A method and system for predicting seasonal variation trend of olea europaea leaf components

Abstract

The application provides a kind of olive leaf component seasonal change trend prediction method and system, related to computer technology field, comprising: obtaining the original spectral absorbance data of target variety olive leaf, and original meteorological data;According to the original spectral absorbance data, component-specific signal separation is carried out, to obtain pure change trajectory;According to the original meteorological data and pure change trajectory, environmental stress modal decomposition is carried out, to obtain differential driving function;According to the differential driving function, metabolic resource allocation network evolution is carried out, to obtain structural emergent features;According to the structural emergent features and pure change trajectory, phase space reconstruction is carried out, to obtain phase space geometric descriptor;According to the phase space geometric descriptor, the best harvesting window decision is made, to predict the best sampling time interval in the current growth cycle.The application provides reliable technical support for the efficient use of olive leaf resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular, to an olive leaf component seasonal change trend prediction method and system. BACKGROUND

[0002] It is well known that olive oil, especially extra virgin olive oil, is recognized as one of the healthiest edible oils, and its unique fatty acid composition and high content of natural antioxidants make olive oil have very high edible value and economic value. However, in the field of research and development of active ingredients in olive leaves, with the vigorous development of the health industry and the increasing demand for high-value utilization of plant resources, the seasonal dynamic monitoring and accurate prediction of active ingredients such as polyphenols, flavonoids and oleuropein in olive leaves have become an important basis for agricultural precision management, drug raw material quality control and functional food development.

[0003] In recent years, researchers have generally used high-performance liquid chromatography, spectral analysis and other technical means to qualitatively and quantitatively analyze plant active ingredients, and combined with meteorological observation data to establish a correlation model between component changes and environmental factors. In the existing technical path, time series analysis, multivariate statistical regression or simple machine learning algorithms are usually used to fit historical monitoring data to predict the change trend of the components. These methods can reflect the overall law of component changes to some extent, but the model construction often relies on the linear or static assumption between meteorological factors and component content, and fails to fully consider the dynamic interaction process between the internal physiological regulation mechanism of plants and environmental stress signals. Especially in dealing with nonlinear and time-varying biological processes, such as critical mutations, path switching and other complex behaviors in the synthesis, transportation and accumulation processes of secondary metabolites, the existing methods are difficult to achieve accurate prediction and meet the urgent needs of accurate judgment of the harvest window in actual production.

[0004] Based on the above-mentioned shortcomings of the prior art, there is an urgent need for an olive leaf component seasonal change trend prediction method and system. SUMMARY

[0005] The purpose of the present application is to provide an olive leaf component seasonal change trend prediction method and system to improve the above-mentioned problems. In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows:

[0006] In a first aspect, the present application provides an olive leaf component seasonal change trend prediction method, comprising:

[0007] obtaining directly measured original spectral absorbance data of target variety olive leaves at non-uniform sampling time points in multiple complete growth years, and continuously monitoring original meteorological data at the same period;

[0008] According to the original spectral absorbance data, component-specific signal separation is performed to obtain a pure change trajectory of each target component;

[0009] According to the original meteorological data and the pure change trajectory, environmental stress modal decomposition is performed to obtain a differential driving function of a meteorological modal on a component metabolic flow;

[0010] According to the differential driving function, metabolic resource allocation network evolution is performed to obtain a structural emergent feature of metabolic allocation under meteorological stress;

[0011] According to the structural emergent feature and the pure change trajectory, a phase space reconstruction of a xylem sap flow-phloem loading coupling system is performed to obtain a phase space geometric descriptor of a source-sink relationship critical state;

[0012] According to the phase space geometric descriptor, an optimal harvesting window decision is made, and by calculating a geodesic line path from a current observation state to a historical optimal harvesting region, an optimal sampling time interval in a current growth cycle is predicted.

[0013] In a second aspect, the application also provides an olive leaf component seasonal change trend prediction method and system, comprising:

[0014] An acquisition module is configured to acquire original spectral absorbance data directly measured at non-uniform sampling time points in multiple complete growth years of a target variety of olive leaves, and original meteorological data continuously monitored at the same period;

[0015] A separation module is configured to perform component-specific signal separation according to the original spectral absorbance data to obtain a pure change trajectory of each target component;

[0016] A decomposition module is configured to perform environmental stress modal decomposition according to the original meteorological data and the pure change trajectory to obtain a differential driving function of a meteorological modal on a component metabolic flow;

[0017] An evolution module is configured to perform metabolic resource allocation network evolution according to the differential driving function to obtain a structural emergent feature of metabolic allocation under meteorological stress;

[0018] A reconstruction module is configured to perform phase space reconstruction of a xylem sap flow-phloem loading coupling system according to the structural emergent feature and the pure change trajectory to obtain a phase space geometric descriptor of a source-sink relationship critical state;

[0019] A decision module is configured to make an optimal harvesting window decision according to the phase space geometric descriptor, and by calculating a geodesic line path from a current observation state to a historical optimal harvesting region, an optimal sampling time interval in a current growth cycle is predicted.

[0020] The beneficial effects of the present application are:

[0021] The present application realizes accurate prediction of seasonal dynamics of active ingredients in olive leaves by constructing a historical database of multi-year spectral and meteorological data, combining component-specific signal separation, environmental stress modal decomposition, metabolic resource allocation network evolution and source-sink relationship phase space reconstruction, and calculating the optimal harvesting window based on geodesic path optimization, effectively improving the accuracy, timeliness of ingredient accumulation trend prediction and scientificity of harvesting decision, providing reliable technical support for efficient utilization of olive leaf resources. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments, and it should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0023] Figure 1 The flowchart of the olive leaf component seasonal change trend prediction method described in the embodiments of the present application;

[0024] Figure 2 The structural diagram of the olive leaf component seasonal change trend prediction system described in the embodiments of the present application;

[0025] Figure 3 The structural diagram of the olive leaf component seasonal change trend prediction device described in the embodiments of the present application.

[0026] Marked in the figure: 800, an olive leaf component seasonal change trend prediction device; 801, a processor; 802, a memory; 803, a multimedia component; 804, an I / O interface; 805, a communication component; 901, an acquisition module; 902, a separation module; 903, a decomposition module; 904, an evolution module; 905, a reconstruction module; 906, a decision module. DETAILED DESCRIPTION

[0027] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0028] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", etc. are only used for differentiation in description, and cannot be understood as indicating or implying relative importance.

[0029] Embodiment 1

[0030] The present embodiment provides a method for predicting seasonal variation trend of olive leaf components.

[0031] Referring to Figure 1 , the method includes steps S100 to S600.

[0032] Step S100, obtaining original spectral absorbance data directly measured at non-uniform sampling time points in multiple complete growth years of target variety olive leaves, and original meteorological data continuously monitored at the same period;

[0033] Step S100 involves the collection of basic data, mainly obtaining two types of data: one is the original spectral absorbance data obtained by non-destructive detection of target variety olive leaves through ultraviolet-visible spectrophotometer or near-infrared spectrometer in multiple complete growth years, which contains spectral response information from ultraviolet to near-infrared band, and can reflect the electronic transition characteristics of phenolic compounds, flavonoids and other compounds in leaves; the other is the original meteorological data recorded synchronously by the automatic weather station set in the planting area, including temperature, relative humidity, sunshine duration, rainfall and other continuous monitoring indexes. The acquisition of these two types of data follows the natural variation law in the actual planting environment, and the non-uniform sampling time points cover the key phenological periods of olive leaf budding, growth and maturity, ensuring the time representativeness and ecological authenticity of the data.

[0034] Step S200, performing component-specific signal separation according to the original spectral absorbance data to obtain the pure change trajectory of each target component;

[0035] Step S200 performs deep mining on the collected spectral data, and decomposes the composite spectral response into independent signals corresponding to specific components through component-specific signal separation technology. This process takes into account the spectral overlap characteristics of multiple phenolic substances in olive leaves, uses signal processing algorithms to extract the pure change trajectory of target components such as total phenols, total flavonoids and oleuropein, and eliminates background interference and co-extraction effects, providing a clear component dynamic benchmark for subsequent analysis.

[0036] Step S300, according to the original meteorological data and the pure change trajectory, performs environmental stress mode decomposition to obtain the differentiated driving function of the meteorological mode to the component metabolic flow;

[0037] Step S300 focuses on analyzing the dynamic correlation between environmental factors and component changes, and converts continuous meteorological data into driving modes with physiological significance through environmental stress mode decomposition. These modes can represent the differential regulation of meteorological conditions on secondary metabolic pathways at different time scales, and then establish a quantitative driving relationship between meteorological modes and component accumulation dynamics through time lag correlation analysis, revealing the internal mechanism of the conversion of environmental stress signals to metabolic flow.

[0038] Step S400, according to the differentiated driving function, performs metabolic resource allocation network evolution to obtain the structural emergence characteristics of metabolic allocation under meteorological stress;

[0039] Step S400 simulates the allocation process of metabolic resources from a system perspective, and converts the aforementioned driving function into a dynamic network model. This network simulates the competitive allocation of limited resources between defensive metabolism and basic metabolism through node activity decay and connection weight change, thereby capturing the structural characteristics of metabolic pathway reconfiguration under meteorological stress, reflecting the physiological strategies of olive leaves adapting to environmental changes.

[0040] Step S500, according to the structural emergence characteristics and the pure change trajectory, performs phase space reconstruction of the xylem sap flow - phloem loading coupling system to obtain the phase space geometric descriptor of the critical state of source-sink relationship;

[0041] Step S500 maps the component dynamics and metabolic allocation characteristics to a high-dimensional geometric space through phase space reconstruction technology, constructing a dynamic model representing the source-sink relationship. This process can capture the geometric characteristics of the system's critical state, and identify the critical point of the qualitative change of the source-sink relationship by analyzing the topological properties of the phase space trajectory, providing a theoretical basis for harvesting decisions.

[0042] Step S600, according to the phase space geometric descriptor, makes the best harvesting window decision, and calculates the geodesic path from the current observation state to the historical optimal harvesting region to predict the best sampling time interval within the current growth cycle.

[0043] Step S600 applies prediction based on the established phase space geometry descriptor, determines the time window when the component accumulation reaches the best state by calculating the geodesic path from the current observation state to the historical optimal harvesting region. This step comprehensively considers the balance between compound accumulation rate and oxidation stability, ensures harvesting at the period before the peak of component concentration and the best stability, and realizes the precision of harvesting decision.

[0044] Further, step S200 includes steps S210 to S230.

[0045] Step S210, according to the original spectral absorbance data, performs spectral data correction processing, and obtains corrected spectral data by eliminating instrument baseline drift and environmental light scattering interference;

[0046] Step S220, according to the corrected spectral data, performs metabolite feature enhancement processing, and obtains a feature-enhanced spectral matrix by performing waveband selection and weighting on the characteristic absorption waveband related to lignification process and terpene lactone synthesis path;

[0047] Step S230, according to the spectral matrix, performs blind source signal separation processing, and obtains the pure change trajectory of each target component by independent component analysis to analyze the independent fluctuation source signal of total phenol, total flavonoids and oleuropein.

[0048] In the preferred embodiment, step S210 first performs spectral data correction processing on the original spectral absorbance data, and obtains corrected spectral data with improved signal-to-noise ratio by eliminating baseline drift (caused by device stability fluctuation) and environmental light scattering interference (such as noise caused by leaf surface reflection) generated in the instrument measurement process. The key calculation involved here is to correct the baseline by polynomial fitting, and the mathematical expression is as follows:

[0049] ;

[0050] In the formula, x' i,λ represents the corrected absorbance value of the i-th sample at the λ-th wavelength point; x i,λ represents the original absorbance measurement value of the i-th sample at the λ-th wavelength point; d represents the order of polynomial fitting, which is used to describe the complexity of baseline drift, and is usually taken as 3; c k represents the k-th item coefficient of the polynomial baseline, which is estimated from the data by least square fitting; λ represents the wavelength index, which is used to identify the specific wavelength position in the spectrum; k is the summation index, which traverses each item of the polynomial from 0 to d.

[0051] On this basis, step S220 performs metabolite feature enhancement processing on the corrected spectral data, and for the lignification process (related to cell wall thickening associated with phenolic polymerization) and the terpene lactone synthesis path (such as the biosynthesis of seco-iridoid compounds to which oleuropein belongs) specific to olive leaf, the absorption bands corresponding to the characteristic functional groups (such as phenolic hydroxyl and enol ether bond) in these paths (for example, 280-320 nm in the ultraviolet region and specific bands in the visible-near infrared region) are screened, and the band weight distribution is allocated to amplify the response signal of the target component, to obtain a spectral matrix highlighting specific metabolite information; this processing is realized by weighted transformation, and the core operation can be represented as:

[0052] ;

[0053] In the formula, X enhanced represents the spectral data matrix after feature enhancement, the rows correspond to samples, and the columns correspond to wavelength points; X corr represents the spectral data matrix after baseline correction, which is the input of the enhancement processing; the table represents Hadamard product (element-by-element multiplication) for weighting each element in the matrix; 1 m represents an m-dimensional all-1 column vector, where m is the number of samples, used to extend the weight vector; w represents the wavelength weight vector; T represents the transpose operation.

[0054] Further, step S230 performs blind source signal separation processing based on the spectral matrix, and uses an independent component analysis algorithm to decouple the mixed spectral signal into independent source signals, so as to analyze the fluctuation mode unique to each of total phenol, total flavone and oleuropein, and finally obtain the pure change trajectory of each component free from co-extraction interference. The mathematical model of the separation process is as follows:

[0055] ;

[0056] In the formula, represents the transpose of the spectral matrix after feature enhancement; A represents a mixing matrix describing the mixing ratio of each independent source signal in the observed spectrum; S represents a source signal matrix, each row of which corresponds to the change trajectory of an independent component (such as total phenol, total flavone or oleuropein).

[0057] Further, step S300 includes steps S310 to S330.

[0058] Step S310 performs periodic mode extraction processing according to the original meteorological data, and obtains meteorological mode components representing environmental stress rhythms by decomposing the time series data of temperature, humidity and sunshine hours into intrinsic mode functions with different periodic characteristics;

[0059] Step S320, time lag correlation analysis processing is performed according to the meteorological modal component and the pure change trajectory, the mutual correlation function of each mode and the component trajectory at multiple time shift points is calculated, the key lag mode driving the differential expression of the secondary metabolite synthesis enzyme gene is identified, and the mode-component correlation pair matched with the time lag is obtained;

[0060] Step S330, the driving intensity quantification processing is performed according to the mode-component correlation pair, the nonlinear transfer function with the mode intensity as the input and the component change rate as the output is established, and the differential driving function of the meteorological mode on the component metabolic flow is fitted.

[0061] Specifically, step S310 first performs periodic modal extraction processing on the original meteorological data, and decomposes the continuous observation sequences of temperature, humidity and sunshine hours into a series of intrinsic modal functions with different time scale characteristics (such as daily cycle, seasonal cycle, etc.) through empirical mode decomposition and other methods (these functions are oscillation components derived from the data itself, which can capture the rhythmic changes of environmental stress), to obtain meteorological modal components representing the dynamic rhythm of environmental stress. The mathematical expression involved in this process is as follows:

[0062] ;

[0063] In the formula, r p-1 (t) represents the residual signal after the p-1th screening, and r0(t) is the original meteorological time series data (such as temperature sequence) when the first screening is performed. m p (t) represents the new residual signal obtained after the pth screening; t represents the time index. The iteration process continues until r p (t) satisfies the definition condition of the intrinsic modal function, and at this time is the extracted meteorological modal component.

[0064] On this basis, step S320 utilizes these meteorological modal components and the obtained pure change trajectory to perform time lag correlation analysis processing, calculates the cross-correlation function (a statistical tool for measuring the similarity of time series under different lags) of each mode and the component trajectory at multiple time shift points, identifies those lag modes that are significantly correlated in time with the expression fluctuation of the olive leaf secondary metabolite (such as phenols, flavonoids) synthesis enzyme gene (for example, some modes may affect the component accumulation in advance for several weeks), and obtains the mode-component correlation pair matched with the time lag; the core of this step is to calculate the cross-correlation function to quantify the similarity under different time lags, and the calculation formula is as follows:

[0065] ;

[0066] In the formula, R xy(τ) represents the cross-correlation coefficient of the meteorological mode sequence x and the composition change trajectory sequence y at the time delay τ, the closer the absolute value is to 1, the stronger the linear correlation is; τ represents the time delay (or lead) step number, τ>0 indicates that the meteorological mode leads the composition change; x t and y t respectively represent the values of the meteorological mode and the composition trajectory at time point t; and represent the sample means of sequences x and y respectively; y t+τ represents the value of the composition trajectory at time point t+τ; N represents the total length of the time series. The calculation traverses a set of τ values to identify the time delay that makes reach the maximum value, thereby determining the dominant time difference between the mode and the composition change.

[0067] Further, step S330 performs driving intensity quantification processing based on the correlation pairs, and by establishing a nonlinear transfer function (such as using a neural network or polynomial fitting to characterize a complex dose-response relationship) with mode intensity as input and composition change rate as output, a differentiated driving function of meteorological modes on composition metabolic flow is fitted, thereby accurately quantifying the dynamic influence intensity of different environmental stress modes on olive leaf active ingredient accumulation. A representative nonlinear model expression is as follows:

[0068] ;

[0069] In the formula, Δy represents the change rate (output) of the target composition; f(·) and represent nonlinear activation functions, such as Sigmoid or ReLU functions, for characterizing complex dose-response relationships; M represents the intensity of one or more meteorological modes; w j and b j are the weight and bias parameters connected to the jth hidden unit respectively; β j is the weight parameter connecting the jth hidden unit to the output layer. The parameters of the model are learned from the “mode-composition” correlation pair data by minimizing the prediction error (such as using gradient descent method), and the fitted function is the differentiated driving function of meteorological modes on composition metabolic flow.

[0070] Further, step S400 includes steps S410 to S430.

[0071] Step S410, according to the differentiated driving function, performs metabolic node network construction processing, and by taking each meteorological mode as a regulation node and the driving function value as the connection weight between nodes, an initial metabolic distribution network reflecting the meteorological stress signal transmission path is constructed, and a meteorological-metabolic interaction network is obtained;

[0072] Step S420, resource competition dynamic evolution processing is performed according to the meteorological-metabolic interaction network, a node activity attenuation and reconnection mechanism is introduced to simulate the competitive allocation process of limited assimilation resources between defensive phenolic compound synthesis and basic metabolism, and a metabolic resource allocation dynamic network evolving with time is obtained;

[0073] Step S430, structure feature emergence identification processing is performed according to the metabolic resource allocation dynamic network, by analyzing the change law of network modularity degree and path stability with meteorological stress intensity, a structural emergence feature representing the critical point of metabolic allocation strategy conversion is extracted.

[0074] Preferably, step S410 first performs metabolic node network construction processing based on a differentiated driving function, meteorological modes with different periodic characteristics are taken as control nodes in the network, and the driving intensity of each mode on component metabolism is quantified as the connection weight between nodes, so as to construct an initial network structure capable of reflecting the transmission path of meteorological stress signals in the metabolic system, and the meteorological-metabolic interaction network directly shows the synergistic or antagonistic relationship of different environmental factors on the secondary metabolic process of olive leaves; on this basis, step S420 performs resource competition dynamic evolution processing on the network, by introducing a node activity attenuation mechanism (simulating the adaptive adjustment of metabolic pathways with environmental changes) and a connection reconnection mechanism (reflecting the dynamic optimization of resource allocation strategy), the competitive allocation process of limited assimilation resources between defensive phenolic compound synthesis and basic metabolic activity is truly simulated, which particularly embodies the physiological adaptability of resource allocation of olive leaves when responding to environmental stress, and finally a metabolic resource allocation network evolving with time is obtained; step S430 deeply mines the dynamic network, by analyzing the change law of network modularity degree (reflecting the aggregation characteristics of functional blocks) and path stability (characterizing the material and energy transport efficiency) with meteorological stress intensity, the turning points when the network structure is reorganized significantly are identified, which correspond to the critical state of fundamental conversion of metabolic allocation strategy of olive leaves, and the structural emergence feature extracted therefrom provides a new observation dimension for understanding the adaptive response of plants to environmental stress.

[0075] Further, step S500 includes step S510 to step S530.

[0076] Step S510, coupled system state vector construction processing is performed according to the structural emergence feature and the pure change trajectory, by combining the structural emergence feature of metabolic allocation and the component change trajectory into a multi-dimensional state variable, a state vector sequence describing the dynamic relationship between source and sink is obtained;

[0077] Step S520, delay-embedding phase space reconstruction processing is performed according to the state vector sequence, a high-dimensional phase space orbit retaining the dynamic characteristics of the system is constructed by calculating the delay parameter and embedding dimension of the state variable, and a phase space trajectory of the system evolution is obtained;

[0078] Step S530, attractor topological feature extraction processing is performed according to the phase space trajectory, geometric invariants representing the source-sink relationship conversion are identified by analyzing the aggregation form and curvature variation law of the trajectory, and a phase space geometric descriptor of the source-sink relationship critical state is obtained.

[0079] It should be noted that step S510 first performs state vector construction processing of the coupled system based on the structural emergence characteristics and the pure change trajectory, multi-dimensionally integrates the structural characteristics (such as modularization degree, path stability, and other topological indexes) generated in the metabolic allocation network evolution process and the concentration change trajectory of each target component, forms a state variable set that can comprehensively reflect the material transportation and distribution dynamics among the source-sink organs, and thus constructs a state vector sequence describing the dynamic evolution process of the olive leaf source-sink system; step S520 then performs delay-embedding phase space reconstruction processing on the state vector sequence, expands the one-dimensional time sequence into a high-dimensional phase space orbit that can retain the intrinsic dynamic characteristics of the system by calculating the optimal delay time parameter and suitable embedding dimension between the state variables, and obtains a phase space trajectory representing the dynamic behavior of the olive leaf source-sink system through the phase space reconstruction technology; step S530 performs attractor topological feature extraction processing on the basis of this, identifies geometric invariants (such as Lyapunov exponent, fractal dimension, and other topological invariants) representing the fundamental conversion of the source-sink relationship by analyzing the aggregation form (such as the distribution density of the trajectory points) and the curvature variation law (reflecting the degree of change in the system state) of the phase space trajectory in the attractor region, these geometric descriptors can accurately capture the critical features of the system from one steady state to another steady state, and finally obtain a set of phase space geometric descriptors for determining the critical state of the source-sink relationship.

[0080] Further, step S600 includes step S610 to step S630.

[0081] Step S610, state mapping processing is performed according to the phase space geometric descriptor and the current observation state, the position coordinates of the current state in the historical geometric descriptor manifold are determined by projecting the component dynamic data of the current growth period into the source-sink relationship phase space, and a current state mapping result is obtained;

[0082] Step S620, harvesting path planning processing is performed according to the current state mapping result, the geodesic line from the current state point to the boundary point of the optimal harvesting region is calculated, and the distribution characteristics of the source-sink relationship conversion critical point on the path are analyzed, and an optimal harvesting path trajectory is obtained.

[0083] Step S630, according to the optimal recovery path trajectory, a time window optimization process is performed, the projection of the path trajectory on the time axis is extracted, the time period when the compound accumulation rate and the oxidation stability reach a balance and is located before the concentration peak is identified, and the best sampling time interval is obtained.

[0084] In the preferred embodiment, step S610 first constructs a phase space geometric descriptor based on historical data and observation data of the current growth period, projects the real-time monitored component dynamic information into the established source-sink relationship phase space through state mapping processing, determines the specific coordinate position of the current state on the multi-dimensional manifold formed by the historical geometric descriptor using manifold learning algorithm, and realizes the accurate docking of real-time observation data and historical law system, which provides an accurate initial state positioning for prediction analysis. On this basis, step S620 performs recovery path planning processing according to the current state mapping result, calculates the geodesic line (the shortest path connecting two points on the manifold) from the current state point to the boundary point of the historical optimal recovery area by differential geometry method, and analyzes the distribution density and variation trend of the source-sink relationship conversion critical point on the path, so as to determine the most ideal evolution path in the component accumulation process, and obtain the optimal recovery path trajectory considering the path length and state conversion stability. Step S630 then performs time window optimization processing based on the foregoing path planning, maps the path trajectory in the geometric space back to the time dimension, analyzes the time sequence characteristics corresponding to the trajectory points, identifies the key time period when the compound accumulation rate curve and the oxidation stability curve reach the best balance state and are located before the component concentration peak, and finally determines the best sampling time interval that can ensure the component content and maintain the stability of the component.

[0085] Embodiment 2:

[0086] As shown in Figure 2 The present embodiment provides an olive leaf component seasonal change trend prediction system, which comprises:

[0087] The acquisition module 901 is configured to acquire original spectral absorbance data directly measured at non-uniform sampling time points in multiple complete growth years of the target variety of olive leaves, and original meteorological data continuously monitored at the same period.

[0088] The separation module 902 is configured to separate the component-specific signals according to the original spectral absorbance data to obtain pure change trajectories of each target component.

[0089] The decomposition module 903 is configured to decompose the environment stress modal according to the original meteorological data and the pure change trajectory to obtain a differential driving function of the meteorological modal on the component metabolic flow.

[0090] an evolution module 904, configured to perform metabolic resource allocation network evolution according to the differential driving function, to obtain a structural emergent feature of metabolic allocation under meteorological stress;

[0091] a reconstruction module 905, configured to perform phase space reconstruction of a xylem sap flow-phloem loading coupling system according to the structural emergent feature and the pure change trajectory, to obtain a phase space geometric descriptor of a source-sink relationship critical state;

[0092] a decision module 906, configured to perform optimal harvesting window decision-making according to the phase space geometric descriptor, to predict an optimal sampling time interval in a current growth period by calculating a geodesic line path from a current observation state to a historical optimal harvesting region.

[0093] In an embodiment of the present application, the separation module 902 includes:

[0094] a first separation unit, configured to perform spectral data correction processing according to original spectral absorbance data, to obtain corrected spectral data by eliminating instrument baseline drift and ambient light scattering interference;

[0095] a second separation unit, configured to perform metabolite feature enhancement processing according to the corrected spectral data, to obtain a feature-enhanced spectral matrix by performing band selection and weighting on feature absorption bands related to lignification process and terpene lactone synthesis pathway;

[0096] a third separation unit, configured to perform blind source signal separation processing according to the spectral matrix, to obtain pure change trajectories of each target component by independent component analysis to analyze independent fluctuation source signals of total phenol, total flavonoids and oleuropein.

[0097] In an embodiment of the present application, the decomposition module 903 includes:

[0098] a first decomposition unit, configured to perform periodic mode extraction processing according to original meteorological data, to obtain meteorological mode components representing environmental stress rhythms by decomposing time series data of temperature, humidity and sunshine hours into intrinsic mode functions with different periodic characteristics;

[0099] a second decomposition unit, configured to perform time lag correlation analysis processing according to the meteorological mode components and the pure change trajectory, to identify key lag modes driving differential expression of secondary metabolite synthesis enzyme genes by calculating cross-correlation functions of each mode and component trajectory at multiple time shift points, to obtain mode-component correlation pairs matched in time lag;

[0100] a third decomposition unit, configured to perform driving intensity quantification processing according to the mode-component correlation pairs, to fit a differential driving function of meteorological modes on component metabolic flow by establishing a nonlinear transfer function with mode intensity as input and component change rate as output.

[0101] In an embodiment of the present application, the evolution module 904 includes:

[0102] The first evolution unit is configured to perform metabolic node network construction processing according to the differentiation driving function, to construct an initial metabolic allocation network reflecting the weather stress signal transmission path by taking each weather mode as a regulation node and the driving function value as a connection weight between nodes, and to obtain a weather-metabolism interaction network;

[0103] The second evolution unit is configured to perform resource competition dynamic evolution processing according to the weather-metabolism interaction network, to simulate the competition and allocation process of limited assimilation resources between defensive phenol synthesis and basic metabolism by introducing a node activity decay and reconnection mechanism, and to obtain a metabolic resource allocation dynamic network evolving over time;

[0104] The third evolution unit is configured to perform structural feature emergence identification processing according to the metabolic resource allocation dynamic network, to extract structural emergence features representing the metabolic allocation strategy conversion critical point by analyzing the variation law of the network modularity degree and path stability with the weather stress intensity.

[0105] In an embodiment of the present application, the reconstruction module 905 includes:

[0106] The first reconstruction unit is configured to perform coupled system state vector construction processing according to the structural emergence features and the pure change trajectory, to combine the structural emergence features of metabolic allocation and the component change trajectory into a multi-dimensional state variable, and to obtain a state vector sequence describing the source- sink dynamic relationship;

[0107] The second reconstruction unit is configured to perform delay embedding phase space reconstruction processing according to the state vector sequence, to calculate the delay parameter and embedding dimension of the state variable, to construct a high-dimensional phase space trajectory preserving the system dynamics characteristics, and to obtain a phase space trajectory of system evolution;

[0108] The third reconstruction unit is configured to perform attractor topological feature extraction processing according to the phase space trajectory, to identify geometric invariants representing the source-sink relationship conversion by analyzing the variation law of the trajectory aggregation form and curvature, and to obtain a phase space geometric descriptor of the source-sink relationship critical state.

[0109] In an embodiment of the present application, the decision module 906 includes:

[0110] The first decision unit is configured to perform state mapping processing according to the phase space geometric descriptor and the current observation state, to project the component dynamic data of the current growth period into the source-sink relationship phase space, to determine the position coordinates of the current state in the historical geometric descriptor manifold, and to obtain a current state mapping result;

[0111] The second decision-making unit is used to perform harvesting path planning based on the current state mapping results. It calculates the geodesic line from the current state point to the boundary point of the optimal harvesting area and analyzes the distribution characteristics of the source-sink relationship conversion critical points on the path to obtain the optimal harvesting path trajectory.

[0112] The third decision unit is used to optimize the time window based on the optimal harvesting path trajectory. By extracting the projection of the path trajectory on the time axis, it identifies the time period when the compound accumulation rate and oxidation stability reach a balance and are located before the concentration peak, thus obtaining the optimal sampling time interval.

[0113] Example 3:

[0114] Corresponding to the above method embodiments, this embodiment also provides a device for predicting the seasonal variation trend of olive leaf components. The device for predicting the seasonal variation trend of olive leaf components described below and the method for predicting the seasonal variation trend of olive leaf components described above can be referred to in correspondence.

[0115] Figure 3 This is a block diagram illustrating a device 800 for predicting the seasonal variation trend of olive leaf components according to an exemplary embodiment. Figure 3 As shown, the olive leaf component seasonal variation trend prediction device 800 may include: a processor 801 and a memory 802. The olive leaf component seasonal variation trend prediction 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.

[0116] The processor 801 is configured to control overall operations of the olive leaf component seasonal variation trend prediction device 800 to complete all or part of the steps in the above-described olive leaf component seasonal variation trend prediction method. The memory 802 is configured to store various types of data to support operations of the olive leaf component seasonal variation trend prediction device 800, which can include, for example, instructions for any application or method operating on the olive leaf component seasonal variation trend prediction device 800, and application-related data, such as contact data, sent and received messages, pictures, audio, video, and the like. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The multimedia component 803 can include a screen and an audio component. The screen can be, for example, a touch screen, and the audio component is configured to output and / or input audio signals. For example, the audio component can include a microphone configured to receive external audio signals. The received audio signals can be further stored in the memory 802 or transmitted through the communication component 805. The audio component also includes at least one speaker configured to output audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, which can be a keyboard, a mouse, a button, and the like. The buttons can be virtual buttons or physical buttons. The communication component 805 is configured to perform wired or wireless communication between the olive leaf component seasonal variation trend prediction device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include a Wi-Fi module, a Bluetooth module, an NFC module.

[0117] In an exemplary embodiment, an olive leaf component seasonal variation trend prediction device 800 can be implemented by one or more Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor or other electronic elements for executing the above-mentioned olive leaf component seasonal variation trend prediction method.

[0118] In another exemplary embodiment, a computer readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the above-mentioned olive leaf component seasonal variation trend prediction method. For example, the computer readable storage medium can be the above-mentioned memory 802 including program instructions, which can be executed by the processor 801 of the above-mentioned olive leaf component seasonal variation trend prediction device 800 to complete the above-mentioned olive leaf component seasonal variation trend prediction method.

[0119] The above description is merely a specific implementation of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the scope of protection of the present application.

Claims

1. A method for predicting seasonal variation trends of olive leaf components, characterized by, The method comprises the following steps: Obtaining original spectral absorbance data directly measured at multiple complete growth years and non-uniform sampling time points of target variety olive leaf, and original meteorological data continuously monitored at the same period; Performing component-specific signal separation according to the original spectral absorbance data to obtain pure change trajectories of each target component; Performing environmental stress mode decomposition according to the original meteorological data and the pure change trajectories to obtain differential driving functions of meteorological modes on component metabolic flow; Performing metabolic resource allocation network evolution according to the differential driving functions to obtain structural emergent characteristics of metabolic allocation under meteorological stress; Performing phase space reconstruction of xylem sap flow-xylem loading coupling system according to the structural emergent characteristics and the pure change trajectories to obtain phase space geometric descriptors of critical state of source-sink relationship; Performing optimal harvesting window decision according to the phase space geometric descriptors, and predicting the optimal sampling time interval in the current growth cycle by calculating the geodesic path from the current observation state to the historical optimal harvesting region.

2. The method of claim 1, wherein the method is characterized by: The method comprises the following steps: Performing spectral data correction processing according to the original spectral absorbance data to eliminate instrument baseline drift and environmental light scattering interference, and obtaining corrected spectral data; Performing metabolite feature enhancement processing according to the corrected spectral data, selecting and weighting the characteristic absorption bands related to lignification process and terpene lactone synthesis pathway to obtain a feature-enhanced spectral matrix; Performing blind source signal separation processing according to the spectral matrix, analyzing the independent fluctuation source signals of total phenol, total flavonoids and oleuropein by independent component analysis to obtain the pure change trajectories of each target component.

3. The method of claim 1, wherein the method is characterized by: The method comprises the following steps: Performing periodic mode extraction processing according to the original meteorological data, decomposing the time series data of temperature, humidity and sunshine hours into intrinsic mode functions with different periodic characteristics to obtain meteorological mode components representing environmental stress rhythm; Performing time lag correlation analysis processing according to the meteorological mode components and the pure change trajectories, calculating the cross-correlation functions of each mode and component trajectory at multiple time shift points to identify the key lag mode driving differential expression of secondary metabolite synthesis enzyme genes, and obtaining mode-component correlation pairs matched in time lag; Performing driving intensity quantification processing according to the mode-component correlation pairs, establishing a nonlinear transfer function with mode intensity as input and component change rate as output to fit the differential driving functions of meteorological modes on component metabolic flow.

4. The method of claim 1, wherein the method is characterized by, The method comprises the following steps: Performing metabolic resource allocation network evolution according to the differential driving functions to obtain structural emergent characteristics of metabolic allocation under meteorological stress, including: According to the differential driving function, a metabolic node network construction process is performed, an initial metabolic allocation network reflecting a meteorological stress signal transmission path is constructed by taking each meteorological mode as a regulation node and a driving function value as a connection weight between nodes, and a meteorological-metabolic interaction network is obtained; According to the meteorological-metabolic interaction network, a resource competition dynamic evolution process is performed, a node activity attenuation and reconnection mechanism is introduced to simulate a competitive allocation process of limited assimilation resources between defensive phenolic synthesis and basic metabolism, and a metabolic resource allocation dynamic network evolving over time is obtained; According to the metabolic resource allocation dynamic network, a structural feature emergence identification process is performed, by analyzing the change law of the network modularity degree and the path stability with the meteorological stress intensity, a structural emergence feature representing a metabolic allocation strategy conversion critical point is extracted.

5. The method of claim 1, wherein the method is characterized by: According to the structural emergence feature and the pure change trajectory, a phase space reconstruction of the xylem sap flow-phloem loading coupling system is performed, and a phase space geometric descriptor of the source-sink relationship critical state is obtained, including: According to the structural emergence feature and the pure change trajectory, a coupling system state vector construction process is performed, a structural emergence feature of metabolic allocation and a component change trajectory are combined into a multi-dimensional state variable, and a state vector sequence describing the dynamic relationship between the source and the sink is obtained; According to the state vector sequence, a delay-embedded phase space reconstruction process is performed, a delay parameter and an embedding dimension of the state variable are calculated, a high-dimensional phase space trajectory retaining the dynamic characteristics of the system is constructed, and a phase space trajectory of the system evolution is obtained; According to the phase space trajectory, an attractor topological feature extraction process is performed, by analyzing the change law of the trajectory aggregation form and the curvature, a geometric invariant representing the source-sink relationship conversion is identified, and a phase space geometric descriptor of the source-sink relationship critical state is obtained.

6. An oil olive leaf component seasonal change trend prediction system characterized by comprising: Including: An acquisition module is configured to acquire original spectral absorbance data directly measured at non-uniform sampling time points in multiple complete growth years of target variety olive leaves, and original meteorological data continuously monitored at the same period; A separation module is configured to separate component-specific signals according to the original spectral absorbance data to obtain pure change trajectories of each target component; A decomposition module is configured to decompose environmental stress modes according to the original meteorological data and the pure change trajectories to obtain differential driving functions of the meteorological modes on the component metabolic flow; An evolution module is configured to evolve a metabolic resource allocation network according to the differential driving functions to obtain a structural emergence feature of metabolic allocation under meteorological stress; A reconstruction module is configured to reconstruct a phase space of a xylem sap flow-phloem loading coupling system according to the structural emergence feature and the pure change trajectories to obtain a phase space geometric descriptor of a source-sink relationship critical state; A decision module is configured to make a decision on an optimal harvesting window according to the phase space geometric descriptor by calculating a geodesic line path from a current observation state to a historical optimal harvesting region to predict an optimal sampling time interval in a current growth period.

7. The system for predicting seasonal variation trend of olive leaf components according to claim 6, characterized in that, The separation module includes: A first separation unit is configured to perform a spectral data correction process on the original spectral absorbance data, to eliminate instrument baseline drift and ambient light scattering interference, and to obtain corrected spectral data; A second separation unit is configured to perform a metabolite feature enhancement process on the corrected spectral data, to select and weight characteristic absorption bands related to lignification process and terpene lactone synthesis pathway, and to obtain a feature-enhanced spectral matrix; A third separation unit is configured to perform a blind source signal separation process on the spectral matrix, to analyze independent fluctuation source signals of total phenol, total flavonoids and oleuropein by independent component analysis, and to obtain pure change trajectories of each target component.

8. The system for predicting seasonal variation trend of olive leaf components according to claim 6, characterized in that, The decomposition module comprises: A first decomposition unit is configured to perform a periodic modal extraction process on the original meteorological data, to decompose time series data of temperature, humidity and sunshine hours into intrinsic modal functions with different periodic characteristics, and to obtain meteorological modal components representing environmental stress rhythms; A second decomposition unit is configured to perform a time lag correlation analysis process on the meteorological modal components and the pure change trajectories, to calculate cross-correlation functions of each modal and component trajectory at multiple time shift points, to identify key lag modes driving differential expression of secondary metabolite synthesis enzyme genes, and to obtain time-lag matched modal-component correlation pairs; A third decomposition unit is configured to perform a driving intensity quantification process on the modal-component correlation pairs, to establish a nonlinear transfer function with modal intensity as input and component change rate as output, and to fit a differential driving function of meteorological modal on component metabolic flow.

9. The system for predicting seasonal variation trend of olive leaf components according to claim 6, characterized in that, The evolution module comprises: A first evolution unit is configured to perform a metabolic node network construction process on the differential driving function, to construct an initial metabolic allocation network reflecting meteorological stress signal transmission paths by taking each meteorological modal as a control node and driving function value as a connection weight between nodes, and to obtain a meteorological-metabolic interaction network; A second evolution unit is configured to perform a resource competition dynamic evolution process on the meteorological-metabolic interaction network, to simulate the competition and allocation process of limited assimilation resources between defensive phenolic synthesis and basic metabolism by introducing node activity decay and reconnection mechanisms, and to obtain a time-evolving metabolic resource allocation dynamic network; A third evolution unit is configured to perform a structure feature emergence identification process on the metabolic resource allocation dynamic network, to analyze the change law of network modularity and path stability with meteorological stress intensity, and to extract structural emergence features representing the critical point of metabolic allocation strategy conversion.

10. The system for predicting seasonal variation trend of olive leaf components according to claim 6, characterized in that, The reconstruction module comprises: A first reconstruction unit is configured to perform a coupled system state vector construction process on the structural emergence features and the pure change trajectories, to combine the structural emergence features of metabolic allocation and the component change trajectories into multi-dimensional state variables, and to obtain a state vector sequence describing the dynamic relationship between source and sink; A second reconstruction unit is configured to perform a delay-embedded phase space reconstruction process on the state vector sequence, to calculate delay parameters and embedding dimensions of state variables, to construct a high-dimensional phase space trajectory preserving the dynamic characteristics of the system, and to obtain a phase space trajectory of system evolution. The third reconstruction unit is configured to perform attractor topological feature extraction processing according to the phase space trajectory, identify geometric invariants representing source-reservoir relationship conversion by analyzing the aggregation form and curvature variation law of the trajectory, and obtain phase space geometric descriptors of the source-reservoir relationship critical state.

Images