This represents the deviation from the target final pH value. This is the real-time value obtained by the pH sensor at the current moment. The preset target final value, The rate of pH decrease within the fermenter. To preset the extrapolation duration, These are the extrapolation weighting coefficients.

[0057] The greater the rate of pH decrease in the fermenter and the closer the current pH is to the target final value, the faster the pH target final value deviation converges; the farther the current pH is from the target final value, the greater the pH target final value deviation, which is used to reflect the remaining gap of the flavor target.

[0058] It should be noted that the reason for choosing the linear extrapolation method to estimate the deviation of the final pH target is that it has high computational efficiency and transparent parameters, making it suitable for online real-time calculation scenarios within a short time window at the end of fermentation, without the need to introduce complex nonlinear fitting.

[0059] The viscosity exceedance risk coefficient is obtained by calculating the ratio of the material viscosity increase slope to the difference between the preset upper viscosity threshold and the current value measured by the online viscometer.

[0060]

[0061] in, This is the viscosity exceeding the limit risk coefficient. The slope of the viscosity increase of the material. The preset upper limit threshold for viscosity, This is the value obtained from the online viscometer at the current moment.

[0062] The greater the slope of the viscosity increase and the closer the current viscosity is to the upper limit, the greater the risk coefficient of viscosity exceeding the limit, which is used to reflect the urgency of exceeding the quality target.

[0063] In step S2, the deviation of the final pH target and the viscosity over-limit risk coefficient are fused in an adversarial manner to quantify the degree of conflict between the two opposing targets, and an intervention command to drive temperature callback is generated when the conflict is significant.

[0064] The dual-objective conflict intensity index is an intermediate quantity generated by the deviation of the final value of the pH target and the viscosity exceedance risk coefficient. It is used to quantify the degree of conflict between the current flavor target gap and the texture target exceedance risk, and is the core basis for determining whether to enter the termination decision window.

[0065] The termination decision window refers to the real-time decision-making period that is initiated after the dual-objective conflict intensity index reaches a set boundary. It is used to weigh the two opposing objectives and issue a fermentation termination command during this period.

[0066] Specifically, the deviation from the target pH value and the viscosity exceedance risk coefficient were both subjected to interval standardization. The calculation formula for interval standardization is as follows:

[0067]

[0068] Where X represents the deviation from the target pH value or the viscosity exceedance risk coefficient within the current sliding window. , These represent the minimum and maximum values ​​of corresponding quantities within the same sliding window. This is the corresponding standardized result.

[0069] Furthermore, the standardized results of the deviation from the final pH target value and the standardized results of the viscosity exceedance risk coefficient are fused using the geometric mean method to generate a dual-objective conflict intensity index:

[0070]

[0071] Where C is the dual-objective conflict intensity index, The result is a standardized result of the deviation from the target pH value. This is the standardized result of the viscosity exceeding the limit risk coefficient.

[0072] The larger the standardized result of the pH target final value deviation and the larger the standardized result of the viscosity exceedance risk coefficient, the larger the dual-target conflict intensity index, reflecting that when the flavor target gap and the texture target exceedance risk increase simultaneously, the antagonistic intensity of the two targets rises to the highest level in tandem.

[0073] Compare the dual-target conflict intensity index with the disturbance warning entry boundary:

[0074] When the dual-target conflict intensity index is greater than or equal to the disturbance warning entry boundary, it is determined that the dual-target conflict has become significant, and the termination decision window is entered, triggering the subsequent temperature intervention instruction generation action.

[0075] When the dual-target conflict intensity index is less than the disturbance warning entry boundary, it is determined that the dual-target conflict is not significant, the current fermentation control state is maintained, and the termination decision window is not entered.

[0076] It should be noted that the method for determining the disturbance warning entry boundary is as follows: Before the fermentation process is finalized, multiple batches of end-stage fermentation tests are conducted for conventional raw material formulas and inoculum amounts. For each batch, the pH decrease rate and material viscosity increase slope in the fermenter are collected synchronously at equal time intervals during the end-stage. The dual-objective conflict intensity index corresponding to each batch is calculated according to the aforementioned method to obtain the sample distribution of the dual-objective conflict intensity index. Subsequently, combined with the qualified marking of the dual-objective trend of each batch by manual evaluation, the lower quartile of the distribution corresponding to the significant conflict between the dual objects and the need to enter the decision window is taken as the disturbance warning entry boundary.

[0077] After entering the termination decision window, the fermentation temperature compensation amount is read from the fermentation tank temperature control unit device. The fermentation temperature compensation amount is the temperature adjustment benchmark amount that the temperature control unit has dynamically calculated based on the current fermentation tank operating conditions.

[0078] Based on the imbalance direction term corresponding to the bi-objective conflict intensity index, the fermentation temperature compensation is adjusted to obtain the temperature intervention command:

[0079]

[0080] in, This is a temperature intervention instruction. This is the amount of compensation for fermentation temperature. As a regulating factor, This represents the direction of the dual-objective imbalance.

[0081] When the standardized result of the viscosity exceedance risk coefficient is greater than the standardized result of the pH target final value deviation, the dual-target imbalance direction term is negative, and the temperature intervention instruction is adjusted to decrease the temperature in the direction of acid suppression to inhibit the gel aging process; when the standardized result of the pH target final value deviation is greater than the standardized result of the viscosity exceedance risk coefficient, the dual-target imbalance direction term is positive, and the temperature intervention instruction is adjusted to increase the temperature in the direction of acid promotion to make up for the remaining gaps in the flavor target.

[0082] It should be explained that the adjustment factor can be set by professionals in the field according to the actual situation. For example, the adjustment factor can be set to 0.15, which will not be elaborated here.

[0083] Through the above-mentioned process of generating the dual-objective conflict intensity index, determining the termination decision window, and outputting temperature intervention commands, the originally irreconcilable dual opposing objectives are transformed into directional adjustment commands that can be executed by the temperature control unit, providing a regulatory basis for the synchronous convergence evaluation of the two objectives in subsequent steps.

[0084] In step S3, the fermenter temperature control unit is driven to perform temperature callback according to the temperature intervention command, and the synchronous convergence status of the two opposite targets is continuously monitored within the termination decision window to determine the timing of issuing the fermentation termination command.

[0085] Specifically, the temperature control unit of the fermenter is driven to perform a temperature correction action according to the temperature intervention command. The temperature control unit adjusts the current fermentation temperature according to the temperature intervention command, so that the relative rate of acid production metabolism and gel aging shifts towards the direction of synchronous convergence of the two objectives.

[0086] Within the termination decision window, the pH time series is continuously collected by the pH sensor device installed on the top of the fermenter at a preset sampling period, and the deviation of the pH target final value is recalculated according to the method in step S1.

[0087] An online viscometer device installed on the side wall of the fermenter continuously collects the viscosity time series of the material and recalculates the viscosity exceedance risk coefficient according to the method in step S1;

[0088] The two recalculated representations in this step differ from their counterparts in step S1 only in that the sampling period is after the temperature callback action takes effect, and are used to reflect the real-time deviation of the two opposing targets after the temperature callback action takes effect.

[0089] The termination timing assessment quantity is an intermediate quantity that reflects the degree of synchronous convergence of two opposing objectives. It is used to uniformly characterize whether the two opposing objectives have fallen into the convergence interval within the termination decision window.

[0090] The deviation from the final pH target value and the viscosity exceedance risk coefficient, re-acquired within the termination decision window, are standardized using interval standardization to obtain the standardized results for the pH target deviation and the viscosity exceedance risk coefficient, respectively. The sum of the absolute values ​​of the two standardized results is then mapped to a termination timing assessment quantity using a reciprocal decay method.

[0091]

[0092] Where E represents the termination timing assessment quantity. To terminate the standardization of the absolute value of the final deviation of the pH target within the decision window, To terminate the standardization of the absolute value of the viscosity exceedance risk coefficient within the decision window, It is a convergence sensitivity factor.

[0093] When the standardized result of the deviation of the final value of the pH target and the standardized result of the viscosity over-limit risk coefficient both approach zero, the cumulative term in parentheses approaches zero, the denominator approaches 1, and the termination timing assessment quantity approaches 1, reflecting that the two opposing targets have simultaneously fallen into the convergence interval; when either term deviates, the denominator increases and the termination timing assessment quantity decreases, reflecting that there are still targets that have not converged.

[0094] It should be noted that the reciprocal decay form mathematically treats the deviations of two opposing targets as symmetrical and equivalent, and automatically maps multiple deviations to a monotonic interval of 0 to 1, which is convenient for comparison with a unified boundary and does not introduce additional directional coupling parameters.

[0095] Compare the termination timing assessment quantity with the termination trigger boundary:

[0096] When the termination timing assessment value is greater than or equal to the termination trigger boundary, it is determined that the two opposing targets have entered the convergence interval synchronously, and a fermentation termination command is issued to drive the fermenter into the cooling process.

[0097] When the termination timing assessment value is less than the termination trigger boundary, it is determined that the two targets have not yet converged synchronously. The temperature callback state driven by the current temperature intervention command is maintained, and the termination timing assessment value is recalculated in the next sampling cycle.

[0098] It should be noted that the method for determining the termination trigger boundary is as follows: Before the fermentation process is finalized, select several batches of samples that are manually judged as qualified termination timings at the end of fermentation. Calculate the termination timing assessment quantity corresponding to each sample using the same method as in this step, and statistically obtain the sample distribution of the termination timing assessment quantity. Then, take the upper quartile of the distribution that corresponds to the acceptable level of simultaneous convergence of the two objectives as the termination trigger boundary.

[0099] In step S4, after the fermentation termination command is issued, the actual stable state of the gel network is continuously observed, and if the gel network is unstable, the temperature intervention command referenced in step S2 is remedial and corrective.

[0100] Specifically, the moment the fermentation termination command is issued is taken as the starting point of the observation period. Within the preset observation time, an online viscometer device installed on the side wall of the fermenter continuously re-collects the material viscosity time series, and the material viscosity growth slope is recalculated according to the method in step S1. The difference between this material viscosity growth slope and the corresponding quantity in step S1 is that its sampling period is after the fermentation termination command is issued, and it is used to reflect the continuous change trend of the gel network after fermentation termination.

[0101] Compare the viscosity growth slope with the viscosity stability fluctuation band:

[0102] When the viscosity growth slope of the material exceeds the viscosity stability fluctuation zone, it is determined that the gel network is still in a state of continued development or reverse contraction, and a remedial correction action is initiated for the temperature intervention command; conversely, when the viscosity growth slope of the material falls into the viscosity stability fluctuation zone, it is determined that the gel network has stabilized, no remedial correction is performed, and the temperature intervention command remains unchanged.

[0103] It should be noted that the viscosity stability fluctuation band can be set by professionals in the field according to the actual situation. For example, the viscosity stability fluctuation band can be set to ±0.05 Pa·s / s, which will not be elaborated here.

[0104] The remedial correction sensitivity factor is a preset factor that reflects the sensitivity of the temperature intervention command to the slope of the material viscosity increase. It is used to convert the slope of the material viscosity increase into the correction direction and magnitude of the temperature intervention command.

[0105] When initiating a corrective action, the material viscosity increase slope is multiplied by the corrective sensitivity factor, and the negative value is used as the corrective increment for the temperature intervention command.

[0106]

[0107] in, This is a remedial correction increment for temperature intervention commands. This represents the slope of the material viscosity increase during the observation period after the fermentation termination command is issued. To remedy the sensitivity factor, the formula is explicitly negative to map the direction of the material viscosity growth slope to the opposite direction of the temperature return.

[0108] If the slope of the material viscosity increase is still positive and the larger the value, the remedial correction increment takes a negative value and the larger the absolute value. The temperature intervention command is reversed to the cooling direction to inhibit the continued development of the gel network. When the slope of the material viscosity increase is negative, the remedial correction increment takes a positive value. The temperature intervention command is reversed to the heating direction to slow down the whey separation tendency caused by the reverse contraction of the gel network.

[0109] It should be noted that the linear feedback correction method was chosen because its response direction is monotonic and its adjustment is sensitive, making it suitable for performing small-scale corrections at the fermentation termination stage when the gel network has been basically formed, thus avoiding overshoot.

[0110] The temperature intervention command is superimposed with the remedial correction increment to obtain the corrected temperature intervention command, and the corrected temperature intervention command is issued to the temperature control unit as the temperature intervention benchmark for subsequent tanks in the same batch.

[0111] Through the above-mentioned process of determining the stable state of the gel network, remediating and correcting temperature intervention commands, and issuing benchmarks, the deviation in the determination of the termination timing assessment is corrected after the fact, further ensuring that the same batch of fermentation products hits the two opposite targets of pH and viscosity simultaneously.

[0112] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0113] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0114] In this document, the singular forms “a,” “an,” and “the” may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that terms such as “comprising / including” or “having” specify the presence of the stated features, integrals, steps, operations, components, parts, or combinations thereof, but do not preclude the possibility of the presence or addition of one or more other features, integrals, steps, operations, components, parts, or combinations thereof. Meanwhile, the term “and / or” as used in this specification includes any and all combinations of the associated listed items.

[0115] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0116] The above description of the disclosed embodiments will enable those skilled in the art to make or use various modifications to these embodiments. It will be readily apparent to those skilled in the art that the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion, characterized by: Includes the following steps: Step S1: Collect the pH decrease rate in the fermenter, calculate the deviation of the final pH target based on the pH decrease rate in the fermenter, collect the viscosity increase slope of the material and analyze the viscosity exceedance risk coefficient. Step S2: Generate a dual-objective conflict intensity index by combining the deviation of the final pH target value and the viscosity over-limit risk coefficient. Determine whether to enter the termination decision window based on the dual-objective conflict intensity index. If the termination decision window is entered, read the fermentation temperature compensation amount. Perform a balance adjustment on the fermentation temperature compensation amount according to the dual-objective conflict intensity index and set the temperature intervention command. Step S3: Drive the fermenter to perform temperature callback according to the temperature intervention command, continuously monitor the deviation of the final pH value and the viscosity over-limit risk coefficient within the termination decision window, generate the termination timing assessment quantity based on the synchronous convergence state of the two, and issue the fermentation termination command based on the termination timing assessment quantity. Step S4: After fermentation is terminated, continuously collect the slope of material viscosity growth. Based on the slope of material viscosity growth, determine whether to perform remedial correction on the temperature intervention command. After performing remedial correction, output the corrected temperature intervention command.

2. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 1, characterized in that: In step S1, the pH inside the fermenter is collected in real time by a pH sensor installed on the top of the fermenter. The pH time series is obtained according to a preset sampling period, and the ratio of the pH difference between two adjacent samples to the sampling time interval is used as the pH decrease rate inside the fermenter. The viscosity of the material is collected in real time by an online viscometer installed on the side wall of the fermenter. The material viscosity time series is obtained by sampling at the same period. The ratio of the difference in material viscosity between two adjacent samples to the sampling time interval is taken as the slope of the material viscosity increase.

3. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 2, characterized in that: In step S1, the difference between the current real-time value of the pH sensor and the preset target final value is obtained to get the current pH remaining deviation. The pH target final value deviation is calculated in combination with the pH decrease rate in the fermenter. The viscosity exceedance risk coefficient is obtained by calculating the ratio of the material viscosity growth slope to the difference between the preset upper limit viscosity threshold and the current value measured by the online viscometer.

4. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 3, characterized in that: In step S2, the deviation of the final pH target and the viscosity exceedance risk coefficient are respectively subjected to interval standardization. The standardized results of the pH target final value deviation and the standardized results of the viscosity exceedance risk coefficient are fused using the geometric mean method to obtain the dual-objective conflict intensity index.

5. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 4, characterized in that: In step S2, the dual-target conflict intensity index is compared with the disturbance warning entry boundary: When the dual-target conflict intensity index is greater than or equal to the disturbance warning entry boundary, it is determined to enter the termination decision window; When the dual-target conflict intensity index is less than the disturbance warning entry boundary, it is determined that the termination decision window will not be entered. The fermentation temperature compensation amount is a temperature adjustment benchmark amount that has been dynamically calculated by the temperature control unit based on the current operating conditions of the fermenter. The temperature intervention command is obtained by performing a balance adjustment on the fermentation temperature compensation amount based on the bi-objective conflict intensity index.

6. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 5, characterized in that: In step S3, the fermenter temperature control unit is driven to perform a temperature callback action according to the temperature intervention command; Within the termination decision window, the pH time series is continuously collected and the deviation of the pH target final value is recalculated at a preset sampling period. The material viscosity time series is collected and the viscosity exceedance risk coefficient is recalculated. The deviation of the final pH target value and the viscosity exceedance risk coefficient reacquired within the termination decision window were respectively processed by interval standardization. The sum of the absolute values ​​of the two standardized results is mapped to the termination timing assessment quantity using the inverse decay form.

7. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 6, characterized in that: In step S3, the termination timing evaluation quantity is compared with the termination trigger boundary: When the evaluation value of the termination timing is greater than or equal to the termination trigger boundary, a fermentation termination command is issued. When the termination timing assessment is less than the termination trigger boundary, the temperature callback state driven by the current temperature intervention command is maintained.

8. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 1, characterized in that: In step S4, the time when the fermentation termination command is issued is taken as the starting point of the observation period. The material viscosity time series is re-acquired within the preset observation period, and the material viscosity growth slope is recalculated. Compare the viscosity growth slope with the viscosity stability fluctuation band: When the viscosity increase slope of the material exceeds the viscosity stability fluctuation zone, a remedial and corrective action is initiated to respond to the temperature intervention command. Conversely, no remedial action is taken, and the temperature intervention command remains unchanged.

9. The method for synergistic regulation of yogurt smoothie fermentation based on multi-source data fusion according to claim 8, characterized in that: In step S4, when the remedial action is initiated, the material viscosity growth slope is multiplied by the remedial correction sensitivity factor and the negative value is used as the remedial correction increment of the temperature intervention command. The temperature intervention command is superimposed with the remedial correction increment to obtain the corrected temperature intervention command.

Images