A food baking processing control system and method

Abstract

This invention discloses a food baking processing control system and method, belonging to the field of intelligent control technology for food processing. The system includes: acquiring the dielectric properties of materials and capturing the dynamic changes in the dielectric loss tangent; simultaneously using time-domain dielectric spectroscopy analysis to identify the phase transition stages of the materials in real time; constructing a dual-channel heat conduction model and dynamically identifying the regional temperature gradient of the baking equipment based on the phase transition stages; and calculating the microwave power adjustment and fan speed ratio adjustment of the baking equipment based on the regional temperature gradient of the baking equipment using a multi-parameter coupled control algorithm, and generating a composite control command set. This invention identifies the phase transition stages of materials in real time using time-domain dielectric spectroscopy analysis, and accurately analyzes the correlation between the relaxation time distribution and the phase transition characteristic spectrum using inverse Laplace transform and dynamic time warping algorithms, achieving dynamic capture of key phase transition nodes such as glass transition and starch gelatinization.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for food processing, and in particular to a food baking processing program control system and method. Background Technology

[0002] In the food baking and processing field, conventional control methods are mainly based on a closed-loop control system of temperature feedback and preset process parameters. Existing technologies typically employ thermocouple arrays to monitor the surface temperature distribution of materials, combined with PID controllers to dynamically adjust microwave power output and fan speed to achieve temperature field uniformity control. Some systems obtain the dielectric loss tangent through frequency domain dielectric spectroscopy analysis to characterize the moisture content and polar molecular activity intensity of materials, thereby aiding in determining the baking stage. This type of method, by establishing a linear mapping relationship between temperature gradient, microwave power, and fan speed, and using multivariate regression models to generate control commands, has already achieved some application in the stability control of baking processes. Furthermore, offline detection data from texture analyzers and colorimeters are often used for post-hoc correction of process parameters to optimize product quality consistency.

[0003] However, the aforementioned methods have room for improvement in the dynamic characteristics analysis and heat conduction modeling of the phase transition stage. While conventional frequency-domain dielectric spectroscopy analysis can obtain the frequency response characteristics of the dielectric loss tangent, it fails to establish a dynamic correlation between the time-domain relaxation current signal and phase transition characteristics, limiting the accuracy of identifying key phase transition nodes such as glass transition and starch gelatinization. On the other hand, a single heat conduction model cannot simultaneously characterize the coupling effect of slow thermal relaxation (dominated by molecular chain motion) and rapid heat transfer (dominated by moisture migration), resulting in deviations between the predicted temperature gradient and the actual phase transition kinetics. These deviations are particularly significant in the starch gelatinization stage, where the high moisture migration rate makes it difficult for traditional PID control algorithms to achieve coordinated adaptive adjustment of microwave power and wind speed, easily leading to localized overheating or underbaking. Summary of the Invention

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

[0005] Therefore, the present invention provides a food baking processing procedure control method to solve the problems of insufficient recognition accuracy in the phase transition stage and temperature field instability caused by inaccurate microwave-wind speed coordinated control.

[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 a food baking processing procedure control method, comprising: acquiring the dielectric properties of the material and capturing the dynamic changes in the dielectric loss tangent, while simultaneously using time-domain dielectric spectroscopy analysis to identify the phase transition stage of the material in real time; constructing a dual-channel heat conduction model and dynamically identifying the regional temperature gradient of the baking equipment according to the phase transition stage of the material; based on the regional temperature gradient of the baking equipment, calculating the microwave power adjustment amount and wind speed ratio adjustment amount of the baking equipment through a multi-parameter coupled control algorithm, and generating a composite control instruction set; acquiring the variation range of the material color difference and texture data, dynamically generating pulse cooling and gradient heating compensation instructions through a fuzzy PID compensator, and dynamically correcting the composite control instruction set.

[0008] As a preferred embodiment of the food baking processing procedure control method of the present invention, the dynamic change of the dielectric loss tangent value refers to obtaining the dielectric loss tangent value by frequency domain dielectric spectrum analysis based on the dielectric constant and loss factor, and performing dynamic tracking analysis using the time-temperature superposition principle to obtain the dynamic change curve of the dielectric loss tangent value.

[0009] As a preferred embodiment of the food baking processing control method of the present invention, the step of using time-domain dielectric spectroscopy analysis to identify the phase transition stage of materials in real time includes the following steps.

[0010] Based on the dynamic change curve of the dielectric loss tangent, the phase transition characteristics of the material are extracted using the sliding window difference method, and the phase transition characteristic spectrum is generated by the dynamic time warping algorithm.

[0011] The time-domain dielectric relaxation current signal of the material is subjected to inverse Laplace transform to generate a relaxation time distribution that is time-aligned with the phase transition characteristic spectrum.

[0012] The minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum was calculated using time-domain dielectric spectroscopy analysis.

[0013] Low and high deviation thresholds are defined and compared with the minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum to identify the glass transition stage, starch gelatinization stage, and setting and maturation stage of the material.

[0014] As a preferred embodiment of the food baking processing control method of the present invention, the steps of constructing a dual-channel heat conduction model and dynamically identifying the regional temperature gradient of the baking equipment according to the phase change stage of the material are as follows:

[0015] The dual-channel heat conduction model includes a slow thermal relaxation channel and a fast heat transfer channel;

[0016] The slow thermal relaxation channel was analyzed using time-domain dielectric relaxation spectroscopy to identify the hysteresis characteristics of molecular chain motion.

[0017] The relaxation activation energy and thermal history temperature characteristics of the material were calibrated by differential scanning calorimetry, and a slow thermal relaxation time spectrum was constructed by Prony series fitting.

[0018] By using the finite element discretization method, the slow thermal relaxation time spectrum is spatially meshed and solved to predict the regional temperature gradient of the baking equipment.

[0019] The rapid heat transfer channel uses TGA-FTIR to measure the rate of moisture diffusion and the rate of change of thermal conductivity during baking. A quantitative relationship between moisture migration and heat conduction is established by fitting the Arrhenius equation. At the same time, the finite volume method is used to predict the regional temperature gradient of the baking equipment.

[0020] Materials at different phase change stages are processed using a dual-channel heat conduction model consisting of a slow thermal relaxation channel and a fast heat transfer channel.

[0021] As a preferred embodiment of the food baking processing procedure control method of the present invention, the generation of the composite control instruction set refers to calculating the microwave power adjustment amount and the wind speed ratio adjustment amount through a multi-parameter coupled control algorithm based on the regional temperature gradient of the baking equipment, and generating the composite control instruction set through a linear interpolation mapping method.

[0022] As a preferred embodiment of the food baking processing procedure control method of the present invention, the steps of obtaining the variation range of material color difference and texture data, and dynamically generating pulse cooling and gradient heating compensation commands through a fuzzy PID compensator are as follows:

[0023] Based on real-time material color difference and texture data, a sliding window statistical method is used to identify the range of variation of material color difference and texture data.

[0024] A fuzzy clustering algorithm is used to divide the variation range of material color difference and texture data into multiple fuzzy subsets, and an IF-THEN control rule is established for each fuzzy subset.

[0025] Based on material color difference and texture data, pulse cooling and gradient heating compensation commands are generated through the IF-THEN control rules of each fuzzy subset.

[0026] As a preferred embodiment of the food baking processing procedure control method of the present invention, the dynamic correction of the composite control instruction set refers to superimposing and correcting the pulse cooling and gradient heating compensation instructions and the composite control instruction set through a fuzzy PID compensator, and using the corrected composite control instruction set to control the baking process in real time.

[0027] Secondly, this invention provides a food baking processing control system, comprising a phase change identification module, a temperature gradient identification module, an instruction generation module, and an instruction correction module. The phase change identification module is used to collect the dielectric properties of the material and capture the dynamic changes in the dielectric loss tangent, while simultaneously using time-domain dielectric spectroscopy analysis to identify the phase change stage of the material in real time. The temperature gradient identification module is used to construct a dual-channel heat conduction model and dynamically identify the regional temperature gradient of the baking equipment based on the phase change stage of the material. The instruction generation module is used to calculate the microwave power adjustment and wind speed ratio adjustment of the baking equipment based on the regional temperature gradient of the baking equipment using a multi-parameter coupled control algorithm, and generate a composite control instruction set. The instruction correction module is used to obtain the variation range of the material's color difference and texture data, dynamically generate pulse cooling and gradient heating compensation instructions using a fuzzy PID compensator, and dynamically correct the composite control instruction set.

[0028] 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 food baking processing program control method as described in the first aspect of the present invention.

[0029] 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 food baking processing program control method as described in the first aspect of the present invention.

[0030] The beneficial effects of this invention are as follows: It identifies material phase transition stages in real time using time-domain dielectric spectroscopy analysis; it accurately analyzes the correlation between relaxation time distribution and phase transition characteristic spectra using inverse Laplace transform and dynamic time warping algorithms, achieving dynamic capture of key phase transition nodes such as glass transition and starch gelatinization. By constructing a dual-channel heat conduction model, and fitting molecular chain motion characteristics based on Prony series and determining moisture migration patterns using TGA-FTIR, a coupled computational framework for slow thermal relaxation and rapid heat transfer is established, achieving accurate matching between phase transition stages and temperature gradient prediction. Attached Figure Description

[0031] 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.

[0032] Figure 1 A flowchart for a food baking and processing procedure control method.

[0033] Figure 2This is a schematic diagram of a food baking processing control system.

[0034] Figure 3 This is a flowchart for identifying the phase transition stages of materials.

[0035] Figure 4 This is a flowchart of a two-channel heat conduction model. Detailed Implementation

[0036] 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.

[0037] 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.

[0038] 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.

[0039] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a food baking processing procedure control method, including the following steps:

[0040] S1. Collect the dielectric properties of the material and capture the dynamic changes in the dielectric loss tangent. At the same time, use time-domain dielectric spectrum analysis to identify the phase transition stage of the material in real time.

[0041] The dielectric properties of a material refer to its dielectric constant and loss factor, which are acquired using a broadband dielectric sensor.

[0042] Based on the dielectric constant and loss factor, the dielectric loss tangent is obtained by frequency domain dielectric spectrum analysis, and the dynamic tracking analysis is performed by the time-temperature superposition principle to obtain the dynamic change curve of the dielectric loss tangent.

[0043] Furthermore, the frequency domain dielectric spectroscopy analysis method scans the dielectric response of the material in a logarithmic step manner within the frequency range of 10Hz to 1MHz. A precision LCR meter combined with a three-electrode measurement method is used to simultaneously acquire the dielectric constant and loss factor. The ratio of the dielectric constant to the loss factor is converted into the dielectric loss tangent value through complex number operations. Using the time-temperature superposition principle, isothermal measurement points are set within the temperature range of 25℃ to 150℃. The horizontal displacement factor and vertical scaling factor are calculated for the frequency domain dielectric spectrum acquired at each temperature point. The WLF equation is used to map the frequency domain data of each temperature range to a reference temperature, constructing a master curve covering 12 orders of magnitude of frequency. The Hanning window weighted moving average algorithm is used to process the master curve data, extracting the dielectric loss tangent value of the characteristic frequency band with a step size of 0.5 octaves. Adaptive piecewise cubic Hermitian interpolation is used to generate a dynamic change curve of the dielectric loss tangent value with a time resolution of 10ms.

[0044] Based on the dynamic change curve of the dielectric loss tangent, the phase transition characteristics of the material are extracted using the sliding window difference method, and the phase transition characteristic spectrum is generated by the dynamic time warping algorithm.

[0045] Furthermore, the dynamic change curve of dielectric loss tangent is slid along the time axis through a rectangular window with a fixed time width. The values ​​of adjacent data points within the window are subjected to first-order difference operations to calculate the local slope change and extreme point density. The segment in the dynamic change curve of dielectric loss tangent where the slope mutation rate exceeds the mutation threshold (range: 0.5-3.0) is captured as a candidate region for phase transition features. The dynamic time warping algorithm nonlinearly aligns the slope change sequence of the candidate region with the standard phase transition template sequence. By constructing a dynamic bending path and a cumulative difference matrix, the time dimension scaling difference is eliminated. The slope mutation rate, characteristic peak intensity, and plateau duration extracted from the multi-window are integrated into a time-aligned set of phase transition feature vectors, forming a phase transition feature spectrum characterizing the glass transition and starch gelatinization process of the material.

[0046] The phase transition characteristics of materials include the abrupt change rate of the slope in the dynamic change of the dielectric loss tangent, the characteristic peak intensity, and the plateau duration (the period during which the dielectric loss tangent remains stable).

[0047] The time-domain dielectric relaxation current signal of the material is simultaneously acquired using a current probe and a high-voltage pulse generator.

[0048] Furthermore, the time-domain dielectric relaxation current signal acquisition process employs a high-voltage pulse generator to apply a square wave excitation with a nanosecond-level rising edge, which is connected to a three-electrode test fixture via a coaxial cable. The material is placed between the electrodes as a medium. The current probe uses a broadband Rogowski coil structure, which is wrapped around the high-voltage loop conductor to capture the polarization current decay waveform in real time. The synchronous triggering mechanism ensures that the leading edge of the excitation pulse is strictly aligned with the oscilloscope acquisition clock. The signal conditioning circuit performs logarithmic amplification and baseline drift compensation on the micro-current signal. Finally, the complete time-domain dielectric current decay curve is obtained by sampling at equal intervals and stored as a time-current amplitude sequence.

[0049] The time-domain dielectric relaxation current signal of the material is subjected to inverse Laplace transform to generate a relaxation time distribution that is time-aligned with the phase transition characteristic spectrum.

[0050] Furthermore, the inverse Laplace transform process takes the time decay curve of the time-domain dielectric relaxation current signal as input, establishes a complex frequency domain integral equation to characterize the relaxation process, and uses the regularized least squares method to solve the ill-conditioned integral equation. High-frequency noise interference is suppressed by truncating singular value decomposition. The transform kernel function is selected from the exponential decay basis function set, and the basis function distribution density is optimized by the mesh refinement algorithm. The time axis of the relaxation time distribution is aligned with the time marker points of the phase transition characteristic spectrum by the linear interpolation method, forming a relaxation time distribution with strict time dimension matching.

[0051] The minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum is calculated using time-domain dielectric spectroscopy analysis. The expression is as follows:

[0052] ;

[0053] in, This represents the minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum. It is a set of paths connecting data points of two signal sequences, consisting of a series of discrete coordinate points. The relaxation time is distributed at the sampling time. amplitude, The first time-domain dielectric relaxation current signal represents the... Each sampling time, The phase transition characteristic spectrum at the sampling time amplitude, It is an index of the sampling time of the time-domain dielectric relaxation current signal. It is the index of the sampling time of the phase transition characteristic spectrum.

[0054] Furthermore, when calculating the minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum using time-domain dielectric spectroscopy analysis, the dynamic time warping algorithm first constructs a two-dimensional grid coordinate system between the relaxation time distribution amplitude sequence and the phase transition characteristic spectrum amplitude sequence, traversing all possible path sets. The connection method of discrete coordinate points; for each candidate path, calculate the relaxation time distribution amplitude corresponding to each coordinate point on the path. With phase transition characteristic spectral amplitude The squared differences are summed to form the total path difference value. A dynamic programming method is used to iteratively update the cumulative difference matrix from the starting point to the ending point, retaining the minimum cumulative difference value for each grid point. Finally, the path set is backtracked... The path with the minimum cumulative difference is found by extracting the sum of squared differences of all coordinate points along the path, and then taking the square root to obtain the minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum. .

[0055] Based on the time-domain dielectric relaxation current signal of historical materials, a low deviation threshold D1 and a high deviation threshold D2 are defined.

[0056] It should be noted that, based on the time-domain dielectric relaxation current signal of historical materials, the distribution range of the minimum cumulative difference D between the relaxation time distribution and the phase transition characteristic spectrum corresponding to the glass transition stage, starch gelatinization stage, and setting and ripening stage is statistically analyzed. The minimum cumulative difference for each glass transition stage is then selected. The upper limit of the distribution is used as the low deviation threshold D1, representing the minimum cumulative difference between the starch gelatinization stage and the setting and ripening stage. The boundary value of the distribution is used as the high deviation threshold D2; the low deviation threshold D1 ranges from 0.5 to 1.5, and the high deviation threshold D2 ranges from 1.5 to 3.0.

[0057] when When the value is ≤D1, the material is considered to be in the glass transition stage; for example, when the minimum cumulative difference... =0.8 ( When ≤D1=1.0), the material is in the glass transition stage, and the dielectric loss tangent shows a slow upward trend.

[0058] When D1 < When the difference is ≤D2, the material is considered to be in the starch gelatinization stage; for example, when the minimum cumulative difference is ≤D2, the material is considered to be in the starch gelatinization stage. =1.8 (D1=1.0< When ≤D2=2.5), the material is in the starch gelatinization stage, at which point the dielectric loss tangent increases rapidly;

[0059] when When the value is greater than D2, the material is considered to be in the shaping and maturation stage.

[0060] When the minimum cumulative difference =3.2 ( When D2 = 2.5, the material is in the stabilization and maturation stage, at which point the dielectric loss tangent tends to be stable.

[0061] S2. Construct a dual-channel heat conduction model and identify the regional temperature gradient of the baking equipment based on the dynamics of the material phase change stage.

[0062] The dual-channel heat conduction model includes a slow thermal relaxation channel and a fast heat transfer channel;

[0063] The slow thermal relaxation channel is as follows:

[0064] Based on the dielectric loss tangent, time-domain dielectric relaxation spectrum analysis is used to identify the hysteresis characteristics of molecular chain motion.

[0065] Furthermore, the process of identifying the hysteresis characteristics of molecular chain motion through time-domain dielectric relaxation spectroscopy analysis is as follows: A broadband dielectric spectrometer acquires the time-domain decay curve of the dielectric loss tangent, and the time-domain response of the polarization current is recorded by applying a step voltage excitation; the analysis focuses on the nonlinear decay characteristics of the dielectric loss tangent over time, and a multi-exponential fitting method is used to separate relaxation peaks at different time scales, calculating the ratio of the full width at half maximum (FWHM) of the main relaxation peak to the peak time constant; the weight of the long-term relaxation component in the total relaxation energy is quantified by evaluating the integral area ratio of the relaxation peak tailing effect; when analyzing the degree of obstruction to the conformational change of the molecular chain, the asymmetric characteristics of the relaxation time distribution are examined; the final identification result is manifested as hysteresis characteristics of molecular chain motion, specifically including kinetic parameters such as broadening of the FWHM of the main relaxation peak and increased span of the main peak time constant distribution.

[0066] Based on the hysteresis characteristics of chain motion, the relaxation activation energy and thermal history temperature characteristics of the material were calibrated by differential scanning calorimetry, and a slow thermal relaxation time spectrum was constructed by Prony series fitting.

[0067] Furthermore, differential scanning calorimetry (DSC) measures the heat flux of the material with temperature at a constant heating rate, determines the glass transition temperature by the endothermic peak initiation point, and calculates the relaxation activation energy by integrating the endothermic peak area. The calibration of the thermal history temperature characteristics is based on multi-stage annealing experiments, recording the enthalpy relaxation recovery curves at different temperature holding stages, and analyzing the relationship between the recovery rate constant and temperature. The relaxation time constant sequence from time-domain dielectric relaxation spectroscopy analysis is used as initial parameters to approximate the decay curve of the slow thermal relaxation process in the form of a linear combination of exponential terms. The amplitude coefficients and time constants of each exponential term are optimized by the least squares method, ultimately constructing the slow thermal relaxation time spectrum.

[0068] The slow thermal relaxation time spectrum is spatially meshed and solved using the finite element method discretization to predict the regional temperature gradient of the baking equipment. The expression is as follows:

[0069] ;

[0070] in, It is the regional temperature gradient of the baking equipment. It is the effective working volume of the baking equipment. It is the slow thermal relaxation strength coefficient. This is the current baking time. It is the spatial curvature of the temperature distribution in baking equipment. It is the first molecular chain motion A time scale It is the total number of timescales of molecular chain motion. It is an index of the time scale of molecular chain motion. It is the first Thermal history correction factor for each time scale It is the molar gas constant. It is the current real-time temperature of the material. It is a characteristic of thermal history temperature;

[0071] Furthermore, the finite element discretization method first discretizes the effective working volume of the baking equipment. Discretize the data into tetrahedral or hexahedral mesh elements, and assign the molecular chain motion timescale from the slow thermal relaxation time spectrum to each element node. Slow thermal relaxation strength coefficient Thermal history correction coefficient and real-time temperature Based on the current baking time Thermal history temperature characteristics Calculate the exponential decay term within each unit. The instantaneous value is obtained, and the spatial curvature of the temperature distribution is solved using the central difference method. Sum the exponential decay terms of each unit. With spatial curvature Multiplication is performed, and numerical integration is conducted within the element volume using the Gaussian integration method. After summing the integral contributions of all mesh elements, the regional temperature gradient of the baking equipment is obtained. This gradient value reflects the coupling effect between the slow thermal relaxation effect and the spatial variation of the temperature field.

[0072] It should be noted that the slow thermal relaxation strength coefficient : 1.0×10² to 5.0×10³, thermal history correction factor 0.1 to 5.0, timescale of molecular chain motion : 1.0 s to 1.0 × 10³ s, molar gas constant : Fixed value 8.314 J / (mol·K).

[0073] The rapid heat transfer channel is as follows:

[0074] The rate of moisture diffusion and the rate of change of thermal conductivity during baking were determined by TGA-FTIR, and a quantitative relationship between moisture migration and heat conduction was established by fitting the Arrhenius equation.

[0075] Furthermore, the TGA-FTIR synchronous acquisition of moisture diffusion rate and thermal conductivity change rate during baking is as follows: the moisture diffusion rate is obtained through the first derivative of the mass loss curve, and the thermal conductivity change rate is determined by comparing the slopes of the material's heat flow response curves at different temperatures; the Arrhenius equation fitting process involves taking the natural logarithm of the measured moisture diffusion rate and the corresponding temperature and then performing a linear regression, with the slope reflecting the activation energy of moisture migration and the intercept characterizing the pre-exponential factor; when establishing a quantitative relationship between moisture migration and heat conduction, the activation energy of moisture migration obtained from the Arrhenius equation and the thermal conductivity change rate are subjected to multiple linear regression to obtain the proportionality coefficient and coupling constant of the moisture diffusion rate and the thermal conductivity change rate, ultimately forming a quantitative relationship describing the change in heat conduction driven by moisture migration during baking.

[0076] Based on the quantitative relationship between moisture migration and heat conduction, the finite volume method (FVM) is used to predict the regional temperature gradient of the baking equipment. The expression is as follows:

[0077]

[0078] in, It is the total number of heat transfer channels dominated by moisture migration. It is the index variable of the heat transfer channel dominated by moisture migration. It is the first Mass and heat transfer coupling coefficients for each heat transfer channel (range: 0.01-1.0). It is the TGA-FTIR number Each heat transfer channel is at the current real-time temperature of the material. The temperature-dependent moisture diffusion rate was measured below. It is the first Time scale of moisture migration characteristics in each heat transfer channel It is the first Apparent enthalpy change of moisture migration in each heat transfer channel;

[0079] Furthermore, the finite volume method determines the effective working volume of the baking equipment. The volume was divided into control volume elements, and the temperature-dependent moisture diffusion rate measured by TGA-FTIR was assigned at the center node of each element. Mass and heat transfer coupling coefficient Moisture migration characteristics time scale and apparent enthalpy change of water migration Based on the current baking time and real-time temperature Calculate the exponential decay term within each unit. The instantaneous value is obtained, and the spatial curvature of the temperature distribution is solved using a second-order central difference scheme. The summation of the exponential decay terms of all heat transfer channels within each control volume unit is then combined with the spatial curvature. Multiply the results and integrate the flux on the element surface using the Gaussian divergence theorem; sum the integral results of all control volume elements to obtain the regional temperature gradient of the baking equipment. This gradient value reflects the dynamic coupling effect between the water migration process and the heat conduction effect.

[0080] It should be noted that the first Apparent enthalpy change of moisture migration in each heat transfer channel This refers to the energy required for moisture to be removed from the material during the starch gelatinization and setting / maturation stages, obtained by scanning the heating rate of the TGA weight loss curve, with a value ranging from 0 to 120 kJ / mol (starch gelatinization stage). =1 is taken as 40–80 kJ / mol, during the conditioning and maturation stage. =2 is taken as 80–120 kJ / mol).

[0081] When the material is in the glass transition stage, the regional temperature gradient of the baking equipment is predicted through a slow thermal relaxation channel.

[0082] When the material is in the starch gelatinization stage and the shaping and ripening stage, the temperature gradient of the baking equipment is predicted by the rapid heat transfer channel.

[0083] Furthermore,

[0084] S3. Based on the regional temperature gradient of the baking equipment, the microwave power adjustment and wind speed ratio adjustment of the baking equipment are calculated through a multi-parameter coupled control algorithm, and a composite control instruction set is generated.

[0085] Based on the regional temperature gradient of the baking equipment, the microwave power adjustment and wind speed ratio adjustment are calculated by a multi-parameter coupled control algorithm.

[0086] The expression for microwave power adjustment is:

[0087] ;

[0088] in, It is in time Microwave power adjustment at that time It is the initial microwave power. It is the weighting coefficient of the regional temperature gradient amplitude on the microwave power adjustment (the value ranges from 0.1 to 0.5). It is the weighting coefficient of the regional temperature gradient change rate (with a value range of 0.05-0.3). It is in time Regional temperature gradient over time;

[0089] Furthermore, the multi-parameter coupled control algorithm first obtains the time... Regional temperature gradient Calculate the magnitude of the temperature gradient in the region. and its rate of change over time The weighting coefficient of the regional temperature gradient magnitude to the microwave power adjustment amount. (Value range 0.1-0.5) Weighting coefficient with the rate of change of regional temperature gradient (Values ​​range from 0.05 to 0.3) Multiply by the corresponding gradient magnitude and rate of change respectively, and then add the initial microwave power. Time was obtained later Microwave power adjustment at time .

[0090] The expression for wind speed ratio adjustment is:

[0091] ;

[0092] in, It is in time Wind speed ratio adjustment amount at time It is the baseline wind speed ratio coefficient. It is the weighting coefficient of the regional temperature gradient amplitude on the wind speed ratio adjustment (the value ranges from 0.05 to 0.25). It is the lag adjustment weighting coefficient of the cumulative effect of regional temperature gradient on wind speed ratio (with a value range of 0.01-0.1).

[0093] Furthermore, the multi-parameter coupled control algorithm first acquires time data in real time. Regional temperature gradient Calculate the magnitude of the temperature gradient in the region. The trapezoidal numerical integration method was used to integrate the data from time 0 to... The regional temperature gradient magnitude within the interval is obtained by cumulative integration. The weighting coefficient of the regional temperature gradient amplitude to the wind speed ratio adjustment. (Value range 0.05-0.25) Multiplied by the magnitude of the regional temperature gradient, this is the lag adjustment weighting coefficient of the cumulative effect of the regional temperature gradient on the wind speed ratio. (Value range 0.01-0.1) Multiply by the cumulative integral value; then combine the weighted result of the two items with the reference wind speed ratio coefficient. Perform linear combination and output time Wind speed ratio adjustment amount .

[0094] Based on microwave power adjustment and wind speed ratio adjustment, a composite control instruction set is generated using a linear interpolation mapping method.

[0095] Furthermore, the microwave power adjustment and wind speed ratio adjustment are first normalized to the same control parameter range. Based on the historical optimal combination database of baking equipment control parameters, the coordinate position of the current adjustment combination is located in the two-dimensional parameter space. The bilinear interpolation algorithm is used to calculate the execution parameter combination corresponding to the four adjacent optimal control points. The parameters are then weighted and fused according to the distance weight to generate a composite control instruction set containing the microwave generator output power curve and the fan speed curve.

[0096] S4. Obtain the variation range of material color difference and texture data, dynamically generate pulse cooling and gradient heating compensation commands through a fuzzy PID compensator, and dynamically correct the composite control command set.

[0097] Texture data refers to the hardness, elasticity, adhesion, and chewiness of materials as measured by a physical property analyzer.

[0098] Material color difference refers to the difference in CIE Lab color space between the material color and the standard color swatch;

[0099] Based on real-time material color difference and texture data, a sliding window statistical method is used to identify the range of variation of material color difference and texture data.

[0100] Furthermore, the sliding window statistical method traverses the real-time material color difference and texture data sequence with a fixed time window length. Within each window, it calculates the mean and standard deviation of the material color difference, as well as statistical measures such as kurtosis and skewness of the texture data. By comparing the differences in statistical measures between adjacent windows, it identifies the fluctuation range of the ΔE value of the color difference in the Lab color space and the range of the coefficient of variation of the texture data (such as hardness and elastic modulus). The window length is dynamically adjusted to capture the parameter drift characteristics at different time scales, and finally outputs the range of changes in the material color difference and texture data. For example, when the standard deviation of the material color difference ΔE value increases by more than 30% within three consecutive windows, it is determined to be an abnormal surface browning rate. At this time, the output color difference variation range is ΔE = 2.5-4.0, and the range of the coefficient of variation of texture hardness is adjusted to 0.15-0.25.

[0101] A fuzzy clustering algorithm is used to divide the variation range of material color difference and texture data into multiple fuzzy subsets, and an IF-THEN control rule is established for each fuzzy subset.

[0102] Furthermore, the fuzzy clustering algorithm first inputs the joint distribution data of color difference and texture data (hardness, adhesion) into the fuzzy C-means clustering process. Through iterative calculation, it optimizes the membership matrix and cluster center positions, dividing the data space into several fuzzy subsets. Each fuzzy subset uses a Gaussian membership function to accurately describe the degree of belonging of color difference-texture data. For example, color difference = in the range of 3.0-5.0 and hardness = in the range of 50-70N constitutes a typical fuzzy subset. Based on the centroid characteristics and boundary conditions of each fuzzy subset, a complete IF-THEN control rule base is constructed, such as "if color difference = belongs to the 3.0-5.0 fuzzy subset and hardness belongs to the 50-70N fuzzy subset, then output the cooling intensity adjustment coefficient 0.5".

[0103] Based on material color difference and texture data, pulse cooling and gradient heating compensation commands are generated through the IF-THEN control rules of each fuzzy subset.

[0104] Furthermore, after the real-time material color difference and texture data are input into the fuzzy subsets divided by the fuzzy clustering algorithm, the membership weight of each subset is calculated. Based on the logical condition matching degree of the corresponding subset in the IF-THEN control rules, the output of multiple rules is fused using the weighted average method. When the material color difference ΔE value belongs to the high-deviation fuzzy subset, a pulse cooling command is triggered, which maps the degree of color difference deviation to a step increment of the cooling fan duty cycle. When the texture parameters (hardness, elastic modulus) belong to the non-ideal state fuzzy subset, a gradient heating compensation command is generated, which calculates the gradual compensation amount of microwave power based on the texture deviation. After defuzzification using the max-min inference method, the final output is a pulse cooling and gradient heating compensation command that can directly control the actuator.

[0105] By using a fuzzy PID compensator, the pulse cooling and gradient heating compensation commands are superimposed and corrected with the composite control command set, and the corrected composite control command set is used to control the baking process in real time.

[0106] Furthermore, the fuzzy PID compensator receives pulse cooling and gradient heating compensation commands, which are superimposed on the microwave power adjustment and wind speed ratio adjustment in the composite control command set. The proportional-integral-derivative parameters are dynamically adjusted according to the real-time error of color difference and texture data. For example, when the color difference ΔE error is large, the weight of the proportional term is increased, and when the texture parameter has accumulated deviation, the effect of the integral term is enhanced. The corrected composite control command set synchronously adjusts the output power of the microwave generator and the fan speed through the actuator to achieve coordinated control of temperature and material state during the baking process.

[0107] This embodiment also provides a food baking processing control system, including: a phase change identification module, a temperature gradient identification module, an instruction generation module, and an instruction correction module; the phase change identification module is used to collect the dielectric properties of the material and capture the dynamic changes in the dielectric loss tangent, while using time-domain dielectric spectrum analysis to identify the phase change stage of the material in real time; the temperature gradient identification module is used to construct a dual-channel heat conduction model and dynamically identify the regional temperature gradient of the baking equipment according to the phase change stage of the material; the instruction generation module is used to calculate the microwave power adjustment and wind speed ratio adjustment of the baking equipment based on the regional temperature gradient of the baking equipment through a multi-parameter coupled control algorithm, and generate a composite control instruction set; the instruction correction module is used to obtain the variation range of the material color difference and texture data, and use a fuzzy PID compensator

[0108] This embodiment also provides a computer device applicable to the food baking processing procedure control method, 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 implement the food baking processing procedure control method proposed in the above embodiment.

[0109] 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.

[0110] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the food baking processing program control method 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.

[0111] In summary, this invention achieves real-time identification of material phase transition stages through time-domain dielectric spectroscopy analysis, and accurately analyzes the correlation between relaxation time distribution and phase transition characteristic spectra using inverse Laplace transform and dynamic time warping algorithms, thus realizing the dynamic capture of key phase transition nodes such as glass transition and starch gelatinization. Furthermore, by constructing a dual-channel heat conduction model, and based on Prony series fitting of molecular chain motion characteristics and TGA-FTIR determination of moisture migration patterns, a coupled computational framework for slow thermal relaxation and rapid heat transfer is established, achieving precise matching between phase transition stages and temperature gradient prediction.

[0112] 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 method for controlling a food baking process, characterized in that: include, The dielectric properties of the material are collected and the dynamic changes in the dielectric loss tangent are captured. At the same time, the time-domain dielectric spectroscopy analysis method is used to identify the phase transition stage of the material in real time. A dual-channel heat conduction model was constructed, and the regional temperature gradient of the baking equipment was dynamically identified based on the material phase change stage. The steps are as follows. The dual-channel heat conduction model includes a slow thermal relaxation channel and a fast heat transfer channel; The slow thermal relaxation channel was analyzed using time-domain dielectric relaxation spectroscopy to identify the hysteresis characteristics of molecular chain motion. The relaxation activation energy and thermal history temperature characteristics of the material were calibrated by differential scanning calorimetry, and a slow thermal relaxation time spectrum was constructed by Prony series fitting. By using the finite element discretization method, the slow thermal relaxation time spectrum is spatially meshed and solved to predict the regional temperature gradient of the baking equipment. The rapid heat transfer channel uses TGA-FTIR to measure the rate of moisture diffusion and the rate of change of thermal conductivity during baking. A quantitative relationship between moisture migration and heat conduction is established by fitting the Arrhenius equation. At the same time, the finite volume method is used to predict the regional temperature gradient of the baking equipment. Materials at different phase change stages are processed using a dual-channel heat conduction model consisting of a slow thermal relaxation channel and a fast heat transfer channel. Based on the regional temperature gradient of the baking equipment, a multi-parameter coupled control algorithm is used to calculate the microwave power adjustment and wind speed ratio adjustment of the baking equipment, and generate a composite control instruction set. The range of variation of material color difference and texture data is obtained, and pulse cooling and gradient heating compensation commands are dynamically generated through a fuzzy PID compensator. The composite control command set is then dynamically corrected.

2. The food baking processing procedure control method as described in claim 1, characterized in that: The dynamic change of the dielectric loss tangent refers to obtaining the dielectric loss tangent by frequency domain dielectric spectrum analysis based on the dielectric constant and loss factor, and then using the time-temperature superposition principle to perform dynamic tracking analysis to obtain the dynamic change curve of the dielectric loss tangent.

3. The food baking processing procedure control method as described in claim 2, characterized in that: The method of using time-domain dielectric spectroscopy to identify the phase transition stage of materials in real time includes the following steps. Based on the dynamic change curve of the dielectric loss tangent, the phase transition characteristics of the material are extracted using the sliding window difference method, and the phase transition characteristic spectrum is generated by the dynamic time warping algorithm. The time-domain dielectric relaxation current signal of the material is subjected to inverse Laplace transform to generate a relaxation time distribution that is time-aligned with the phase transition characteristic spectrum. The minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum was calculated using time-domain dielectric spectroscopy analysis. Low and high deviation thresholds are defined and compared with the minimum cumulative difference between the relaxation time distribution and the phase transition characteristic spectrum to identify the glass transition stage, starch gelatinization stage, and setting and maturation stage of the material.

4. The food baking processing procedure control method as described in claim 1, characterized in that: The generation of the composite control instruction set refers to calculating the microwave power adjustment and wind speed ratio adjustment based on the regional temperature gradient of the baking equipment through a multi-parameter coupled control algorithm, and generating the composite control instruction set through a linear interpolation mapping method.

5. The food baking processing procedure control method as described in claim 1, characterized in that: The steps for obtaining the variation range of material color difference and texture data, and dynamically generating pulse cooling and gradient heating compensation commands through a fuzzy PID compensator, are as follows: Based on real-time material color difference and texture data, a sliding window statistical method is used to identify the range of variation of material color difference and texture data. A fuzzy clustering algorithm is used to divide the variation range of material color difference and texture data into multiple fuzzy subsets, and an IF-THEN control rule is established for each fuzzy subset. Based on material color difference and texture data, pulse cooling and gradient heating compensation commands are generated through the IF-THEN control rules of each fuzzy subset.

6. The food baking processing procedure control method as described in claim 5, characterized in that: The dynamic correction of the composite control instruction set refers to superimposing and correcting the pulse cooling and gradient heating compensation instructions and the composite control instruction set using a fuzzy PID compensator, and then using the corrected composite control instruction set to control the baking process in real time.

7. A food baking processing procedure control system, based on the food baking processing procedure control method according to any one of claims 1 to 6, characterized in that: It includes a phase change recognition module, a temperature gradient recognition module, an instruction generation module, and an instruction correction module; The phase transition identification module is used to collect the dielectric properties of materials and capture the dynamic changes in the dielectric loss tangent. At the same time, it uses time-domain dielectric spectrum analysis to identify the phase transition stage of materials in real time. The temperature gradient identification module is used to construct a dual-channel heat conduction model and dynamically identify the regional temperature gradient of the baking equipment based on the phase change stage of the material. The instruction generation module is used to calculate the microwave power adjustment and wind speed ratio adjustment of the baking equipment based on the regional temperature gradient of the baking equipment through a multi-parameter coupled control algorithm, and generate a composite control instruction set. The instruction correction module is used to obtain the variation range of material color difference and texture data, dynamically generate pulse cooling and gradient heating compensation instructions through a fuzzy PID compensator, and dynamically correct the composite control instruction set.

8. 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 food baking processing program control method according to any one of claims 1 to 6.

9. 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 food baking processing program control method according to any one of claims 1 to 6.

Images