Intelligent collaborative method and system for multi-pot acid soup base simmering

Abstract

The application discloses an intelligent coordination method and system for multi-pot acid soup base stewing, and relates to the technical field of automatic control of food processing.The method comprises the following steps: collecting current data and vibration data of stirring motors of each stewing pot during the stewing process of multi-pot acid soup base; obtaining shearing characteristic data based on the current data, and performing variational modal decomposition on the vibration data to eliminate noise modes caused by solid impacts and reconstruct heat characteristic data; calculating the coherence spectrum entropy of the shearing characteristic data and the heat characteristic data in the frequency domain, constructing a stability constraint penalty term based on the coherence spectrum entropy and a preset demulsification threshold, and establishing a game benefit model containing the stability constraint penalty term; based on the game benefit model, taking heating efficiency and stirring uniformity as game subjects to solve a Nash equilibrium solution, determining the steam supply adjustment amount and the stirring intensity adjustment amount corresponding to each stewing pot, and coordinating and distributing each stewing pot.The method can significantly reduce the intensity of manual intervention and the risk of quality fluctuation.

Description

Technical Field

[0001] This invention relates to the field of automated control technology in food processing, and in particular to an intelligent collaborative method and system for cooking multi-jar sour soup base. Background Technology

[0002] The process of making sour soup base involves complex phase transitions of starch gelatinization and oil emulsification, which is a typical non-Newtonian fluid viscosity-changing thermal processing process. Current industrial production mostly adopts single-variable temperature feedback control, that is, monitoring the temperature through wall-mounted sensors to adjust the steam opening.

[0003] However, this control method has significant technical shortcomings in multi-tank parallel cooking scenarios. On the one hand, as the sour soup enters the middle and later stages of cooking, the viscosity of the material increases non-linearly, and the fluid is prone to overall rigid rotation, leading to a near-zero flow velocity at the boundary layer of the pot wall and a sudden increase in thermal resistance. The wall-attached sensors are affected by the thermal inertia of the wall layer, resulting in a severe lag in the detection data and an inability to promptly report early signs of local scorching. On the other hand, there is a physical coupling conflict between the heating rate and the stirring intensity. To overcome the high viscosity thermal resistance, the stirring shear force needs to be increased, but excessive shearing will damage the fragile oil-water emulsion system of the sour soup, causing mechanical demulsification and stratification. Existing technologies lack real-time perception of the fluid rheological state and a multivariate decoupling mechanism, making it difficult to achieve consistent quality control of the base material across multiple tanks while preventing scorching and achieving emulsification. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent collaborative method for cooking sour soup base in multiple pots to solve the following problems in the cooking process: the nonlinear change in fluid viscosity causes the overall rigid rotation, resulting in a lag in the monitoring of the temperature attached to the wall, and the blind spot in the identification of local scorching. In addition, there is a physical coupling contradiction between heating rate and stirring intensity, where improving heat transfer efficiency can easily lead to mechanical demulsification, while reducing stirring shear can easily lead to boundary layer coking.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent collaborative method for cooking multiple pots of sour soup base, which includes: collecting current data and vibration data of the stirring motor of each cooking pot during the cooking process of multiple pots of sour soup base;

[0008] Shear feature data is obtained based on current data, and variational mode decomposition is performed on vibration data to remove noise modes of solid impact and reconstruct thermal feature data.

[0009] Calculate the coherent spectral entropy of shear feature data and thermal feature data in the frequency domain, construct a stability constraint penalty term based on the coherent spectral entropy and a preset demulsification threshold, and establish a game payoff model that includes the stability constraint penalty term.

[0010] Based on the game payoff model, the Nash equilibrium solution is solved with heating efficiency and stirring uniformity as the main game subjects. The steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank are determined, and the allocation of steam to each boiling tank is coordinated.

[0011] As a preferred embodiment of the intelligent collaborative method for cooking multi-tank sour soup base as described in this invention, the method involves: collecting current data and vibration data of the stirring motor of each cooking tank during the cooking process of multi-tank sour soup base;

[0012] Shear feature data is obtained based on current data, and variational mode decomposition is performed on vibration data to remove noise modes of solid impact and reconstruct thermal feature data.

[0013] Calculate the coherent spectral entropy of shear feature data and thermal feature data in the frequency domain, construct a stability constraint penalty term based on the coherent spectral entropy and a preset demulsification threshold, and establish a game payoff model that includes the stability constraint penalty term.

[0014] Based on the game payoff model, the Nash equilibrium solution is solved with heating efficiency and stirring uniformity as the main game subjects. The steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank are determined, and the allocation of steam to each boiling tank is coordinated.

[0015] As a preferred embodiment of the intelligent collaborative method for cooking multi-jar sour soup base as described in this invention, the calculation of the coherence spectral entropy of shear feature data and thermal feature data in the frequency domain includes: extracting spectral features from the shear feature data, identifying the main frequency component and harmonic components of the stirring driving energy, and determining the frequency domain interval containing the main frequency component and harmonic components as the dominant frequency band of the shear feature.

[0016] The full-band auto-power spectral density of shear feature data and thermal feature data, as well as the full-band cross-power spectral density between shear feature data and thermal feature data, are calculated separately. A frequency domain weighting function is constructed with the dominant frequency band of shear feature as the weight center, and the full-band cross-power spectral density is weighted using the frequency domain weighting function to obtain the weighted cross-power spectral density.

[0017] Based on the weighted cross-power spectral density and the full-band self-power spectral density, the amplitude squared coherence coefficient distribution is calculated, the amplitude squared coherence coefficient distribution is normalized, and the Shannon entropy is calculated. The Shannon entropy is then used as the coherence spectral entropy.

[0018] As a preferred embodiment of the intelligent collaborative method for cooking multi-jar sour soup base as described in this invention, the construction of the stability constraint penalty term includes: constructing a stability constraint penalty term based on the coherence spectrum entropy; using the time domain amplitude of the shear feature data to represent the macroscopic viscosity of the fluid, and establishing a dynamic mapping relationship between the demulsification threshold and the macroscopic viscosity of the fluid to obtain the rheological adaptive safety threshold.

[0019] Set a safety buffer bandwidth and determine the safety buffer lower limit by combining the rheological adaptive safety threshold; when the coherence spectral entropy is higher than the safety buffer lower limit, calculate the penalty value in the form of a logarithmic function; when the coherence spectral entropy is lower than or equal to the safety buffer lower limit, calculate the penalty value in the form of a quadratic function.

[0020] The calculated penalty value is used as the stability constraint penalty term in the game payoff model.

[0021] As a preferred embodiment of the intelligent collaborative method for preparing multi-pot sour soup base according to the present invention, the establishment of the game payoff model includes: defining a heating efficiency game subject and a stirring uniformity game subject; setting the steam supply adjustment amount as the strategy variable of the heating efficiency game subject, and taking maximizing the temperature rise rate as the control objective of the heating efficiency game subject; setting the stirring intensity adjustment amount as the strategy variable of the stirring uniformity game subject, and taking minimizing the fluctuation variance of the shear feature data as the control objective of the stirring uniformity game subject;

[0022] Construct payoff functions for the heating efficiency game subject and the stirring uniformity game subject. The payoff functions include a performance gain term, an energy consumption cost term, and the stability constraint penalty term.

[0023] As a preferred embodiment of the intelligent collaborative method for cooking multi-jar sour soup base as described in this invention, the method of solving the Nash equilibrium solution with heating efficiency and stirring uniformity as the main game subjects includes: constructing the physical feasible domain of the stirring intensity adjustment amount, removing the dominant frequency band of shear features from the physical feasible domain, and constructing a discontinuous stirring strategy search space.

[0024] The temporal amplitude of shear feature data is used as the basis for determining the game dominance; when the temporal amplitude is lower than the preset viscosity threshold, the heating efficiency game subject is taken as the dominant party and the stirring uniformity game subject is taken as the subordinate party, and the game payoff model is iteratively solved in the stirring strategy search space.

[0025] When the time-domain amplitude is higher than the viscosity critical value, the game payoff model is iteratively solved within the stirring strategy search space, with the stirring uniformity game subject as the dominant party and the heating efficiency game subject as the subordinate party.

[0026] After determining that the iterative solution has converged, the obtained Nash equilibrium solution is identified as the steam supply adjustment amount and the stirring intensity adjustment amount.

[0027] As a preferred embodiment of the intelligent collaborative method for cooking sour soup base in multiple pots as described in this invention, the iterative solution of the game payoff model includes obtaining a preset basic update step size, a damping sensitivity index, and a benchmark turbulence entropy value, wherein the benchmark turbulence entropy value is configured as a standard entropy value characterizing the sour soup base in a rarefied turbulent state.

[0028] During the iterative solution process, the current coherence spectrum entropy is read in real time; the ratio of the current coherence spectrum entropy to the reference turbulence entropy value is calculated to obtain the real-time safety ratio; the damping sensitivity index is used to perform a power operation on the real-time safety ratio to obtain the step size correction coefficient.

[0029] The update step size for the current iteration cycle is calculated by multiplying the basic update step size by the step size correction coefficient, and the steam supply adjustment amount and stirring intensity adjustment amount are updated using the update step size.

[0030] Secondly, the present invention provides an intelligent collaborative system for the cooking of multiple pots of sour soup base, including: a data module for collecting current data and vibration data of the stirring motor of each cooking pot during the cooking process of multiple pots of sour soup base;

[0031] The reconstruction module acquires shear feature data based on current data, performs variational mode decomposition on vibration data, removes noise modes from solid impact, and reconstructs thermal feature data.

[0032] The calculation module calculates the coherent spectral entropy of shear feature data and thermal feature data in the frequency domain, constructs a stability constraint penalty term based on the coherent spectral entropy and a preset demulsification threshold, and establishes a game payoff model that includes the stability constraint penalty term.

[0033] The execution module, based on a game payoff model, uses heating efficiency and stirring uniformity as the main game subjects to solve the Nash equilibrium solution, determines the steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank, and coordinates the allocation of resources among the boiling tanks.

[0034] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the intelligent collaborative method for preparing multi-jar sour soup base as described in the first aspect of the present invention.

[0035] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the intelligent collaborative method for preparing multi-jar sour soup base as described in the first aspect of the present invention.

[0036] The beneficial effects of this invention are as follows: By simultaneously collecting current and vibration data of the stirring motor during the cooking process of multiple pots of sour soup base, shear characteristics and thermal characteristics are constructed respectively, achieving decoupled perception of fluid rheological state and thermal cavitation behavior, avoiding the control lag problem caused by traditional reliance on a single temperature or empirical parameter; by calculating coherent spectral entropy based on the dominant frequency band of shear characteristics, and introducing a frequency domain weighting and rheological adaptive stability constraint penalty mechanism, the risks of demulsification and burning can be identified in advance during the cooking process; by constructing a game payoff model, and combining the search space of discontinuous stirring strategy, the determination of the dominant power of time domain amplitude, and the entropy-oriented adaptive step size, the coordinated optimization control of heating and stirring at different viscosity stages is achieved, thereby improving heating efficiency and consistency of multi-pot cooking while ensuring emulsification stability, and significantly reducing the intensity of human intervention and the risk of quality fluctuation. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 A flowchart for an intelligent collaborative method for preparing sour soup base from multiple pots.

[0039] Figure 2 A flowchart of the game payoff model for an intelligent collaborative method for cooking sour soup base in multiple pots.

[0040] Figure 3 A computer device diagram for an intelligent collaborative method for preparing sour soup base from multiple pots. Detailed Implementation

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0042] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0043] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0044] Reference Figures 1-3 This is one embodiment of the present invention, which provides an intelligent collaborative method for preparing sour soup base in multiple containers, comprising the following steps:

[0045] S1: Collect current and vibration data of the stirring motor of each cooking tank during the cooking process of multiple sour soup bases; obtain shear feature data based on the current data, and perform variational mode decomposition on the vibration data to remove the noise mode of solid impact and reconstruct thermal feature data.

[0046] In a multi-tank sour soup cooking system, a stirring drive unit is used as a single composite sensing node to perform synchronous data acquisition.

[0047] Real-time torque and current data of the mixer motor inverter can be read via fieldbus. This data directly reflects the natural low-pass filtering characteristics of the fluid resistance torque during impeller rotation. Composite vibration data transmitted via the rigid stirring shaft were collected using a MEMS sensor installed inside the stirring reducer housing at a sampling frequency of 5120 Hz. The vibration data includes high-frequency vibrations from bubble thermalization and transient vibrations from solid impacts.

[0048] Since there is a clear physical correspondence between torque current and apparent fluid viscosity, and it is unaffected by high-frequency cavitation noise, the collected real-time torque current data can be directly used. Defined as cut feature data No mode decomposition is required:

[0049]

[0050] To achieve adaptive noise removal, vibration data is first analyzed. Frequency domain topology determines the number of decomposition layers in VMD. Vibration data within the current time window Perform a Fast Fourier Transform (FFT) to calculate its power spectral density. The formula is expressed as:

[0051]

[0052] in, Indicates time; The power spectral density function representing the signal; Represents the angular frequency variable; This indicates the total number of signal sampling points within the current time window; The base of the natural logarithm; Represents the imaginary unit ( .

[0053] The calculated power spectral density curve A full-band peak search is performed to locate all local maxima as candidate energy ridges. Based on a preset noise floor threshold and minimum band spacing, the candidate energy ridges are screened, retaining effective energy ridges that simultaneously satisfy the constraints of amplitude significance (i.e., amplitude significantly higher than background noise) and band independence (i.e., adjacent peaks are not adjacent). The total number of effective energy ridges is counted and directly assigned to the number of real-time modes required for variational mode decomposition at the current moment. .

[0054] Utilizing real-time computing A variational constraint model for vibration data is constructed. The goal is to find... One intrinsic mode component This minimizes the sum of the estimated bandwidths for each mode. The variational constraint model formula is as follows:

[0055]

[0056]

[0057] in, The decomposition yields the first... One intrinsic mode function component; Indicates the first The center frequency corresponding to each modal component; Indicates time The partial derivative operator; Dirac Function (unit impulse function); Represents pi; The square of the norm is used to measure the bandwidth energy of a signal.

[0058] The alternating direction multiplier method (ADMM) is used to iteratively solve the model in the frequency domain, and the final output is... One intrinsic mode component .

[0059] Because the vibration data contains transient strong noise generated by solid objects such as bones impacting the stirring shaft, it is necessary to use statistical features to accurately remove it and retain the pure bubble cavitation characteristics.

[0060] For each modal component Calculate its standardized kurtosis The formula is as follows:

[0061]

[0062] in, Indicates the first The standardized kurtosis values ​​of each modal component; Indicates the sample point index of a discrete signal; Indicates the first The modal component in the th ... Discrete values ​​of each sampling point; : indicates the first The arithmetic mean of the modal components.

[0063] In this embodiment, a kurtosis threshold is set. If a certain mode It was determined to be a solid-object impact noise mode (with non-stationary pulse characteristics), and the mode was directly eliminated.

[0064] The remaining effective modal components are superimposed in the time domain to obtain thermal feature data. .

[0065]

[0066] in, The final determined thermal characteristic data represents the cavitation state of the bubbles; Valid represents the set of valid modes retained after kurtosis filtering. The thermal characteristic data is not a thermodynamic temperature measurement, but rather a cavitation vibration characterization quantity at the fluid dynamics level. This data is collected using vibration sensors, aiming to indirectly and without hysteresis quantify the thermal input state during the cooking process by monitoring the mechanical vibration intensity of bubble formation and collapse.

[0067] By acquiring torque and current data from the stirring motor, the physical mapping relationship between fluid resistance and motor load is directly utilized, enabling lag-free sensing of overall viscosity changes in the sour soup. Compared to traditional contact viscometers, using current signals as the shear characteristic data source avoids the engineering challenges of sensor probe scaling and corrosion in high-temperature acidic environments. By monitoring minute load fluctuations transmitted through the rigid stirring shaft, the micro-rheological characteristics of the fluid's transition from Newtonian to non-Newtonian fluids can be captured, providing a pure physical quantity input for subsequent precise control.

[0068] Variational mode decomposition (VMD) is used to process high-frequency vibration data, overcoming the frequency aliasing defect of traditional Fourier transform when processing non-stationary signals. The environment for simmering sour soup is extremely harsh, with the sounds of bursting bubbles, bone impacts, and mechanical transmission noise all mixed together. VMD, through iteratively searching for the optimal solution of the variational model, can adaptively decompose complex vibration signals into intrinsic modal components with different center frequencies. This signal processing method does not rely on preset basis functions, thus effectively removing irrelevant background noise while preserving the cavitation characteristics of bubbles (reflecting heat exchange efficiency).

[0069] Introducing kurtosis statistics as a signal filtering criterion solves the problem of difficult-to-remove solid particle impact interference. Vibrations generated by solid materials impacting the agitator exhibit typical pulse characteristics, statistically manifested as abnormally high kurtosis values. By calculating the kurtosis of each modal component and setting a threshold, modal components containing impact characteristics can be directly identified and discarded. The remaining low-kurtosis modal components are reconstructed into thermal feature data, ensuring that the final input system data only contains bubble cavitation information reflecting fluid thermal dynamics, greatly improving the signal-to-noise ratio and the reliability of the features.

[0070] S2: Calculate the coherence spectral entropy of shear feature data and thermal feature data in the frequency domain, construct a stability constraint penalty term based on the coherence spectral entropy and a preset demulsification threshold, and establish a game payoff model that includes the stability constraint penalty term.

[0071] In the later stages of simmering the sour soup, as the viscosity of the base increases, the fluid is prone to rigid motion following the rotation of the stirring paddle (i.e., overall rigid rotation). At this point, the heat exchange efficiency drops sharply, and the soup is highly susceptible to burning. To accurately monitor this risk, the system locks onto the main frequency band of the stirring energy injection. Specifically, a Fast Fourier Transform (FFT) is performed on the shear characteristic data (characterizing the load on the stirring motor) output in step S1. The fundamental frequency peak of the stirring motor is searched in the spectrum. and its predecessor Significant harmonic components. The frequency domain interval containing the fundamental peak and the significant harmonic components is defined as the dominant frequency band of the bleaching characteristic.

[0072]

[0073] in, This indicates the frequency band dominated by the blurring characteristics, which is the set of frequency domain intervals that need to be monitored.

[0074] Indicates the harmonic order index variable; Indicates the preset highest harmonic order; This indicates the peak value of the fundamental frequency of the stirring motor; This indicates the preset half-bandwidth.

[0075] When the sour soup base undergoes severe gelatinization or shows signs of coking, the thermal characteristic signal (bubble vibration) will lose its randomness and instead exhibit a highly synchronized periodic fluctuation within that frequency band with the stirring frequency.

[0076] Calculate the full-band autopower spectral density of shear characteristic data and thermal characteristic data respectively. , and cross-power spectral density The squared coherence coefficient of the full-band amplitude was obtained. Next, a frequency domain weighting function is constructed with the frequency band dominated by the blurred features as the weight center. We weight the cross-power spectral density to construct a normalized probability distribution sequence. :

[0077]

[0078] Finally, the weighted coherence spectral entropy of the shear band is calculated. :

[0079]

[0080] in, Represents frequency The amplitude squared coherence coefficient at the location; Represents frequency variables; : Represents the cross-power spectral density between shear feature data and thermal feature data; This represents the auto-power spectral density of the shear feature data; The self-power spectral density represents the thermal characteristic data; Representing discrete frequency points The weighted normalized probability distribution value at the location; Indicates the first A discrete frequency point; Indicates the frequency domain weighted window function in The value at; Indicates the summation index variable; Indicates the total number of valid frequency points; This represents the weighted coherence spectrum entropy of the sheared frequency band.

[0081] The 'coherence spectral entropy' introduced here serves as a quantitative indicator to measure the synchronization complexity between mechanical stirring and thermal cavitation response. Based on the principle of information entropy, when this entropy value is high, it indicates that the coupling between stirring energy and thermal characteristics in the frequency domain exhibits a wide and disordered distribution, meaning that the fluid is uniformly mixed internally without obvious dead zones or resonant structures at fixed frequencies. Conversely, when this entropy value decreases, it indicates that energy is highly concentrated at a specific frequency (such as the stirring master frequency), suggesting that the fluid may have undergone overall rigid rotation (following rotation) or stratification, meaning that the mechanical motion and heat transfer process has lost its due random turbulent characteristics.

[0082] when A higher value indicates that the stirring energy effectively breaks up the gelatinized clumps, the red oil droplets in the sour soup are evenly dispersed, the emulsion is stable, and the heat convection is smooth.

[0083] when At lower frequencies, it indicates that energy is highly concentrated at the stirring frequency, creating dead zones inside the sour soup. The stirring paddle exerts a strong mechanical cutting effect on the already formed red oil emulsion layer, indicating the risk of mechanical demulsification and scorching at the bottom of the pot.

[0084] Sour soup base is a typical non-Newtonian fluid, with its viscosity increasing exponentially with cooking time, and its sensitivity to shear failure also increases accordingly. Therefore, the safety threshold must be dynamically adjusted according to viscosity. First, the mean temporal amplitude of the shear characteristic data within the current time window is calculated. , used to represent the real-time gelatinization viscosity of the sour soup. A dynamic mapping relationship between the anti-demulsification threshold and the real-time gelatinization viscosity is established to obtain the rheologically adaptive safety threshold.

[0085]

[0086] in, The preset basic security entropy threshold, This is the threshold gain coefficient. Rheological sensitivity coefficient; The mean temporal amplitude of the shear feature data is used to characterize the real-time gelatinization viscosity of the sour soup.

[0087] The larger the number, the thicker the sour soup; the thicker the sour soup, the more it needs to be. The higher the viscosity, the more difficult it is for the sour soup to dissipate heat. A higher level of disordered mixing must be maintained to prevent localized overheating and burning.

[0088] Set a safety buffer lower limit Construct a stability constraint penalty term for the stability of sour soup.

[0089]

[0090] in, This represents the penalty term for stability constraints; This represents the obstacle penalty weighting coefficient; Indicates the lower limit of the safety buffer (by...) Confirmed, among which (for buffer bandwidth); The constant coefficients of the quadratic function segment are used to ensure a smooth connection. This represents the constant coefficient of the intercept of the function at the connection point.

[0091] when When this occurs, it indicates that the sour soup is on the critical edge of breaking the emulsion or coking, and at this time the function outputs a strong penalty signal.

[0092] Define the players in the heating efficiency game, with the steam valve opening as their strategy variable. The control objective is to maximize the rate of temperature rise. Define the game subject for achieving uniformity of mixing, with the mixing motor frequency as the strategy variable. The control objective is to minimize the variance of the shearing feature data. .

[0093] Establish the payoff functions for the two subjects respectively. and The model is as follows:

[0094]

[0095]

[0096] in, This represents the payoff function value of the main players in the heating efficiency game. Indicates the weighting of heating performance gain; This indicates the rate of temperature rise, which is the direct control target of the heating element; Indicates the weight of heating energy consumption cost; This indicates the steam valve opening degree, which is a strategy variable for the heating system. The payoff function value represents the payoff of the players in the game of mixing uniformity. Indicates the weighting of stirring performance gain; The variance of the shear characteristic data represents the direct control target of the mixing body; This represents a small constant to prevent the denominator from being zero; Indicates the weight of stirring energy consumption cost; This represents the frequency of the stirring motor, which is a strategy variable for the main stirring component.

[0097] By using stability constraint penalty terms Simultaneously, the revenue functions of two entities are introduced, establishing a "risk-sharing" mechanism at the mathematical model level. Once the sour soup exhibits signs of overall rigid rotation (i.e....) reduce), The value will increase dramatically, leading to and At the same time, it decreased significantly.

[0098] A shear-band-weighted coherent spectral entropy was constructed, establishing a composite index capable of quantifying the synergy between "stirring input" and "thermal response." Single temperature or viscosity indices cannot describe the microscopic mixing state of fluids, while coherent spectral entropy, by analyzing the correlation between two heterogeneous signals in the frequency domain, reveals the efficiency of mechanical energy conversion to thermal energy. When fluids exhibit stratification or overall rigid rotation, the coherence between mechanical shear and heat exchange decreases significantly, leading to a reduction in entropy. This calculation method makes hidden flow field anomalies quantifiable and monitorable.

[0099] By establishing and weighting the dominant frequency bands for gelatinization characteristics, the monitoring focus is precisely locked onto the frequency range most prone to fluid resonance. During the gelatinization process of sour soup, the fluid structure undergoes drastic changes at specific frequencies. By weighting the frequency bands containing the fundamental frequency and significant harmonics, the influence of signal changes within these key frequency bands on the overall entropy value is artificially amplified. This processing mechanism makes the control system extremely sensitive to early signs of burning, enabling it to detect subtle deterioration trends at the fluid dynamics level before any anomalies occur in the macroscopic temperature field.

[0100] A rheologically adaptive hybrid barrier penalty term is established to set up a safety barrier for the control system that dynamically tightens with viscosity. Traditional fixed threshold control cannot adapt to the entire process of the sour soup changing from thin to thick. By establishing a functional mapping between the safety threshold and real-time viscosity, the allowable entropy fluctuation range decreases as the fluid becomes thicker. Once the monitored data approaches this dynamic boundary, the value of the penalty function output increases exponentially. This mechanism forces subsequent optimization algorithms to prioritize safety under critical conditions, eliminating the risk of blindly pursuing heating efficiency while ignoring physical boundaries.

[0101] S3: Based on the game payoff model, the Nash equilibrium solution is solved with heating efficiency and stirring uniformity as the main game subjects. The steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank are determined, and the allocation of each boiling tank is coordinated.

[0102] like Figure 3 As shown, based on the dominant frequency band of the shear feature, it is mapped to the corresponding frequency range of the stirring motor. Subsequently, within the physically permissible frequency range of the stirring motor, the stirring frequency range is deleted, thereby constructing a discontinuous stirring strategy search space.

[0103] when When the viscosity is below the preset critical value, the sour soup is determined to be in a thin state, and the system prioritizes calculating the optimal strategy of the heating efficiency game subject.

[0104] when When the viscosity is higher than or equal to the preset critical viscosity value, the sour soup is determined to be in a thick state. The system prioritizes calculating the optimal strategy of the game subject of stirring uniformity, and the heating strategy needs to be optimized under the premise of meeting the stirring requirements.

[0105] In this embodiment, the reference turbulent entropy value is obtained through online calibration during the initial stage of cooking: in the initial feeding stage of cooking the sour soup (when the starch has not yet gelatinized and the fluid is in a low-viscosity Newtonian fluid state), the stirring motor is controlled to run at its rated speed for a preset calibration time window. Within this time window, the shear band weighted coherence spectrum entropy is calculated in real time, and the arithmetic mean of all entropy values ​​within this time window is calculated and defined as the reference turbulent entropy value.

[0106] A preset base update step size and damping sensitivity index are set. During subsequent iterative solutions, the current shear band weighted coherence spectral entropy is read in real time. The current shear band weighted coherence spectral entropy is divided by the baseline turbulence entropy value to obtain a real-time safety ratio reflecting the degree of deviation of the current mixing state from the ideal state. Using the damping sensitivity index as the exponent, the real-time safety ratio is exponentially calculated to obtain the step size correction coefficient. The base update step size is multiplied by the step size correction coefficient to obtain the actual update step size for the current iteration cycle.

[0107] Using the actual update step size calculated above, the game payoff model is solved online iteratively under the determined dominance mode.

[0108] During the solution process, for each iteratively updated policy variable, it is forcibly projected into its respective feasible region; especially for It is necessary to ensure that it falls within the search space of discontinuous stirring strategies. The iteration ends when the change in the strategy variable between two consecutive iterations is less than a preset convergence threshold, and the converged result is recorded. and The final control command is issued to the actuator; when the sour soup is thin, a large steam opening combined with moderate stirring intensity rapidly raises the temperature; when the sour soup is thick and at risk, a strong stirring command that avoids the resonant frequency band is forcibly executed, and the steam opening is actively reduced to match the limited heat exchange capacity. Thus, dynamic synergy between heating and stirring is achieved at the physical level, maximizing cooking efficiency while preventing local scorching and demulsification.

[0109] By constructing a discontinuous stirring strategy search space, the possibility of resonance is blocked from a purely physical and logical perspective. Traditional control algorithms typically search for optimal solutions in a continuous domain, easily straying into specific speed ranges that could trigger resonance. By mapping the identified dominant frequency band to the motor frequency range and forcibly eliminating it from the feasible domain, a control no-go zone is artificially created. The optimization algorithm operates within this discontinuous space, and the output stirring commands naturally avoid dangerous frequencies, achieving proactive physical isolation between overall rigid rotation and the risk of resonance and scorching.

[0110] A master-slave game mechanism based on viscosity criteria was introduced to resolve the inherent coupling conflict between heating rate and stirring uniformity. During the thinner stage of the sour soup, heating efficiency takes precedence to maximize production capacity; however, during the thicker stage where viscosity exceeds a critical value, the mechanism is forcibly switched to prioritizing stirring uniformity, ensuring that during the risk accumulation period, the stirring strategy prioritizes disrupting laminar flow and preventing coking. This dynamic switching of decision-making order simulates the technological wisdom of "high heat for thin soup, frequent stirring for thick soup" from human experience, achieving a dynamic balance of multi-objective optimization.

[0111] Implementing entropy-guided adaptive variable step size adjustment provides the control system with stability protection similar to that of a damper. When far from the risk region, a larger step size ensures the system's rapid response to deviations; however, when the coherence spectrum entropy decreases and the system approaches the safety margin, the calculation step size decays exponentially according to the power rule.

[0112] This embodiment also provides an intelligent collaborative system for preparing sour soup base in multiple containers, including:

[0113] The data module collects current and vibration data of the stirring motors in each of the multiple pots of sour soup base during the cooking process.

[0114] The reconstruction module acquires shear feature data based on current data, performs variational mode decomposition on vibration data, removes noise modes from solid impacts, and reconstructs thermal feature data.

[0115] The calculation module calculates the coherence spectral entropy of shear feature data and thermal feature data in the frequency domain, constructs a stability constraint penalty term based on the coherence spectral entropy and a preset demulsification threshold, and establishes a game payoff model that includes the stability constraint penalty term.

[0116] The execution module, based on a game payoff model, uses heating efficiency and stirring uniformity as the main game subjects to solve the Nash equilibrium solution, determines the steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank, and coordinates the allocation of resources among the boiling tanks.

[0117] This embodiment also provides a computer device applicable to the intelligent collaborative method for cooking multiple pots of sour soup base, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent collaborative method for cooking multiple pots of sour soup base as proposed in the above embodiment.

[0118] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0119] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the intelligent collaborative method for preparing sour soup base in multiple containers as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A smart collaborative method for preparing sour soup base in multiple pots, characterized in that, include: During the cooking process of multiple batches of sour soup base, the current and vibration data of the stirring motors in each cooking tank were collected. Shear feature data is obtained based on current data, and variational mode decomposition is performed on vibration data to remove noise modes of solid impact and reconstruct thermal feature data. Calculate the coherent spectral entropy of shear feature data and thermal feature data in the frequency domain, construct a stability constraint penalty term based on the coherent spectral entropy and a preset demulsification threshold, and establish a game payoff model that includes the stability constraint penalty term. Based on the game payoff model, the Nash equilibrium solution is solved with heating efficiency and stirring uniformity as the main game subjects. The steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank are determined, and the allocation of steam to each boiling tank is coordinated.

2. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 1, characterized in that: The variational mode decomposition of vibration data includes performing spectral analysis on the vibration data within the current time window, identifying the number of effective energy ridges in the power spectral density, and using the number of effective energy ridges as the real-time mode number at the current moment; constructing a variational constraint model using the real-time mode number, and solving it using the alternating direction multiplier method to obtain several intrinsic mode components. Calculate the normalized kurtosis value of each intrinsic mode component, and mark the intrinsic mode components with normalized kurtosis values ​​greater than a preset kurtosis threshold as solid impact noise modes and remove them; The thermal feature data is obtained by superimposing the remaining intrinsic mode components after removing the solid impact noise mode.

3. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 2, characterized in that: The calculation of the coherence spectral entropy of shear feature data and thermal feature data in the frequency domain includes extracting spectral features from the shear feature data, identifying the dominant frequency component and harmonic components of the stirring driving energy, and determining the frequency domain interval containing the dominant frequency component and harmonic components as the dominant frequency band of the shear feature. The full-band auto-power spectral density of shear feature data and thermal feature data, as well as the full-band cross-power spectral density between shear feature data and thermal feature data, are calculated separately. A frequency domain weighting function is constructed with the dominant frequency band of shear feature as the weight center, and the full-band cross-power spectral density is weighted using the frequency domain weighting function to obtain the weighted cross-power spectral density. Based on the weighted cross-power spectral density and the full-band self-power spectral density, the amplitude squared coherence coefficient distribution is calculated, the amplitude squared coherence coefficient distribution is normalized, and the Shannon entropy is calculated. The Shannon entropy is then used as the coherence spectral entropy.

4. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 3, characterized in that: The construction of the stability constraint penalty term includes: constructing a stability constraint penalty term based on the coherence spectrum entropy; using the time-domain amplitude of the shear feature data to represent the macroscopic viscosity of the fluid, and establishing a dynamic mapping relationship between the demulsification threshold and the macroscopic viscosity of the fluid to obtain the rheological adaptive safety threshold. Set a safety buffer bandwidth and determine the safety buffer lower limit by combining the rheological adaptive safety threshold; when the coherence spectral entropy is higher than the safety buffer lower limit, calculate the penalty value in the form of a logarithmic function; when the coherence spectral entropy is lower than or equal to the safety buffer lower limit, calculate the penalty value in the form of a quadratic function. The calculated penalty value is used as the stability constraint penalty term in the game payoff model.

5. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 4, characterized in that: The establishment of the game payoff model includes defining a heating efficiency game subject and a stirring uniformity game subject; setting the steam supply adjustment amount as the strategy variable of the heating efficiency game subject, and taking maximizing the temperature rise rate as the control objective of the heating efficiency game subject; setting the stirring intensity adjustment amount as the strategy variable of the stirring uniformity game subject, and taking minimizing the fluctuation variance of the shear characteristic data as the control objective of the stirring uniformity game subject. Construct payoff functions for the heating efficiency game subject and the stirring uniformity game subject. The payoff functions include a performance gain term, an energy consumption cost term, and the stability constraint penalty term.

6. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 5, characterized in that: The method of finding the Nash equilibrium solution with heating efficiency and stirring uniformity as the main game subjects includes: constructing the physical feasible region of the stirring intensity adjustment amount, removing the dominant frequency band of shear features from the physical feasible region, and constructing a discontinuous stirring strategy search space. The temporal amplitude of shear feature data is used as the basis for determining the game dominance; when the temporal amplitude is lower than the preset viscosity threshold, the heating efficiency game subject is taken as the dominant party and the stirring uniformity game subject is taken as the subordinate party, and the game payoff model is iteratively solved in the stirring strategy search space. When the time-domain amplitude is higher than the viscosity critical value, the game payoff model is iteratively solved within the stirring strategy search space, with the stirring uniformity game subject as the dominant party and the heating efficiency game subject as the subordinate party. After determining that the iterative solution has converged, the obtained Nash equilibrium solution is identified as the steam supply adjustment amount and the stirring intensity adjustment amount.

7. The intelligent collaborative method for preparing multi-pot sour soup base as described in claim 6, characterized in that: The iterative solution of the game payoff model includes obtaining a preset basic update step size, damping sensitivity index and benchmark turbulence entropy value, wherein the benchmark turbulence entropy value is configured as a standard entropy value characterizing the sour soup base in a rarefied turbulent state. During the iterative solution process, the current coherence spectrum entropy is read in real time; Calculate the ratio of the current coherence spectral entropy to the reference turbulence entropy value to obtain the real-time safety ratio; use the damping sensitivity index to perform a power operation on the real-time safety ratio to obtain the step size correction coefficient; The update step size for the current iteration cycle is calculated by multiplying the basic update step size by the step size correction coefficient, and the steam supply adjustment amount and stirring intensity adjustment amount are updated using the update step size.

8. An intelligent collaborative system for cooking multi-jar sour soup base, based on the intelligent collaborative method for cooking multi-jar sour soup base as described in any one of claims 1 to 7, characterized in that: The data module collects current and vibration data of the stirring motors in each of the multiple pots of sour soup base during the cooking process. The reconstruction module acquires shear feature data based on current data, performs variational mode decomposition on vibration data, removes noise modes from solid impact, and reconstructs thermal feature data. The calculation module calculates the coherent spectral entropy of shear feature data and thermal feature data in the frequency domain, constructs a stability constraint penalty term based on the coherent spectral entropy and a preset demulsification threshold, and establishes a game payoff model that includes the stability constraint penalty term. The execution module, based on a game payoff model, uses heating efficiency and stirring uniformity as the main game subjects to solve the Nash equilibrium solution, determines the steam supply adjustment amount and stirring intensity adjustment amount corresponding to each boiling tank, and coordinates the allocation of resources among the boiling tanks.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the intelligent collaborative method for cooking multi-jar sour soup base as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the intelligent collaborative method for cooking multi-jar sour soup base as described in any one of claims 1 to 7.

Images