Frequency domain fitting-based lycopene content parameter prediction method, medium and electronic equipment

Abstract

The invention discloses a lycopene content parameter prediction method based on frequency domain fitting, which mainly comprises the following steps of: correcting a lycopene content coefficient and related physical quantity, constructing a function to carry out frequency domain fitting, convexly optimizing a characteristic value function, carrying out calculation simplification on a characteristic value function reconstruction model in a model error compensation function expression, and finally, carrying out frequency domain fitting. And performing frequency domain expansion on the basis of performing characteristic analysis on the established lycopene content model, converting the learning process of the lycopene content model into a convex optimization problem, and finally obtaining a controllable lycopene content parameter optimization model. According to the method, the limitation conditions of four physical quantities including lightness, color direction, hue and saturation and the relation between the limitation conditions and characteristic value functions are considered at the same time, and a method is provided for convenient determination and nondestructive testing of lycopene.

Description

Technical Field

[0001] The invention relates to the technical field of food detection, in particular to a lycopene content parameter optimization model method established by a frequency domain fitting method. Background Art

[0002] Lycopene is widely found in tomatoes, tomato products, watermelon, grapefruit and other fruits. It is the main pigment in ripe tomatoes, a carotenoid found in plant foods, and one of the most common carotenoids. Lycopene has deep red needle-shaped crystals, which not only have strong antioxidant activity, anti-cancer and anti-cancer effects, but also have the advantages of being soluble in chloroform, benzene and oils but insoluble in water.

[0003] The main methods for predicting lycopene content parameters include traditional chemical determination, predicting lycopene content through the regression equation of color difference value on lycopene content, and using machine learning algorithms to predict lycopene based on color difference value. Among these methods, the ones with strong generalization ability and small sample learning characteristics can effectively avoid problems such as overfitting, local minimization, and multidimensionality. The learning ability and optimization ability depend largely on the selection of parameters, and different input variables will correspond to different optimal parameters.

[0004] For example, the literature [Ke Jiansai, Yang Feiping, Jin Yezhou, Jiang Wuyi. Determination of lycopene content in tomatoes by ultra-high performance liquid chromatography [J]. Zhejiang Agricultural Sciences, 2024, 65 (7): 1683-1686.] uses traditional chemical methods for determination. This method establishes a method for determining the lycopene content in tomatoes using ultra-high performance liquid chromatography, which can save a lot of time and reagents, making pretreatment simpler, more economical, and more environmentally friendly. The method for determining the lycopene content in tomatoes disclosed in publication number CN101750284A calculates and predicts the lycopene content in tomatoes by establishing a kinetic equation of lycopene based on the lightness value and temperature of tomato fruit, thereby providing an important reference for the picking, storage and transportation conditions of tomatoes. This method has a good implementation method, but fails to take into account the limiting conditions of four physical quantities such as lightness, color direction, hue and saturation at the same time. The spectral detection method for detecting the lycopene content in agricultural products disclosed in publication number CN115046961A is implemented by a spectrometer, which includes: a light source device, a focusing device, a chopping device, a filtering device, a detection device and a processing device. This method mainly tends to be a method of equipment and instruments.

[0005] Based on this, the present invention proposes to use frequency domain fitting in the method for optimizing the lycopene content parameter model, which can avoid the above problems to a limited extent. At the same time, no patents and literature on the application of frequency domain fitting in the prediction of lycopene content are found. In summary, the existing lycopene content parameter prediction methods are restricted by the prediction accuracy and estimation speed, and the convex optimization problem applied to frequency domain fitting has not been applied to the accurate prediction of lycopene content. Summary of the invention

[0006] The present invention aims to solve one of the technical problems existing in the related art at least to a certain extent.

[0007] The purpose of the present invention is to provide a frequency domain fitting method for predicting lycopene content parameters, ensure the accuracy of lycopene content prediction, and establish a parameter optimization model that can simultaneously take into account multiple physical quantity constraints.

[0008] In order to achieve the above-mentioned object, the present invention provides a method for predicting lycopene content parameters by frequency domain fitting, comprising the following steps:

[0009] S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content;

[0010] S2. According to the corrected lycopene content coefficient, a constructor is used to fit the relevant physical quantities to obtain a model error compensation function for compensating for the model error in the lycopene content prediction model;

[0011] S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and expanding the eigenvalue function in the frequency domain through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function;

[0012] S4. Reconstructing the convex optimization problem model of the eigenvalue correction function, and obtaining the lycopene content prediction reconstruction model based on frequency domain fitting;

[0013] S5. The characteristic value correction function and the lycopene content prediction reconstruction model are adjusted to be in an iterative form with the related physical quantities, so as to obtain a lycopene content parameter optimization model, which is equivalent to the actual lycopene content prediction model.

[0014] A further preferred technical solution of the present invention is that, in step S1, the color difference of tomatoes of different maturity is measured, and the lycopene content coefficient and related physical quantities affecting the determination of lycopene content are corrected; the specific method is:

[0015] S11. Use a colorimeter to measure the color difference of tomatoes of different maturity levels;

[0016] S12, according to the lightness and color direction analysis of tomatoes; the lightness reference value is changed from the initial value a - Corrected to target value a + , the color direction reference value is changed from the initial value b - Corrected to target value b + The hue reference value is determined by the initial value F - Corrected to target value F + , the saturation reference value is determined by the initial value δ - Corrected to target value δ + ;

[0017] S13, according to the color difference amplitude λ of tomatoes at different maturity stages, four physical quantity parameter values ​​of brightness reference value a, color direction reference value b, hue reference value F and saturation reference value δ are adjusted in real time to modify the lycopene content coefficient K to K * :

[0018]

[0019] Where ξ is the slack variable.

[0020] As a preferred embodiment, the lycopene content coefficient K generated by defining the four physical quantities of brightness, color direction, hue and saturation is consistent with the variation law of color difference of tomatoes at different maturity stages. The lycopene content coefficient K is defined as:

[0021]

[0022] Where η is the number of smoothing points and n is the number of samples.

[0023] Preferably, the model error compensation function for compensating the model error in the lycopene content prediction model in step S2 is expressed as:

[0024]

[0025] In the formula, K * is the corrected lycopene content coefficient; r is the correlation coefficient; μ is the number of explanatory variables; σ is the standard estimation variance; t is the time constant; θ(t) is the linear insensitive loss function; dθ is the differential expression of θ;

[0026] y(a, b, F, δ) is the eigenvalue function, where a is the tomato brightness reference value, b is the tomato color direction reference value, F is the tomato hue reference value, and δ is the tomato saturation reference value.

[0027] Preferably, in step S3, the eigenvalue function introduced in the model error compensation function is constructed into a structure expression, and the eigenvalue function is expanded in the frequency domain by Fourier transform, and its learning process is converted into a convex optimization problem to obtain the eigenvalue correction function; the specific method is:

[0028] S31. Construct a function expression for the eigenvalue function y(a, b, F, δ) in the model error compensation function expression, and establish the relationship between the four physical quantities of brightness, color direction, hue and saturation and the eigenvalue function:

[0029]

[0030] In the formula, x i Indicates the i-th parameter variable in the sample space: a(x i )、b(x i )、F(x i ) and (x i ) represent the i-th optimized value of brightness, color direction, hue and saturation in the sample space respectively;

[0031] S32. The eigenvalue function y(a, b, F, δ) is expanded in the frequency domain by Fourier transform, and its learning process is converted into a convex optimization problem. The expression of the eigenvalue correction function is:

[0032]

[0033] In the formula, a n (K * ) is the difference function of the corrected lycopene content coefficient; τ i is the Lagrange multiplier in quadratic programming; N is the total number of samples; θ(K * ) is the coefficient of each harmonic term in the Fourier decomposition expansion, and its magnitude is only related to the corrected lycopene content coefficient K * When the corrected lycopene content coefficient K * When it increases to a certain extent, its eigenvalue correction function decreases sharply, thus obtaining the local optimal solution.

[0034] Preferably, the lycopene content prediction reconstruction model expression obtained in step S4 is:

[0035]

[0036] Where, T i (x) represents the i-th kernel parameter in the sample space; cosx is the cosine function of x, and dx represents the differential form of x.

[0037] Preferably, the lycopene content parameter optimization model expression obtained in step S5 is:

[0038]

[0039] In the formula, is the t-th iteration speed of the i-th variable; is the t-th iteration state of the i-th variable, and ω represents the inertia weight, which is used to measure the speed of parameter optimization.

[0040] In another aspect, the present invention provides a non-transitory computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions enable a computer to execute the above-mentioned method for predicting lycopene content parameters by frequency domain fitting.

[0041] Yet another aspect of the present invention provides an electronic device, comprising: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus, and the processor calls logic instructions in the memory to execute the above-mentioned frequency domain fitting lycopene content parameter prediction method.

[0042] In yet another aspect, the present invention provides a computer program product, comprising a computer program stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer executes the above-mentioned frequency domain fitting method for predicting lycopene content parameters.

[0043] Beneficial effects: The lycopene content parameter optimization model established by the frequency domain fitting method of the present invention eliminates the model error, is equivalent to the actual lycopene content prediction model, achieves accurate prediction of lycopene content, and provides a prediction method for convenient determination and non-destructive detection of lycopene;

[0044] Compared with the lycopene content prediction method based on particle swarm optimization, the lycopene content parameter prediction method based on the present invention takes into account the limiting conditions of four physical quantities such as brightness, color direction, hue and saturation, thereby solving the problems of many factors affecting the prediction process, mutual interference of parameter sensitivity, and difficulty in establishing an accurate optimization model.

[0045] Compared with the lycopene content prediction method of support vector machine, the present invention reconstructs the model of the convex optimization problem of the eigenvalue correction function so that the obtained large-scale sample parameters can converge within a limited time, which can significantly improve the lycopene content prediction response speed and reduce the parameter optimization speed, effectively improve the search efficiency, and reduce the serious mutual correlation coupling between the variables. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The flowchart of the method for predicting lycopene content parameters by frequency domain fitting;

[0047] Figure 2 The figure is a comparison diagram between the prediction result obtained by the prediction method of the present invention and the calibration result. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0049] Combine the following Figure 1-Figure 2 The present invention describes the lycopene content parameter prediction method, medium and electronic device based on frequency domain fitting.

[0050] Example 1: This example provides a method for predicting lycopene content parameters by frequency domain fitting.

[0051] This embodiment mainly aims at analyzing the measurement error of the lycopene content prediction model. The method of the present invention mainly includes the following steps: Figure 1 As shown:

[0052] The lycopene content coefficient and related physical quantities are corrected, the constructor is used for frequency domain fitting, the eigenvalue function is convexly optimized, and the eigenvalue function reconstruction model in the model error compensation function expression is simplified. Finally, based on the characteristic analysis of the established lycopene content model, the frequency domain is expanded, and its learning process is converted into a convex optimization problem, and finally a controllable lycopene content parameter optimization model is obtained.

[0053] The method proposed in the present invention focuses more on the method of combining instruments with algorithms, which can more effectively predict the lycopene content, save costs, and further improve efficiency.

[0054] The lycopene content parameter prediction method based on frequency domain fitting of the present invention specifically comprises the following steps:

[0055] S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content;

[0056] S11. Use a colorimeter to measure the color difference of tomatoes of different maturity levels;

[0057] S12, according to the lightness and color direction analysis of tomatoes; the lightness reference value is changed from the initial value a - Corrected to target value a + , the color direction reference value is changed from the initial value b - Corrected to target value b + The hue reference value is determined by the initial value F- Corrected to target value F + , the saturation reference value is determined by the initial value δ - Corrected to target value δ + ;

[0058] S13. After the brightness reference value, color direction reference value, hue reference value and saturation reference value are corrected, the lycopene content coefficient cannot be fully applied to the determination of lycopene content and needs to be readjusted.

[0059] Since the lycopene content coefficient K generated by the four physical quantities of brightness, color direction, hue and saturation and the color difference of tomatoes at different maturity stages are considered to be consistent in this embodiment, the lycopene content coefficient K is defined as:

[0060]

[0061] In the formula, η is the number of smoothing points, and n represents the number of samples. The target variable is the determination of lycopene content, which means that the four physical quantities of brightness, color direction, hue, and saturation are characteristic color attributes related to lycopene. Accurately and quickly obtaining the parameter convergence time of the lycopene content coefficient K is the key to combining model simplification with accuracy.

[0062] S14. In order to obtain the parameter convergence time, according to the color difference amplitude λ of tomatoes at different maturity stages, the four physical quantity parameter values ​​of brightness reference value a, color direction reference value b, hue reference value F and saturation reference value δ are adjusted in real time, and the lycopene content coefficient K is modified to K * :

[0063]

[0064] Where ξ is the slack variable.

[0065] S2, according to the corrected lycopene content coefficient K * , the constructor fits the four physical quantities of brightness, color direction, hue and saturation that affect lycopene, and obtains the model error compensation function that compensates the model error in the lycopene content prediction model; the model error compensation function expression is:

[0066]

[0067] Where r is the correlation coefficient; μ is the number of explanatory variables; σ is the standard estimation variance; t is the time constant; θ(t) is the linear insensitive loss function; dθ is the differential expression of θ; y(a, b, F, δ) is the eigenvalue function.

[0068] S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and expanding the eigenvalue function in the frequency domain through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function;

[0069] S31. From the calculation process of the model error compensation function, it can be seen that since the eigenvalue function y(a, b, F, δ) is added to the lycopene content model error compensation function, the calculation difficulty is increased, and it is necessary to update and optimize the eigenvalue function based on the lycopene content model error compensation function. At the same time, in order to simplify the complexity of the calculation of the model error compensation function and to facilitate the establishment of the relationship between the four physical quantities of brightness, color direction, hue and saturation and the eigenvalue function, the eigenvalue function constructor expression introduced in the model error compensation function is as follows:

[0070]

[0071] In the formula, x i Indicates the i-th parameter variable in the sample space: a(x i )、b(x i )、F(x i ) and (x i ) represent the i-th optimized value of brightness, color direction, hue and saturation in the sample space respectively;

[0072] S32. The eigenvalue function y(a, b, F, δ) is expanded in the frequency domain by Fourier transform, and its learning process is converted into a convex optimization problem. The expression of the eigenvalue correction function is:

[0073]

[0074] In the formula, a n (K * ) is the difference function of the corrected lycopene content coefficient; τ i is the Lagrange multiplier in quadratic programming; N is the total number of samples; θ(K * ) is the coefficient of each harmonic term in the Fourier decomposition expansion, and its magnitude is only related to the corrected lycopene content coefficient K * When the corrected lycopene content coefficient K * When it increases to a certain extent, its eigenvalue correction function decreases sharply, thus obtaining the local optimal solution.

[0075] S4. Reconstruct the convex optimization problem model of the eigenvalue correction function, and obtain the lycopene content prediction reconstruction model based on frequency domain fitting; the lycopene content prediction reconstruction model expression is:

[0076]

[0077] Where, T i (x) represents the i-th kernel parameter in the sample space; cosx is the cosine function of x, and dx represents the differential form of x.

[0078] S5. The realization of brightness, color direction, hue and saturation control is the key to the reconstruction of the lycopene content prediction model. The eigenvalue correction function and the lycopene content prediction reconstruction model are adjusted to the form of iteration with brightness, color direction, hue and saturation, and the lycopene content parameter optimization model is obtained as follows:

[0079]

[0080] In the formula, is the t-th iteration speed of the i-th variable; is the t-th iteration state quantity of the i-th variable, ω represents the inertia weight, which is used to measure the speed of parameter optimization;

[0081] The lycopene content parameter optimization model established by frequency domain fitting eliminated the model error and was equivalent to the actual lycopene content prediction model, achieving accurate prediction of lycopene content. Figure 2 shown.

[0082] Embodiment 2: This embodiment provides a non-transitory computer-readable storage medium, on which computer instructions are stored, the computer instructions causing the computer to execute a lycopene content parameter prediction method based on frequency domain fitting, the method comprising the following steps:

[0083] S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content;

[0084] S2. According to the corrected lycopene content coefficient, a constructor is used to fit the relevant physical quantities to obtain a model error compensation function for compensating for the model error in the lycopene content prediction model;

[0085] S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and expanding the eigenvalue function in the frequency domain through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function;

[0086] S4. Reconstructing the convex optimization problem model of the eigenvalue correction function, and obtaining the lycopene content prediction reconstruction model based on frequency domain fitting;

[0087] S5. The characteristic value correction function and the lycopene content prediction reconstruction model are adjusted to be in an iterative form with the related physical quantities, so as to obtain a lycopene content parameter optimization model, which is equivalent to the actual lycopene content prediction model.

[0088] Embodiment 3: This embodiment provides an electronic device, which may include: a processor, a communications interface, a memory, and a communications bus, wherein the processor, the communications interface, and the memory communicate with each other via the communications bus. The processor may call the logic instructions in the memory to execute the lycopene content parameter prediction method based on frequency domain fitting, and the method includes the following steps:

[0089] S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content;

[0090] S2. According to the corrected lycopene content coefficient, a constructor is used to fit the relevant physical quantities to obtain a model error compensation function for compensating for the model error in the lycopene content prediction model;

[0091] S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and expanding the eigenvalue function in the frequency domain through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function;

[0092] S4. Reconstructing the convex optimization problem model of the eigenvalue correction function, and obtaining the lycopene content prediction reconstruction model based on frequency domain fitting;

[0093] S5. The characteristic value correction function and the lycopene content prediction reconstruction model are adjusted to be in an iterative form with the related physical quantities, so as to obtain a lycopene content parameter optimization model, which is equivalent to the actual lycopene content prediction model.

[0094] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0095] Embodiment 4: This embodiment provides a computer program product, which includes a computer program. The computer program can be stored in a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute a frequency domain fitting method for predicting lycopene content parameters. The method includes the following steps:

[0096] S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content;

[0097] S2. According to the corrected lycopene content coefficient, a constructor is used to fit the relevant physical quantities to obtain a model error compensation function for compensating for the model error in the lycopene content prediction model;

[0098] S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and expanding the eigenvalue function in the frequency domain through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function;

[0099] S4. Reconstructing the convex optimization problem model of the eigenvalue correction function, and obtaining the lycopene content prediction reconstruction model based on frequency domain fitting;

[0100] S5. The characteristic value correction function and the lycopene content prediction reconstruction model are adjusted to be in an iterative form with the related physical quantities, so as to obtain a lycopene content parameter optimization model, which is equivalent to the actual lycopene content prediction model.

[0101] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0102] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting lycopene content parameters by frequency domain fitting, characterized in that: The steps include: S1. Measure the color difference of tomatoes with different maturity, and correct the lycopene content coefficient and related physical quantities that affect the determination of lycopene content; S2. According to the corrected lycopene content coefficient, a constructor is used to fit the relevant physical quantities to obtain a model error compensation function for compensating for the model error in the lycopene content prediction model; S3, constructing a function expression of the eigenvalue function introduced in the model error compensation function, and performing frequency domain expansion of the eigenvalue function through Fourier transform, converting its learning process into a convex optimization problem, and obtaining the eigenvalue correction function; S4. Reconstructing the convex optimization problem model of the eigenvalue correction function, and obtaining the lycopene content prediction reconstruction model based on frequency domain fitting; S5. The characteristic value correction function and the lycopene content prediction reconstruction model are adjusted to be in an iterative form with the related physical quantities, so as to obtain a lycopene content parameter optimization model, which is equivalent to the actual lycopene content prediction model.

2. The method for predicting lycopene content parameters by frequency domain fitting according to claim 1, characterized in that: In step S1, the color difference of tomatoes of different maturity is measured, and the lycopene content coefficient and related physical quantities affecting the determination of lycopene content are corrected; the specific method is: S11. Use a colorimeter to measure the color difference of tomatoes of different maturity levels; S12, according to the lightness and color direction analysis of tomatoes; the lightness reference value is changed from the initial value a - Corrected to target value a + , the color direction reference value is changed from the initial value b - Corrected to target value b + The hue reference value is determined by the initial value F - Corrected to target value F + , the saturation reference value is determined by the initial value δ - Corrected to the target value δ + ; S13, according to the color difference amplitude λ of tomatoes at different maturity stages, four physical quantity parameter values ​​of brightness reference value a, color direction reference value b, hue reference value F and saturation reference value a are adjusted in real time, and the lycopene content coefficient K is modified to K * : Where ξ is the slack variable.

3. The method for predicting lycopene content parameters by frequency domain fitting according to claim 2, characterized in that: The lycopene content coefficient K generated by defining the four physical quantities of brightness, color direction, hue and saturation is consistent with the change law of color difference of tomatoes at different maturity stages. The lycopene content coefficient K is defined as: In the formula, η is the number of smoothing points and n is the number of samples.

4. The method for predicting lycopene content parameters by frequency domain fitting according to claim 1, characterized in that: The model error compensation function expression for compensating the model error in the lycopene content prediction model in step S2 is: In the formula, K * is the corrected lycopene content coefficient; r is the correlation coefficient; μ is the number of explanatory variables; σ is the standard estimation variance; t is the time constant; θ(t) is the linear insensitive loss function; dθ is the differential expression of θ; y(a, b, F, δ) is the eigenvalue function, where a is the tomato brightness reference value, b is the tomato color direction reference value, F is the tomato hue reference value, and δ is the tomato saturation reference value.

5. The method for predicting lycopene content parameters by frequency domain fitting according to claim 4, characterized in that: Step S3 constructs a function expression for the eigenvalue function introduced in the model error compensation function, and expands the eigenvalue function in the frequency domain through Fourier transform, converts its learning process into a convex optimization problem, and obtains the eigenvalue correction function; the specific method is: S31. Construct a function expression for the eigenvalue function y(a, b, F, δ) in the model error compensation function expression, and establish the relationship between the four physical quantities of brightness, color direction, hue and saturation and the eigenvalue function: In the formula, x i Indicates the i-th parameter variable in the sample space: a(x i )、b(x i )、F(x i ) and (x i ) represent the i-th optimized value of brightness, color direction, hue and saturation in the sample space respectively; S32. The eigenvalue function y(a, b, F, δ) is expanded in the frequency domain by Fourier transform, and its learning process is converted into a convex optimization problem. The expression of the eigenvalue correction function is: In the formula, a n (K * ) is the difference function of the corrected lycopene content coefficient; τ i is the Lagrange multiplier in quadratic programming; N is the total number of samples; θ(K * ) is the coefficient of each harmonic term in the Fourier decomposition expansion, and its magnitude is only related to the corrected lycopene content coefficient K * When the corrected lycopene content coefficient K * When it increases to a certain extent, its eigenvalue correction function decreases sharply, thus obtaining the local optimal solution.

6. The method for predicting lycopene content parameters by frequency domain fitting according to claim 5, characterized in that: The lycopene content prediction reconstruction model expression obtained in step S4 is: Where, T i (x) represents the i-th kernel parameter in the sample space; cosx is the cosine function of x, and dx represents the differential form of x.

7. The method for predicting lycopene content parameters by frequency domain fitting according to claim 6, characterized in that: The lycopene content parameter optimization model expression obtained in step S5 is: In the formula, is the t-th iteration speed of the i-th variable; is the t-th iteration state of the i-th variable, and ω represents the inertia weight, which is used to measure the speed of parameter optimization.

8. A non-transitory computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions enable a computer to execute the lycopene content parameter prediction method based on frequency domain fitting according to any one of claims 1 to 7.

9. An electronic device, comprising: A processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus, and the processor calls the logic instructions in the memory to execute the lycopene content parameter prediction method based on frequency domain fitting according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, wherein the computer program is stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer executes the lycopene content parameter prediction method based on frequency domain fitting according to any one of claims 1 to 7.

Images