Method and system for dynamic monitoring of polyphenol components in olive leaf based on principal component analysis

Abstract

The application provides a kind of dynamic monitoring method and system of olive leaf polyphenol components based on principal component analysis, it is related to computer technology field, comprising: obtaining original data;According to the original data, the spectrum interval is separated, and the purified polyphenol characteristic spectrum sequence is obtained;According to the purified polyphenol characteristic spectrum sequence, spectral deconvolution processing is carried out, and the polyphenol subcomponent concentration evolution matrix is obtained;According to the polyphenol subcomponent concentration evolution matrix, the principal component is reconstructed, and the reconstruction result is obtained;According to the reconstruction result, track fitting is carried out, by fitting the principal component score track into the kinetic network model containing hydroxytyrosol glycoside hydrolysis and oxidative condensation step, the fitting result is obtained;According to the fitting result, prediction is carried out, the evolution end point of hydroxytyrosol free state content and its derivative relative proportion is predicted, and the dynamic monitoring result is obtained.The application realizes the accurate monitoring and prediction of the dynamic conversion process of specific polyphenol components such as hydroxytyrosol glycoside and oleuropein in olive leaves.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to a method and system for dynamic monitoring of polyphenol components in olive leaves based on principal component analysis. Background Technology

[0002] Olive leaves are rich in a variety of bioactive polyphenols, such as oleuropein, hydroxytyrosol and their derivatives. These components have been proven to have significant health benefits such as antioxidant, anti-inflammatory and cardiovascular disease prevention, making them of great development value in the food, health products and pharmaceutical fields. In order to assess and ensure the quality stability and efficacy reliability of its raw materials and products, it is urgent to accurately monitor the dynamic transformation behavior of polyphenol components during processing or storage. Currently, existing technologies mainly use conventional spectroscopic or chromatographic methods to intermittently determine the total amount of polyphenols or a few marker components. These methods have significant limitations when applied to monitoring complex dynamic processes: on the one hand, the strong background interference from moisture, cellulose, etc. in the olive leaf matrix makes it difficult for existing methods to purely extract the characteristic information of various polyphenols from the mixed signals, resulting in coarse analysis of polyphenol component changes; on the other hand, polyphenol components undergo complex transformations such as glycoside hydrolysis and oxidative polymerization during processing and storage, while existing methods can usually only provide component data at individual time points, failing to clearly characterize the correlation and transformation pathways between specific subcomponents such as hydroxytyrosol glycosides and oleuropein, and even more difficult to extract transformation patterns that can predict future trends from these discrete data, thus limiting the accurate assessment and effective control of their stability and efficacy evolution.

[0003] Based on the shortcomings of the existing technologies, there is an urgent need for a method and system for dynamic monitoring of olive leaf polyphenol components based on principal component analysis. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for dynamic monitoring of olive leaf polyphenol components based on principal component analysis, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] In a first aspect, this application provides a method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis, including:

[0006] Obtain raw data, which is raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, including vibrational and electronic transition responses of polyphenol characteristic functional groups.

[0007] Based on the original data, spectral intervals were separated to obtain the characteristic spectral sequences of purified polyphenols.

[0008] Based on the purified polyphenol characteristic spectral sequence, spectral deconvolution processing is performed. By decomposing the time-series spectrum into the base spectrum of the polyphenol subcomponents and their respective concentration profiles over time, the concentration evolution matrix of the polyphenol subcomponents is obtained.

[0009] Principal component reconstruction is performed based on the polyphenol subcomponent concentration evolution matrix. The concentration ratio relationship between polyphenol subcomponents over time is used as a dynamic constraint on the covariance structure of the reconstructed data. Dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction results.

[0010] Based on the reconstruction results, the polyphenol conversion kinetic trajectory was fitted. The fitting result was obtained by fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides.

[0011] Based on the fitting results, predictions are made, and the evolution endpoint of the free hydroxytyrosol content and the relative proportion of its derivatives is predicted by extrapolating the kinetic model to the target storage or processing conditions, thus obtaining dynamic monitoring results.

[0012] Secondly, this application also provides a method and system for dynamic monitoring of olive leaf polyphenol components based on principal component analysis, including:

[0013] The acquisition module is used to acquire raw data, which is the raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, containing the vibration and electronic transition responses of polyphenol characteristic functional groups.

[0014] A separation module is used to perform spectral region separation based on the original data to obtain the characteristic spectral sequence of purified polyphenols.

[0015] The processing module is used to perform spectral deconvolution processing based on the purified polyphenol characteristic spectral sequence, and obtain the polyphenol sub-component concentration evolution matrix by decomposing the time-series spectrum into the base spectrum of the polyphenol subcomponent and its respective concentration profile over time.

[0016] The reconstruction module is used to perform principal component reconstruction based on the polyphenol subcomponent concentration evolution matrix. By using the concentration ratio relationship between polyphenol subcomponents evolving over time as a dynamic constraint condition for the covariance structure of the reconstructed data, dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction result.

[0017] The fitting module is used to fit the polyphenol conversion kinetic trajectory based on the reconstruction result. The fitting result is obtained by fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides.

[0018] The prediction module is used to make predictions based on the fitting results, extrapolate the dynamic model to the target storage or processing conditions, predict the evolution endpoint of the free content of hydroxytyrosol and the relative proportion of its derivatives, and obtain dynamic monitoring results.

[0019] The beneficial effects of this invention are as follows:

[0020] This invention achieves purer, more accurate, and quantifiable monitoring and prediction of the dynamic transformation process of specific polyphenol components such as hydroxytyrosol glycosides and oleuropein in olive leaves by purifying polyphenol characteristics from the original spectrum, accurately analyzing the independent concentration changes of polyphenol subcomponents, using their metabolic relationships to constrain and reconstruct the principal components to extract clear transformation trends, and constructing a quantitative kinetic model based on these trends. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating a method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis, as described in an embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of the structure of a dynamic monitoring system for olive leaf polyphenol components based on principal component analysis, as described in an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of the structure of a dynamic monitoring device for olive leaf polyphenol components based on principal component analysis, as described in an embodiment of the present invention.

[0025] Figure 4 Thermograph of polyphenol subcomponent concentration evolution;

[0026] Figure 5 A heatmap showing the concentration clustering of polyphenol subcomponents (with contour lines);

[0027] Figure 6 This is a curve showing the change in concentration of polyphenol subcomponents over time.

[0028] The diagram is labeled as follows: 800, a dynamic monitoring device for olive leaf polyphenol components based on principal component analysis; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component; 901, acquisition module; 902, separation module; 903, processing module; 904, reconstruction module; 905, fitting module; 906, prediction module. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0030] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0031] Example 1:

[0032] This embodiment provides a method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis.

[0033] See Figure 1 The figure shows that the method includes steps S100 to S600.

[0034] Step S100: Obtain raw data. The raw data is the raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, which includes the vibration and electronic transition response of polyphenol characteristic functional groups.

[0035] Step S200: Separate spectral regions based on the original data to obtain the characteristic spectral sequences of purified polyphenols;

[0036] Step S300: Perform spectral deconvolution processing based on the characteristic spectral sequence of purified polyphenols. By decomposing the time-series spectrum into the base spectrum of polyphenol subcomponents and their respective concentration profiles over time, the concentration evolution matrix of polyphenol subcomponents is obtained.

[0037] Step S400: Based on the polyphenol subcomponent concentration evolution matrix, principal component reconstruction is performed. By using the concentration ratio relationship between polyphenol subcomponents evolving over time as a dynamic constraint on the covariance structure of the reconstructed data, dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction results.

[0038] Step S500: Fit the polyphenol conversion kinetic trajectory based on the reconstruction results. By fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides, the fitting results are obtained.

[0039] Step S600: Based on the fitting results, make predictions and extrapolate the kinetic model to the target storage or processing conditions to predict the evolution endpoint of the free hydroxytyrosol content and the relative proportion of its derivatives, and obtain dynamic monitoring results.

[0040] Specifically, this embodiment provides a method for dynamic monitoring of polyphenol components in olive leaves, aiming to achieve precise tracking of component transformation through multi-level data processing, serving the need for stability control of active ingredients in the development of pharmaceuticals, cosmetics, and functional foods. Step S100 involves acquiring raw data, specifically including near-infrared and mid-infrared absorption spectral data of olive leaves collected at continuous monitoring time points. These data are captured in real time by a non-destructive spectrometer to capture the response signals of polyphenol characteristic functional groups, forming a matrix covering the changes in absorption intensity across the entire spectrum over time. The acquisition method simulates the actual planting or processing environment, achieved through periodic sampling or online monitoring, thereby providing a high-temporal-resolution basic dataset for subsequent analysis and supporting the non-destructive assessment of dynamic changes in components during resource utilization. Step S200 involves spectral interval separation, screening out characteristic intervals directly related to polyphenol functional groups from the raw full-band spectrum, eliminating interference from impurities such as proteins and sugars, and generating purified polyphenol characteristic spectral sequences. This process enhances the signal-to-noise ratio through mathematical algorithms, ensures data specificity, adapts to the complex composition of agricultural by-products, and improves monitoring accuracy. Step S300 performs spectral deconvolution processing, using algorithms such as multivariate curve resolution to decompose the time-series spectrum into the base spectra and concentration profiles of polyphenol subcomponents, obtaining a concentration evolution matrix. This processing intuitively presents the dynamics of subcomponents by analyzing overlapping spectral signals, highlighting its analytical capabilities in multicomponent interaction studies and providing a quantitative basis for transformation mechanism analysis. Step S400 performs principal component reconstruction, using the concentration ratio relationship between polyphenol subcomponents as a dynamic constraint, and performs covariance structure analysis on the concentration evolution matrix to extract dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation. This reconstruction captures the key features of the component transformation network, supporting the application of trend prediction in quality control. Step S500 completes the fitting of polyphenol transformation kinetic trajectories, fitting the principal component score trajectory into a kinetic network model including the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides. By quantifying the reaction rate constant, the actual transformation path is simulated, providing a theoretical basis for process optimization. Step S600 makes predictions based on the fitting results, using a kinetic model extrapolated to target storage or processing conditions (such as specific temperature and humidity) to predict the evolution endpoint of the free hydroxytyrosol content and the relative proportion of its derivatives, thus achieving a closed loop from data to application and improving the efficiency of industrial decision-making. The overall method combines spectroscopic techniques with kinetic modeling to achieve efficient and non-destructive monitoring of polyphenol component transformation, demonstrating significant practical value.

[0041] Example 2:

[0042] In a preferred embodiment, step S200 includes steps S210 to S230.

[0043] Step S210: Based on the original data, perform non-phenolic background absorption contribution separation processing. By identifying and subtracting the broadband background absorption bands attributed to the hydrogen bond stretching vibration of leaf water and the carbon-oxygen bond vibration of cellulose skeleton from the full-band spectrum, the intermediate spectral sequence is obtained.

[0044] Step S220: Based on the intermediate spectral sequence, the spectral range of polyphenol characteristic functional groups is extracted. By calculating the first derivative spectrum and locking the characteristic response ranges of carbon-carbon double bond skeletal vibration of benzene ring, oxygen-hydrogen bond stretching vibration of phenolic hydroxyl group and carbon-oxygen-carbon asymmetric stretching vibration of glycosidic bond, the candidate polyphenol characteristic spectral sequence is obtained.

[0045] Step S230: Based on the characteristic spectral sequences of candidate polyphenols, perform characteristic interval optimization and sequence integration processing. By introducing principal component loading vectors as weights, evaluate and weightedly fuse the signal-to-noise ratio and sensitivity to concentration changes of different characteristic intervals to obtain the characteristic spectral sequences of purified polyphenols.

[0046] Specifically, step S200 purifies the spectral data through three progressive sub-steps: First, step S210 performs non-phenolic background absorption contribution separation processing. Specifically, it identifies and subtracts broadband background absorption bands attributable to hydrogen bond stretching vibrations of leaf water and carbon-oxygen bond vibrations of the cellulose skeleton from the full-band spectrum. These interfering signals mainly correspond to the oxygen-hydrogen bond vibrations of water molecules and the carbon-oxygen bond vibrations of polysaccharide structures. Their contributions are eliminated through mathematical subtraction, resulting in an intermediate spectral sequence, thus initially highlighting polyphenol-related signals in the complex matrix of olive leaves. The core of the mathematical subtraction operation is to construct a mathematical model of the background spectrum and separate it from the original spectrum. The specific steps are as follows:

[0047] First, a baseline drift model is established through polynomial fitting, and then the optimal fitting curve is determined using the least squares method to eliminate broadband background interference. The mathematical expression for this process is:

[0048] ;

[0049] Where a is the polynomial coefficient vector, m is the polynomial order, and λ i Let y be the wave number at point i. i α is the absorbance, I is the set of baseline point indices, i is the index value of the baseline point, and α is the index value of the baseline point. j are the polynomial coefficients, where j is the summation index.

[0050] Next, weighted least squares method is used for accurate separation of background components, and adaptive weight adjustment is used to highlight the polyphenol feature signal. Its mathematical expression is:

[0051] ;

[0052] in, Let X be the background component weight vector, X be the reference spectral matrix, and W be the weight matrix. Here, is the spectrum after baseline correction, and T is the transpose of the matrix.

[0053] Finally, precise subtraction of background contribution is achieved through matrix operations to obtain the purified spectrum:

[0054] ;

[0055] in, For the final purified spectrum, This is the spectrum after baseline correction.

[0056] This method uses a progressive mathematical model to gradually eliminate background interference from moisture and cellulose in olive leaves, effectively improving the detection sensitivity of polyphenol characteristic signals.

[0057] Next, step S220 performs spectral region extraction processing for polyphenol characteristic functional groups based on the intermediate spectral sequence. First-derivative spectral calculations are used to pinpoint the characteristic response regions of the carbon-carbon double bond skeleton vibration of the benzene ring, the oxygen-hydrogen bond stretching vibration of the phenolic hydroxyl group, and the carbon-oxygen-carbon asymmetric stretching vibration of the glycosidic bond. First-derivative spectroscopy is a mathematical transformation that enhances spectral peak resolution. By sharpening overlapping peaks, it improves the specific capture of polyphenol functional groups, resulting in candidate polyphenol characteristic spectral sequences. This processing uses mathematical differentiation methods to sharpen overlapping spectral peaks, improving the accuracy of identifying characteristic absorption bands. The specific mathematical expression is as follows:

[0058] ;

[0059] Where D(λ) i ) represents the wave number λ i The first derivative value at point A, where A represents the absorbance value and λ represents the wavenumber, A(λ) i+1 ), A(λ i-1 ) represents λ i+1 and λ i-1 The absorbance value at the wavenumber, where d represents the differential sign. This calculation uses the central difference method to approximate differential operations, effectively enhancing the information on the rate of change of the spectral curve.

[0060] Feature interval locking is based on the identification of first derivative extrema:

[0061] ;

[0062] in, For the extracted feature interval wavenumber set, As a preset threshold, D(λ) i ) indicates that at wavenumber λ i The first derivative value calculated at that point, The first derivative D(λ) iThe derivative of (i.e., the second derivative) is used to locate the extreme points of the derivative.

[0063] Interval optimization is achieved through signal-to-noise ratio weighting:

[0064] ;

[0065] Where w(λ) i ) represents the wave number λ i The weighting factor at σ D (λ i ) is the standard deviation of the first derivative within the local window. This weight is used to evaluate the signal-to-noise ratio quality of each feature interval.

[0066] This method highlights the carbon-carbon double bond skeletal vibration of the benzene ring (approximately 1600 cm⁻¹) through differential calculations. -1 ), phenolic hydroxyl oxygen-hydrogen bond stretching vibration (approximately 3200-3600 cm). -1 ) and the carbon-oxygen-carbon asymmetric stretching vibration of glycosidic bonds (approximately 1000-1200 cm⁻¹). -1 The characteristic response of polyphenols can effectively separate overlapping spectral peaks, enhance the specific capture ability of polyphenol characteristic functional groups, and provide high-quality candidate characteristic spectral sequences for subsequent analysis.

[0067] Finally, step S230 performs feature interval optimization and sequence integration processing, introducing principal component loading vectors as weights. Principal component loading vectors are indicators that characterize the importance of variables in statistical analysis. The signal-to-noise ratio and sensitivity to concentration changes of different feature intervals are evaluated, and low-contribution intervals are eliminated through weighted fusion. Finally, purified polyphenol characteristic spectral sequences are generated. This optimization process adapts to the monitoring needs of dynamic changes in polyphenol components in olive leaves, ensuring the reliability and representativeness of the data in subsequent analyses.

[0068] Example 3:

[0069] In a preferred embodiment, step S300 includes steps S310 to S330.

[0070] Step S310: Based on the characteristic spectral sequence of purified polyphenols, perform preliminary base spectrum extraction of polyphenol subclasses. By calculating the correlation between the differences in spectra at each time point, separate the spectral components that have independent changing trends in the time domain to obtain the base spectrum of candidate polyphenol subclasses.

[0071] Step S320: Based on the candidate polyphenol subclass base spectra, perform base spectrum identification and purification based on the standard spectral library. By calculating the similarity between the candidate base spectra and the standard spectral library of typical polyphenol subclasses in olive leaves, the candidate base spectra are identified and corrected to pure base spectra corresponding to hydroxytyrosol glycosides, oleuropein and flavonol glycosides, and the identified polyphenol subclass base spectra are obtained.

[0072] Step S330: Based on the identified polyphenol subclass base spectra and the purified polyphenol characteristic spectral sequences, perform concentration profile analysis and matrix construction. Using the identified base spectra as a benchmark, calculate the contribution of each base spectrum in the spectrum at each time point using alternating least squares optimization inversion to generate a concentration profile matrix characterizing the evolution of the concentration of each polyphenol subcomponent over time.

[0073] Specifically, in this embodiment, step S310 first performs preliminary base spectrum extraction of polyphenol subclasses. By calculating the correlation between the differences in spectra at different time points in the purified polyphenol characteristic spectral sequences, spectral components exhibiting independent changing trends in the time domain are identified, thereby separating candidate polyphenol subclass base spectra. This process utilizes the statistical characteristics of spectral variability to preliminarily distinguish the signals of different polyphenol subcomponents, laying the foundation for subsequent identification. Next, step S320 performs base spectrum identification and purification based on a standard spectral library. By calculating the similarity between the candidate base spectra and the standard spectral library of typical polyphenol subclasses in olive leaves, the candidate base spectra are further refined. The basic spectra are selected, corrected, and identified as pure basic spectra corresponding to specific polyphenol subclasses such as hydroxytyrosol glycosides, oleuropein, and flavonol glycosides, resulting in the basic spectra of the identified polyphenol subclasses. This identification process utilizes known standards to improve accuracy and ensure the correspondence between the basic spectra and the actual components. Finally, step S330 performs concentration profile analysis and matrix construction. Using the identified basic spectra as a benchmark, the contribution of each basic spectrum at each time point is calculated using an alternating least squares optimization mathematical inversion algorithm, generating a concentration profile matrix characterizing the concentration evolution of each polyphenol subcomponent over time, thus fully presenting the dynamic trajectory of polyphenol transformation. The entire process, from initial separation to precise identification and quantitative analysis, progressively deepens data mining and achieves efficient tracking of specific subcomponents in complex mixtures through spectral deconvolution technology.

[0074] In the concentration profile analysis and matrix construction process, the core procedure involves using an iterative optimization framework to decompose the observed complex spectral data into two factors with distinct physical meanings. The basic idea is to assume that the measured spectral data matrix can be approximated by the product of a set of known fundamental spectrum matrices and a concentration matrix to be determined. By defining an error function to measure the goodness of the approximation, and using this as the objective, while fixing one factor, the other factor is optimally solved. This process is repeated iteratively until the solution update reaches a stable state. The resulting concentration matrix then characterizes the dynamic trajectory of each component over time. The inversion process is as follows:

[0075] First, define the objective function to be minimized:

[0076] ;

[0077] Where D is the observed spectral data matrix, W is the fundamental spectrum matrix, and H is the concentration profile matrix. is the Frobenius norm of the matrix, used to measure the reconstruction error.

[0078] Secondly, the iterative process follows the principle of alternating optimization, and its update rule is as follows:

[0079] With the concentration matrix H fixed, update the fundamental spectrum matrix W:

[0080] ;

[0081] This formula obtains the current optimal basis matrix W by solving the least squares problem under the condition of fixed H.

[0082] With the fundamental spectrum matrix W fixed, the concentration matrix H is updated:

[0083] ;

[0084] Similarly, this formula solves the least squares problem under the condition of fixed W, updating the concentration matrix H. The iteration will continue until the changes in the updates of matrices W and H in two adjacent iterations are less than the preset convergence threshold. At this point, the algorithm reaches stability, and the obtained concentration matrix H is the dynamic profile representing the evolution of the concentration of each polyphenol subclass over time.

[0085] Figures 4-6 This is a visualization result of the polyphenol conversion dynamics monitored in an embodiment of the method. Figure 4 and Figure 5 The system visually presents the concentration changes of the five main polyphenol components during the monitoring period. The color intensity directly corresponds to the concentration level, clearly revealing the consumption of precursor substances (hydroxytyrosol glycosides, oleuropein) and the accumulation of product substances (hydroxytyrosol, oleuropein ligands, etc.). Figure 6 The specific trajectories of the concentration evolution of each component over time were further characterized. These results demonstrate that the proposed method effectively separates overlapping spectral signals through spectral deconvolution, and the obtained concentration matrix accurately reflects the chemical kinetics of polyphenol transformation, providing a reliable data foundation for subsequent trajectory fitting and stage identification.

[0086] Example 4:

[0087] In a preferred embodiment, step S400 includes steps S410 to S430.

[0088] Step S410: Based on the polyphenol subcomponent concentration evolution matrix, a concentration ratio matrix is ​​constructed. By calculating the concentration ratios of hydroxytyrosol glycosides and oleuropein, as well as different flavonol glycosides, at each time point, a dynamic concentration ratio matrix characterizing potential metabolic transformation relationships is generated.

[0089] Step S420: Based on the polyphenol subcomponent concentration evolution matrix and dynamic concentration ratio matrix, perform constrained covariance structure reconstruction processing. By using the stable proportional relationship reflected by the dynamic concentration ratio matrix as a regularization term and incorporating it into the covariance calculation process, a target covariance matrix that can amplify metabolic correlation signals and suppress random fluctuations is constructed.

[0090] Step S430: Principal component extraction is performed based on the target covariance matrix. By performing eigenvalue decomposition on the matrix, the principal components whose eigenvector directions best match the decreasing trend of hydroxytyrosol glycoside concentration and increasing trend of oleuropein concentration are selected, thus obtaining the reconstruction results that reflect the dynamics of the core metabolic transformation pathway of polyphenols.

[0091] In the preferred embodiment of Example 4, step S400 uses a multi-stage data reconstruction strategy to deeply analyze the intrinsic dynamic mechanism of polyphenol metabolic transformation: step S410 uses concentration ratio calculation to convert absolute concentration data into a relative relationship matrix, dynamically capturing the ratio changes between key components such as hydroxytyrosol glycosides and oleuropein, thereby indirectly reflecting the activity intensity of metabolic pathways such as glycoside hydrolysis or oxidative condensation. This method weakens the noise interference caused by environmental fluctuations through ratio processing, highlighting the stability of the transformation trend; step S420 introduces regularized covariance reconstruction, incorporating the stable proportional relationship contained in the concentration ratio matrix as a mathematical regularization term into the covariance calculation. The regularization technique strengthens the synergistic change pattern between polyphenol subcomponents by penalizing random fluctuations, constructing a target covariance matrix that can amplify metabolic correlation signals. This processing effectively distinguishes between the real transformation process and random variation, improving the robustness of signal extraction in complex plant matrices.

[0092] In regularized covariance reconstruction, the target covariance matrix is ​​usually obtained by solving an optimization problem, and its form can be expressed as:

[0093] ;

[0094] in, Let be the target covariance matrix to be solved, and Σ be the optimization variable, representing the covariance matrix to be determined during the optimization process. Let S be the trace of the matrix, which is the sum of the elements on the main diagonal of the matrix, and let S be the sample covariance matrix, which is directly calculated from the observed polyphenol subcomponent concentration data. Let be the regularization parameter; R(Σ, Θ) is the regularization term, a function of Σ and Θ, used to penalize the difference between Σ and the prior matrix Θ, thus introducing domain knowledge (i.e., the stable proportional relationship between polyphenol subcomponents) as a constraint into the optimization process; Θ is the prior matrix. A common specific form of the regularization term R(Σ, Θ) is the squared Frobenius norm:

[0095] ;

[0096] This form forces the estimated covariance matrix Σ to approximate the prior matrix Θ, thereby amplifying metabolic correlation signals and suppressing random fluctuations.

[0097] Therefore, the target covariance matrix It is the solution to the above optimization problem. Its construction enhances the synergistic changes (i.e. metabolic associations) between polyphenol subcomponents, while regularization distinguishes the real transformation process from random variation.

[0098] Step S430 extracts principal components through feature decomposition, selecting the feature vectors that best match the trends of hydroxytyrosol consumption and oleuropein accumulation to quantify the dynamic evolution of the core metabolic pathway. Principal component analysis reduces the dimensionality of high-dimensional data to the key trend directions, thus intuitively presenting the dominant mechanism of polyphenol conversion. The entire process, from ratio quantification to covariance optimization and trend extraction, gradually deepens the analysis of the metabolic network structure. Through the combination of mathematical modeling and statistical constraints, it provides kinetic-level theoretical support for process optimization.

[0099] Example 5:

[0100] In a preferred embodiment, step S500 includes steps S510 to S530.

[0101] Step S510: Based on the principal component score trajectory in the reconstruction results, perform correlation analysis between trajectory segments and transformation steps. By identifying the inflection points and different rate stages of change of the score trajectory, correlate them with the rapid consumption stage of hydroxytyrosol glycoside hydrolysis and the slow accumulation stage of subsequent oxidative condensation in time sequence to obtain the segmented kinetic trajectory.

[0102] Step S520: Based on the kinetic trajectory, construct the kinetic network structure of the series reaction. By constructing a two-stage series reaction network topology structure based on the segmented temporal relationship, with hydroxytyrosol glycoside as the starting reactant, hydrolyzing to generate hydroxytyrosol, and then oxidatively condensing to generate dimer or polymer products, the kinetic network model to be solved is obtained.

[0103] Step S530: Based on the kinetic network model and the segmented kinetic trajectories, the model parameters are solved and optimized. By substituting the data of different trajectory segments into the rate equations of the corresponding reaction steps for iterative fitting, the reaction order, rate constant, and activation energy parameter set of the hydrolysis reaction and oxidative condensation reaction are obtained as the fitting result.

[0104] Specifically, step S510 utilizes the morphological characteristics of the principal component score trajectory to identify trajectory inflection points and differences in change rates, dividing the kinetic trajectory into temporal segments corresponding to the rapid consumption stage of hydroxytyrosol glycoside hydrolysis and the slow accumulation stage of oxidative condensation. This segmentation identification method based on trajectory differential characteristics can effectively capture the nonlinear changes in reaction rate during the conversion process, highlighting the dynamic differences in different reaction steps. Subsequently, step S520 constructs a two-level tandem reaction network topology based on the segmented temporal logic. Using hydroxytyrosol glycoside as the starting reactant, it generates hydroxytyrosol intermediates through hydrolysis, and then forms dimer or polymeric final products through oxidative condensation. This network structure accurately reflects the chemical pathway essence of polyphenol conversion through the sequential arrangement of reaction steps. Finally, step S530 uses an iterative fitting algorithm to solve for parameters. The purpose of the iterative fitting process is to determine the unknown parameters in the tandem reaction kinetic model through optimization algorithms, so that the concentration change trend predicted by the model achieves the best match with the experimentally observed principal component score trajectory or the concentration trajectory obtained by its inversion. This process typically employs the core idea of ​​nonlinear least squares, which involves constructing an objective function to quantify the difference between model predictions and experimental observations, and then using numerical optimization algorithms to iteratively adjust model parameters to minimize the objective function value. The process is as follows:

[0105] First, based on the established two-stage cascade reaction network, a set of ordinary differential equations describing the time evolution of the concentrations of each component is constructed. This cascade reaction kinetic model contains two consecutive reaction steps, and the reaction rate equation is expressed as:

[0106] ;

[0107] Where r1 is the instantaneous reaction rate of the hydrolysis of hydroxytyrosol glycoside (A0); k1 is the rate constant of the hydrolysis reaction, the value of which reflects the speed of the reaction; [A0] is the instantaneous concentration of B0 in the reaction system; n1 is the reaction order of the hydrolysis reaction; r2 is the instantaneous reaction rate of the oxidative condensation reaction of hydroxytyrosol (B0); k2 is the rate constant of the oxidative condensation reaction; [B0] is the instantaneous concentration of B0 in the reaction system; and n2 is the reaction order of the oxidative condensation reaction.

[0108] The system of ordinary differential equations is as follows:

[0109] ;

[0110] in, The value represents the rate of change of the concentration of hydroxytyrosol glycoside (A0) over time; the negative sign indicates that it was consumed due to the reaction. The concentration of hydroxytyrosol (B0) changes over time. Its formation originates from the hydrolysis of A0, while its consumption originates from its own oxidative condensation. t represents the rate of change in the concentration of the dimer / polymer (CO) over time, generated by the oxidative condensation of B0; t is the reaction time.

[0111] Subsequently, the segmented kinetic trajectory data obtained from the experiment, i.e., the concentrations of each component or their associated principal component scores at different time points, were compared with the predicted concentrations obtained from the model at the corresponding time points. The objective function was constructed by calculating the sum of squared differences between the two. The objective function is:

[0112] ;

[0113] in, This means minimizing the subsequent summation expression by adjusting the values ​​of parameters k1, n1, k2, n2; N is the total number of experimental observation time points; t i This refers to the i-th observation time point; , , This indicates at time point t i The concentration values ​​of components A0, B0, and C0 obtained from experimental observation (or from principal component score inversion); , , This indicates at time point t i The model predicts the concentration values ​​of components A0, B0, and C0 by numerically solving the above set of ordinary differential equations.

[0114] Next, reasonable initial estimates are set for the kinetic parameters to be solved. Based on this, a numerical solution and iterative optimization loop is entered: in each iteration, the ordinary differential equations are first solved numerically to obtain the concentration time series predicted by the model under the current parameters; then, the objective function value corresponding to the current parameters is calculated; the optimization algorithm automatically updates the parameter values ​​based on the gradient information of the objective function with respect to the parameters, aiming to reduce the objective function value in the next iteration. This process is repeated until the change in parameters or the decrease in the objective function value is less than a pre-set convergence threshold, at which point the optimal kinetic parameter solution is obtained.

[0115] Example 6:

[0116] In a preferred embodiment, step S600 includes steps S610 to S630.

[0117] Step S610: Map the target condition parameters according to the fitting results. By mapping the temperature and humidity factors in the target storage or processing conditions to the corresponding reaction rate constants, and setting the initial concentration of the model according to the initial polyphenol subcomponent concentration evolution matrix, the kinetic model initialized under the target conditions is obtained.

[0118] Step S620: Perform numerical integration and evolution simulation of the polyphenol conversion process according to the kinetic model. By performing numerical integration on the coupled kinetic differential equations, simulate and calculate the continuous change curve of the concentration of each polyphenol sub-component from the initial time to the preset end time, and obtain the simulated evolution curve.

[0119] Step S630: Based on the simulated evolution curve, calculate the evolution endpoint of the target component and integrate the results. Extract the absolute concentration of hydroxytyrosol in the free state at the preset endpoint from the simulated evolution curve, calculate its relative proportion with the total concentration of all condensed derivatives, and integrate to generate the final dynamic monitoring results.

[0120] Specifically, in this embodiment, step S610 performs target condition parameter mapping, using kinetic principles to convert temperature and humidity parameters in the actual storage or processing environment into reaction rate constants. For example, temperature and reaction rate are correlated through the Arrhenius equation. Simultaneously, the initial concentration of the model is set based on the initial polyphenol subcomponent concentration evolution matrix, ensuring the initialization of the kinetic model under real boundary conditions. This process internalizes external environmental variables into model parameters, enhancing the practicality and adaptability of the prediction. Next, step S620 performs numerical integration and evolutionary simulation processing, using numerical algorithms such as the Runge-Kutta method to solve the coupled kinetic differential equations, simulating the continuous change curves of the concentrations of each polyphenol subcomponent from the initial time to the preset endpoint time. High-precision numerical integration captures the nonlinear dynamics in the conversion process, generating a time-resolved simulated evolution trajectory. Finally, step S630 calculates the evolution endpoint and integrates the results, extracting the absolute concentration of hydroxytyrosol in its free state at the endpoint from the simulation curve and calculating its relative proportion to the total concentration of the condensed derivative, quantifying the conversion efficiency and product distribution, and generating comprehensive dynamic monitoring results. The entire process, through progressive processing of parameterization, simulation, and integration, achieves predictable control of polyphenol conversion behavior, providing theoretical support for quality management of olive leaves during storage or processing.

[0121] Example 7:

[0122] like Figure 2 As shown, this embodiment provides a dynamic monitoring system for olive leaf polyphenol components based on principal component analysis. The system includes:

[0123] The acquisition module 901 is used to acquire raw data, which is the raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, containing the vibration and electronic transition responses of polyphenol characteristic functional groups.

[0124] Separation module 902 is used to perform spectral region separation based on raw data to obtain the characteristic spectral sequence of purified polyphenols;

[0125] The processing module 903 is used to perform spectral deconvolution processing based on the characteristic spectral sequence of purified polyphenols. By decomposing the time-series spectrum into the base spectrum of polyphenol sub-components and their respective concentration profiles over time, the concentration evolution matrix of polyphenol sub-components is obtained.

[0126] Reconstruction module 904 is used to perform principal component reconstruction based on the polyphenol subcomponent concentration evolution matrix. By using the concentration ratio relationship between polyphenol subcomponents evolving over time as a dynamic constraint on the covariance structure of the reconstructed data, dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction result.

[0127] Fitting module 905 is used to fit the polyphenol conversion kinetic trajectory based on the reconstruction results. The fitting result is obtained by fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides.

[0128] The prediction module 906 is used to make predictions based on the fitting results, extrapolate the dynamic model to the target storage or processing conditions, predict the evolution endpoint of the free content of hydroxytyrosol and the relative proportion of its derivatives, and obtain dynamic monitoring results.

[0129] In one specific embodiment of this application, the separation module 902 includes:

[0130] The first separation unit is used to perform non-phenolic background absorption contribution separation processing based on the original data. By identifying and subtracting the broadband background absorption bands attributed to the hydrogen bond stretching vibration of leaf water and the carbon-oxygen bond vibration of cellulose skeleton from the full-band spectrum, the intermediate spectral sequence is obtained.

[0131] The second separation unit is used to extract the spectral range of polyphenol characteristic functional groups based on the intermediate spectral sequence. By calculating the first derivative spectrum and locking the characteristic response range of carbon-carbon double bond skeleton vibration of benzene ring, oxygen-hydrogen bond stretching vibration of phenolic hydroxyl group and carbon-oxygen-carbon asymmetric stretching vibration of glycosidic bond, the candidate polyphenol characteristic spectral sequence is obtained.

[0132] The third separation unit is used to optimize the characteristic regions and integrate the sequences based on the characteristic spectral sequences of candidate polyphenols. By introducing principal component loading vectors as weights, the signal-to-noise ratio and sensitivity to concentration changes of different characteristic regions are evaluated and weighted to obtain the characteristic spectral sequences of purified polyphenols.

[0133] In one specific embodiment of this application, the processing module 903 includes:

[0134] The first processing unit is used to perform preliminary base spectrum extraction of polyphenol subclasses based on the characteristic spectral sequences of purified polyphenols. By calculating the correlation between the differences in spectra at each time point, spectral components with independent changing trends in the time domain are separated to obtain the base spectra of candidate polyphenol subclasses.

[0135] The second processing unit is used to perform basic spectrum identification and purification based on the basic spectrum of candidate polyphenol subclasses using a standard spectral library. By calculating the similarity between the candidate basic spectrum and the standard spectral library of typical polyphenol subclasses in olive leaves, the candidate basic spectrum is identified and corrected to pure basic spectra corresponding to hydroxytyrosol glycosides, oleuropein and flavonol glycosides, thus obtaining the basic spectrum of the identified polyphenol subclasses.

[0136] The third processing unit is used to perform concentration profile analysis and matrix construction based on the identified polyphenol subclass base spectra and the characteristic spectral sequences of purified polyphenols. By using the identified base spectra as a reference, the contribution of each base spectrum in the spectrum at each time point is calculated using alternating least squares optimization inversion, and a concentration profile matrix characterizing the concentration evolution of each polyphenol subcomponent over time is generated.

[0137] In one specific embodiment of this application, the reconstruction module 904 includes:

[0138] The first reconstruction unit is used to construct a concentration ratio matrix based on the polyphenol subcomponent concentration evolution matrix. By calculating the concentration ratios of hydroxytyrosol glycosides and oleuropein, as well as different flavonol glycosides, at each time point, a dynamic concentration ratio matrix characterizing potential metabolic transformation relationships is generated.

[0139] The second reconstruction unit is used to perform constrained covariance structure reconstruction based on the polyphenol subcomponent concentration evolution matrix and the dynamic concentration ratio matrix. By using the stable proportional relationship reflected by the dynamic concentration ratio matrix as a regularization term and incorporating it into the covariance calculation process, a target covariance matrix that can amplify metabolic correlation signals and suppress random fluctuations is constructed.

[0140] The third reconstruction unit is used to perform principal component extraction based on the target covariance matrix. By performing eigenvalue decomposition on the matrix, the principal component whose eigenvector direction best matches the decreasing trend of hydroxytyrosol glycoside concentration and increasing trend of oleuropein concentration is selected, thus obtaining the reconstruction result that reflects the dynamics of the core metabolic transformation pathway of polyphenols.

[0141] In one specific embodiment of this application, the fitting module 905 includes:

[0142] The first fitting unit is used to perform correlation analysis between trajectory segments and transformation steps based on the principal component score trajectories in the reconstruction results. By identifying the inflection points and different rate stages of change of the score trajectory, it is temporally correlated with the rapid consumption stage of hydroxytyrosol glycoside hydrolysis and the slow accumulation stage of subsequent oxidative condensation to obtain the segmented kinetic trajectory.

[0143] The second fitting unit is used to construct the cascade reaction kinetic network structure based on the kinetic trajectory. By constructing a two-stage cascade reaction network topology structure based on the segmented temporal relationship, starting with hydroxytyrosol glycoside as the initial reactant, hydrolyzing to generate hydroxytyrosol, and then oxidatively condensing to generate dimer or polymer products, the kinetic network model to be solved is obtained.

[0144] The third fitting unit is used to solve and optimize the model parameters based on the kinetic network model and the segmented kinetic trajectories. By substituting the data of different trajectory segments into the rate equations of the corresponding reaction steps for iterative fitting, the set of reaction orders, rate constants and activation energy parameters of the hydrolysis reaction and the oxidative condensation reaction are obtained as the fitting results.

[0145] In one specific embodiment of this application, the prediction module 906 includes:

[0146] The first prediction unit is used to map the target condition parameters based on the fitting results. It maps the temperature and humidity factors in the target storage or processing conditions to the corresponding reaction rate constants and sets the initial concentration of the model based on the initial polyphenol subcomponent concentration evolution matrix to obtain the kinetic model initialized under the target conditions.

[0147] The second prediction unit is used to perform numerical integration and evolution simulation of the polyphenol conversion process based on the kinetic model. By performing numerical integration on the coupled kinetic differential equations, it simulates and calculates the continuous change curve of the concentration of each polyphenol subcomponent from the initial time to the preset end time, and obtains the simulated evolution curve.

[0148] The third prediction unit is used to calculate the evolution endpoint of the target component and integrate the results based on the simulated evolution curve. It extracts the absolute concentration of hydroxytyrosol in the free state at the preset endpoint from the simulated evolution curve and calculates its relative proportion with the total concentration of all condensed derivatives, and integrates the results to generate the final dynamic monitoring results.

[0149] Example 8:

[0150] Corresponding to the above method embodiments, this embodiment also provides a dynamic monitoring device for olive leaf polyphenol components based on principal component analysis. The dynamic monitoring device for olive leaf polyphenol components based on principal component analysis described below and the dynamic monitoring method for olive leaf polyphenol components based on principal component analysis described above can be referred to in correspondence.

[0151] Figure 3 This is a block diagram illustrating a dynamic monitoring device 800 for olive leaf polyphenol components based on principal component analysis, according to an exemplary embodiment. Figure 3As shown, the olive leaf polyphenol component dynamic monitoring device 800 based on principal component analysis may include: a processor 801 and a memory 802. The olive leaf polyphenol component dynamic monitoring device 800 may also include one or more of the following: a multimedia component 803, an I / O interface 804, and a communication component 805.

[0152] The processor 801 controls the overall operation of the olive leaf polyphenol component dynamic monitoring device 800 based on principal component analysis to complete all or part of the steps in the aforementioned method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis. The memory 802 stores various types of data to support the operation of the olive leaf polyphenol component dynamic monitoring device 800. This data may include, for example, instructions for any application or method operating on the device, as well as application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using 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 Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the olive leaf polyphenol component dynamic monitoring device 800 based on principal component analysis and other devices. Wireless communication includes, for example, Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, and an NFC module.

[0153] In an exemplary embodiment, an olive leaf polyphenol component dynamic monitoring device 800 based on principal component analysis may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the aforementioned method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis.

[0154] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When executed by a processor, these program instructions implement the steps of the above-described method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis. For example, the computer-readable storage medium may be the memory 802 including the program instructions, which may be executed by a processor 801 of a device 800 for dynamic monitoring of olive leaf polyphenol components based on principal component analysis to complete the above-described method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis.

[0155] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for dynamic monitoring of polyphenolic components in olive leaves based on principal component analysis, characterized in that, include: Obtain raw data, which is raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, including vibrational and electronic transition responses of polyphenol characteristic functional groups. Based on the original data, spectral intervals were separated to obtain the characteristic spectral sequences of purified polyphenols. Based on the purified polyphenol characteristic spectral sequence, spectral deconvolution processing is performed. By decomposing the time-series spectrum into the base spectrum of the polyphenol subcomponents and their respective concentration profiles over time, the concentration evolution matrix of the polyphenol subcomponents is obtained. Principal component reconstruction is performed based on the polyphenol subcomponent concentration evolution matrix. The concentration ratio relationship between polyphenol subcomponents over time is used as a dynamic constraint on the covariance structure of the reconstructed data. Dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction results. Based on the reconstruction results, the polyphenol conversion kinetic trajectory was fitted. The fitting result was obtained by fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides. Based on the fitting results, predictions are made, and the evolution endpoint of the free hydroxytyrosol content and the relative proportion of its derivatives is predicted by extrapolating the kinetic model to the target storage or processing conditions, thus obtaining dynamic monitoring results. Specifically, principal component reconstruction is performed based on the polyphenol subcomponent concentration evolution matrix. The concentration ratios of polyphenol subcomponents evolving over time are used as dynamic constraints on the covariance structure of the reconstructed data. Dynamic principal components characterizing the correlation between precursor consumption and product accumulation are extracted to obtain the reconstruction results, including: Based on the polyphenol subcomponent concentration evolution matrix, a concentration ratio matrix is ​​constructed. By calculating the concentration ratios of hydroxytyrosol glycosides and oleuropein, as well as different flavonol glycosides, at each time point, a dynamic concentration ratio matrix characterizing potential metabolic transformation relationships is generated. Based on the polyphenol subcomponent concentration evolution matrix and the dynamic concentration ratio matrix, a constrained covariance structure reconstruction process is performed. By using the stable proportional relationship reflected by the dynamic concentration ratio matrix as a regularization term and incorporating it into the covariance calculation process, a target covariance matrix that can amplify metabolic correlation signals and suppress random fluctuations is constructed. Principal component extraction is performed based on the target covariance matrix. By performing eigenvalue decomposition on the matrix, the principal component whose eigenvector direction best matches the decreasing trend of hydroxytyrosol glycoside concentration and increasing trend of oleuropein concentration is selected, and the reconstruction result reflecting the dynamics of the core metabolic transformation pathway of polyphenols is obtained. The process involves fitting a polyphenol transformation kinetic trajectory based on the reconstruction results. This is achieved by fitting the principal component score trajectory to a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides, resulting in the following fitting results: Based on the principal component score trajectory in the reconstruction results, the correlation analysis between the trajectory segment and the transformation step is performed. By identifying the inflection point and different rate of change stages of the score trajectory, it is temporally correlated with the rapid consumption stage of hydroxytyrosol glycoside hydrolysis and the slow accumulation stage of subsequent oxidative condensation to obtain the segmented kinetic trajectory. Based on the aforementioned kinetic trajectory, a series reaction kinetic network structure is constructed. By following the segmented temporal relationship, a two-stage series reaction network topology is constructed, which starts with hydroxytyrosol glycosides as reactants, hydrolyzes to generate hydroxytyrosol, and then undergoes oxidative condensation to generate dimer or polymer products. This yields the kinetic network model to be solved. Based on the kinetic network model and the segmented kinetic trajectories, the model parameters are solved and optimized. By substituting the data of different trajectory segments into the rate equations of the corresponding reaction steps for iterative fitting, the reaction order, rate constant, and activation energy parameter set of the hydrolysis reaction and oxidative condensation reaction are obtained as the fitting result.

2. The method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis according to claim 1, characterized in that, Based on the original data, spectral region separation was performed to obtain the characteristic spectral sequences of purified polyphenols, including: Based on the original data, non-phenolic background absorption contribution separation processing was performed. By identifying and subtracting the broadband background absorption bands attributable to the hydrogen bond stretching vibration of leaf water and the carbon-oxygen bond vibration of cellulose skeleton from the full-band spectrum, the intermediate spectral sequence was obtained. Based on the intermediate spectral sequence, the spectral range of polyphenol characteristic functional groups is extracted. By calculating the first derivative spectrum and locking the characteristic response range of carbon-carbon double bond skeletal vibration of benzene ring, oxygen-hydrogen bond stretching vibration of phenolic hydroxyl group and carbon-oxygen-carbon asymmetric stretching vibration of glycosidic bond, the candidate polyphenol characteristic spectral sequence is obtained. Based on the candidate polyphenol characteristic spectral sequences, feature interval optimization and sequence integration processing are performed. By introducing principal component loading vectors as weights, the signal-to-noise ratio and sensitivity to concentration changes of different feature intervals are evaluated and weighted fusion to obtain the purified polyphenol characteristic spectral sequences.

3. The method for dynamic monitoring of olive leaf polyphenol components based on principal component analysis according to claim 1, characterized in that, Based on the purified polyphenol characteristic spectral sequence, spectral deconvolution processing is performed. By decomposing the time-series spectrum into the base spectra of polyphenol subcomponents and their respective concentration profiles over time, the polyphenol subcomponent concentration evolution matrix is ​​obtained, including: Based on the purified polyphenol characteristic spectral sequence, preliminary base spectrum extraction of polyphenol subclasses was performed. By calculating the correlation between the differences in spectra at each time point, spectral components with independent changing trends in the time domain were separated to obtain candidate polyphenol subclass base spectra. Based on the candidate polyphenol subclass base spectra, base spectrum identification and purification were performed based on a standard spectral library. By calculating the similarity between the candidate base spectra and the standard spectral library of typical polyphenol subclasses in olive leaves, the candidate base spectra were identified and corrected to pure base spectra corresponding to hydroxytyrosol glycosides, oleuropein and flavonol glycosides, thus obtaining the identified polyphenol subclass base spectra. Based on the identified polyphenol subclass base spectra and the purified polyphenol characteristic spectral sequences, concentration profile analysis and matrix construction were performed. By using the identified base spectra as a benchmark, the contribution of each base spectrum in the spectrum at each time point was calculated using alternating least squares optimization inversion, thereby generating a concentration profile matrix characterizing the evolution of the concentration of each polyphenol subcomponent over time.

4. A dynamic monitoring system for olive leaf polyphenol components based on principal component analysis, characterized in that, include: The acquisition module is used to acquire raw data, which is the raw near-infrared and mid-infrared absorption spectrum data of olive leaves at continuous monitoring time points, containing the vibration and electronic transition responses of polyphenol characteristic functional groups. A separation module is used to perform spectral region separation based on the original data to obtain the characteristic spectral sequence of purified polyphenols. The processing module is used to perform spectral deconvolution processing based on the purified polyphenol characteristic spectral sequence, and obtain the polyphenol sub-component concentration evolution matrix by decomposing the time-series spectrum into the base spectrum of the polyphenol subcomponent and its respective concentration profile over time. The reconstruction module is used to perform principal component reconstruction based on the polyphenol subcomponent concentration evolution matrix. By using the concentration ratio relationship between polyphenol subcomponents evolving over time as a dynamic constraint condition for the covariance structure of the reconstructed data, dynamic principal components characterizing the correlation trend between precursor consumption and product accumulation are extracted to obtain the reconstruction result. The fitting module is used to fit the polyphenol conversion kinetic trajectory based on the reconstruction result. The fitting result is obtained by fitting the principal component score trajectory into a kinetic network model that includes the hydrolysis and oxidative condensation steps of hydroxytyrosol glycosides. The prediction module is used to make predictions based on the fitting results, extrapolate the dynamic model to the target storage or processing conditions, predict the evolution endpoint of the free content of hydroxytyrosol and the relative proportion of its derivatives, and obtain dynamic monitoring results. The reconstruction module includes: The first reconstruction unit is used to construct a concentration ratio matrix based on the concentration evolution matrix of the polyphenol subcomponents. By calculating the concentration ratios of hydroxytyrosol glycosides and oleuropein, as well as different flavonol glycosides, at each time point, a dynamic concentration ratio matrix characterizing potential metabolic transformation relationships is generated. The second reconstruction unit is used to perform constrained covariance structure reconstruction processing based on the polyphenol subcomponent concentration evolution matrix and the dynamic concentration ratio matrix. By using the stable proportional relationship reflected by the dynamic concentration ratio matrix as a regularization term and incorporating it into the covariance calculation process, a target covariance matrix that can amplify metabolic correlation signals and suppress random fluctuations is constructed. The third reconstruction unit is used to perform principal component extraction processing based on the target covariance matrix. By performing eigenvalue decomposition on the matrix, the principal component whose eigenvector direction best matches the decreasing trend of hydroxytyrosol glycoside concentration and increasing trend of oleuropein concentration is selected, and the reconstruction result reflecting the dynamics of the core metabolic transformation pathway of polyphenols is obtained. The fitting module includes: The first fitting unit is used to perform correlation analysis processing between trajectory segments and transformation steps based on the principal component score trajectory in the reconstruction result. By identifying the inflection point and different rate of change stages of the score trajectory, it is temporally correlated with the rapid consumption stage of hydroxytyrosol glycoside hydrolysis and the slow accumulation stage of subsequent oxidative condensation to obtain the segmented kinetic trajectory. The second fitting unit is used to construct a series reaction kinetic network structure based on the kinetic trajectory. By constructing a two-stage series reaction network topology structure based on the segmented temporal relationship, with hydroxytyrosol glycoside as the starting reactant, hydrolysis to generate hydroxytyrosol, and then oxidative condensation to generate dimer or polymer products, a kinetic network model to be solved is obtained. The third fitting unit is used to solve and optimize the model parameters based on the kinetic network model and the segmented kinetic trajectories. By substituting the data of different trajectory segments into the rate equations of the corresponding reaction steps for iterative fitting, the set of reaction orders, rate constants and activation energy parameters of the hydrolysis reaction and the oxidative condensation reaction are obtained as the fitting result.

5. The dynamic monitoring system for olive leaf polyphenol components based on principal component analysis according to claim 4, characterized in that, The separation module includes: The first separation unit is used to perform non-phenolic background absorption contribution separation processing based on the original data. By identifying and subtracting the broadband background absorption bands attributed to the hydrogen bond stretching vibration of leaf water and the carbon-oxygen bond vibration of cellulose skeleton from the full-band spectrum, an intermediate spectral sequence is obtained. The second separation unit is used to extract the spectral range of polyphenol characteristic functional groups based on the intermediate spectral sequence. By calculating the first derivative spectrum and locking the characteristic response range of carbon-carbon double bond skeleton vibration of benzene ring, oxygen-hydrogen bond stretching vibration of phenolic hydroxyl group and carbon-oxygen-carbon asymmetric stretching vibration of glycosidic bond, the candidate polyphenol characteristic spectral sequence is obtained. The third separation unit is used to perform feature interval optimization and sequence integration processing based on the candidate polyphenol feature spectral sequences. By introducing principal component loading vectors as weights, the signal-to-noise ratio and sensitivity to concentration changes of different feature intervals are evaluated and weighted to obtain the purified polyphenol feature spectral sequences.

6. The dynamic monitoring system for olive leaf polyphenol components based on principal component analysis according to claim 4, characterized in that, The processing module includes: The first processing unit is used to perform preliminary base spectrum extraction processing of polyphenol subclasses based on the purified polyphenol characteristic spectral sequence, and to separate spectral components with independent changing trends in the time domain by calculating the difference correlation between spectra at each time point, thereby obtaining candidate polyphenol subclass base spectra. The second processing unit is used to perform basic spectrum identification and purification based on the candidate polyphenol subclass basic spectrum. By calculating the similarity between the candidate basic spectrum and the standard spectral library of typical polyphenol subclasses in olive leaves, the candidate basic spectrum is identified and corrected to pure basic spectrum corresponding to hydroxytyrosol glycoside, oleuropein and flavonol glycoside, and the identified polyphenol subclass basic spectrum is obtained. The third processing unit is used to perform concentration profile analysis and matrix construction based on the identified polyphenol subclass base spectra and the purified polyphenol characteristic spectral sequences. By using the identified base spectra as a reference, the contribution of each base spectrum in the spectrum at each time point is calculated using alternating least squares optimization inversion, and a concentration profile matrix characterizing the concentration evolution of each polyphenol subcomponent over time is generated.

Images