A method for analyzing an index of phenotypic digital diversity of plants

Abstract

The present application relates to the technical field of plant phenotype analysis, and particularly relates to a plant phenotype digital diversity index analysis method. The method effectively solves the problems of single data dimension, discretization limitation, insufficient dynamics and lack of system maintenance scheme in the existing plant phenotype analysis method. The method collects environment-phenotype dynamic data by constructing a heterogeneous monitoring network, converts discrete phenotype data into continuous probability space by combining a three-dimensional ellipsoid domain model, realizes accurate evaluation and dynamic maintenance of plant phenotype diversity through sub-domain decomposition, dynamic correction, heat map generation, deep tracking and inverse mapping analysis. The present application can effectively capture genotype-environment interaction characteristics, accurately locate phenotype response risk sites and lag units, and provide a systematic solution for plant variety-environment adaptation and diversity maintenance, thereby improving the dynamics, accuracy and practicality of plant phenotype analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plant phenotype analysis, and in particular to a plant phenotype digital diversity index analysis method. BACKGROUND

[0002] Plant phenotype diversity is a core indicator for species to adapt to environmental fluctuations and maintain ecological stability. Its accurate analysis is of great significance for agricultural breeding, germplasm resource protection and ecosystem management. Traditional phenotype analysis methods rely on manual measurement or single sensor data, and focus on static morphological indicators such as plant height and leaf area. However, they cannot capture dynamic phenotype characteristics such as canopy spectral response, leaf dynamic oscillation and root electrical signal conduction. Therefore, the correlation analysis between environmental factors and phenotype response has dimensional limitations, and cannot fully reflect the complex mechanism of genotype and environment interaction.

[0003] The existing technology has significant bottlenecks in data processing: discrete phenotype data is difficult to convert into a continuous probability distribution model, making it impossible to quantitatively express the phenotype evolution rule under environmental stress gradient. At the same time, the strength of genotype and environment interaction is estimated by empirical formula, and there is a lack of visual mathematical model, making it difficult to accurately locate the fluctuation risk points and genetic advantage areas of phenotype diversity, which restricts the depth and application value of phenotype analysis.

[0004] In terms of dynamic response and decision support, traditional methods have a lag in phenotype feedback to environmental stress, and cannot real-time correct key parameters such as stomatal conductance threshold. Moreover, there is a lack of a systematic method to deduce optimal environmental parameters from phenotype characteristics, resulting in a dependence on empirical judgment for the adaptation of varieties and environment, and making it difficult to achieve active maintenance and accurate regulation of phenotype diversity. Therefore, it is an urgent need in the field of plant phenotype research to develop a digital analysis method that integrates multi-dimensional dynamic data, quantifies interaction mechanisms, real-time tracks responses and generates decision-making schemes. SUMMARY

[0005] The present application provides a plant phenotype digital diversity index analysis method, which effectively solves the problems of single data dimension, discrete limitation, lack of dynamicity and lack of systematic maintenance scheme in existing plant phenotype analysis methods. This method collects environment-phenotype dynamic data by building a heterogeneous monitoring network, converts discrete phenotype data into a continuous probability space by using a three-dimensional ellipsoid model, and realizes accurate evaluation and dynamic maintenance of plant phenotype diversity through sub-domain decomposition, dynamic correction, heat map generation, deep tracking and inverse mapping analysis. The present application can effectively capture genotype-environment interaction characteristics, accurately locate phenotype response risk points and lag units, and provide a systematic solution for plant variety-environment adaptation and diversity maintenance, thereby improving the dynamicity, accuracy and practicality of plant phenotype analysis.

[0006] The technical solution of the present application is as follows:

[0007] The technical scheme of the present application provides a plant phenotype digital diversity index analysis method, comprising:

[0008] The plant phenotype analysis constructs a heterogeneous monitoring network through a plant digital phenotype collection and analysis system, cooperatively collects multispectral radiation data above a plant canopy, leaf oscillation frequency, and root system electrical signal topology, and generates an environment-phenotype correlation matrix;

[0009] Based on the environment-phenotype dynamic matrix, genotype and environmental stress parameters are extracted in combination with a historical variety database to construct a three-dimensional ellipsoidal domain model;

[0010] According to the three-dimensional ellipsoidal domain model, non-uniform subdomain decomposition is implemented along the principal axis direction: high-resolution subdivision is performed in the environment stress severe area, and low-resolution subdivision is performed in the stable area, so that each subdomain corresponds to a unique phenotype response mode feature, forming a phenotype response fingerprint library;

[0011] For the divided subdomains, a dedicated plant digital phenotype collection and analysis system tracking unit is configured, and when leaf oscillation abnormalities are detected, root impedance scanning is triggered, and the subdomain response boundary is dynamically determined based on the organ cascade feedback mechanism;

[0012] Based on the three-dimensional ellipsoidal domain model, the subdomains and corresponding subdomain response boundaries are labeled to generate a three-dimensional heat map;

[0013] In response to the three-dimensional heat map result, deep tracking is started in the red high-risk area, and the deep tracking mode is: controlling the plant digital phenotype collection and analysis system to capture hidden stress sources along the canopy vorticity, and positioning the phenotype response lag unit;

[0014] Based on the phenotype characteristics of the blue stable area and the diagnosis result of the lag unit, the optimal environmental parameter combination is backstepped through inverse ellipsoid mapping, a variety-environment adaptation matrix is generated, and a diversity maintenance scheme is output.

[0015] As a further option of the method, the multispectral radiation data collection uses a multispectral imaging system carrying multiple narrowband wavelength sensors to obtain plant canopy reflectance information in the visible to near-infrared band, and through chlorophyll content, water state, and photosynthetic efficiency modeling, the digital representation of key physiological parameters is realized;

[0016] The leaf oscillation frequency measurement uses a high-speed camera and an inertial measurement unit to jointly capture the leaf motion trajectory, and through Fourier transform, the main vibration spectrum is analyzed;

[0017] The root system electrical signal topology analysis includes: using a microelectrode array implanted in the soil to measure rhizosphere conductivity changes, combining ion flow analysis and action potential recording, and evaluating the ability of the root system to absorb water and nutrients.

[0018] As a further option of this method, the construction of the environment-phenotype dynamic matrix includes time alignment, spatial standardization and dimensionality reduction of multi-source data, wherein the dimensionality reduction method uses principal component analysis or wavelet transform to extract key phenotypic features;

[0019] The environment-phenotype dynamic matrix contains the joint distribution of environmental variables and phenotypic characteristics, which is expressed as:

[0020] ;

[0021] in, represents the environment-phenotype dynamic matrix, Indicates the The sample in measurements of environmental-phenotypic traits, is the sample size, is the characteristic dimension.

[0022] As a further option of this method, the construction of the three-dimensional ellipsoid domain model includes:

[0023] Based on the joint covariance matrix of the genotype data matrix and the environmental stress parameter matrix, the eigenvalues ​​and eigenvectors were calculated to determine the direction and scale of the ellipsoid's major axis.

[0024] The curvature of the ellipsoid is quantified by the ratio of the eigenvalues ​​of any two principal axes and is used to characterize the strength of the genotype-environment interaction;

[0025] The standard equation of the three-dimensional ellipsoid domain model is:

[0026] ;

[0027] in, is the genotype feature vector, is the environmental stress parameter vector, is the inverse of the joint covariance matrix;

[0028] Among them, the three-dimensional ellipsoid domain model distinguishes different types of phenotypic response patterns through the ellipsoid curvature and principal axis gradient. The ellipsoid stretching direction represents the area dominated by environmental stress, while the high curvature direction reflects the area with strong genetic stability.

[0029] As a further option of this method, the non-uniform subdomain decomposition strategy uses an adaptive meshing algorithm to dynamically adjust the subdomain resolution based on the ellipsoid principal axis gradient function and local curvature evaluation; quadtree / octree recursive subdivision is implemented in high-gradient regions to achieve refined modeling and classification of stress response patterns;

[0030] The non-uniform subdomain division includes: calculating the gradient and curvature in the main axis direction for each subdomain, identifying areas with severe environmental stress and improving the resolution, while retaining the low-resolution division of stable areas to balance computational efficiency and modeling accuracy.

[0031] As a further option of this method, the phenotypic response fingerprint library construction includes:

[0032] PDPAS uses autoencoder technology to encode the extracted phenotypic response features to generate compact and representative fingerprint vectors;

[0033] PDPAS introduces a locality-sensitive hashing algorithm to map feature vectors into a hash space, so that similar feature vectors have a higher probability of collision.

[0034] The PRFD system uses a mini-batch stochastic gradient descent algorithm to regularly update the parameters of the autoencoder to ensure that the encoder can adapt to the latest phenotypic response characteristics;

[0035] PDPAS combines phenotypic response patterns with known genotype-environment interactions to construct a semantic network, where nodes represent phenotypic response patterns and edges represent similarities or causal relationships between patterns.

[0036] The key features stored in the phenotypic response fingerprint library include: spectral oscillation ratio, vibration damping coefficient, root conductivity entropy and other highly distinguishable phenotypic response indicators.

[0037] As a further option of this method, the organ cascade feedback mechanism constructs a cross-organ coupling feedback model by integrating the real-time physiological signals of leaf oscillation and root impedance, and dynamically adjusts the subdomain response boundary based on the ellipsoid surface gradient field.

[0038] As a further option of this method, the three-dimensional heat map generation includes: using the three main axes of the ellipsoid domain model as a three-dimensional coordinate system, mapping each subdomain and its response boundary to the space, and using a color coding strategy, with red hot spots marking diversity fluctuation risk sites and blue stable areas marking genetic advantage expression sites;

[0039] The three-dimensional heat map introduces a knowledge graph to enhance the semantic annotation function, allowing users to interactively query the genotype background, historical response trends and potential stress types of the subdomain.

[0040] As a further option of this method, the in-depth tracking of hidden stress sources includes: using lidar and infrared thermal imaging technology to capture the trajectory of canopy vorticity changes, combining meteorological sensor data to identify areas of abnormal air flow, thereby locating potential hidden stress sources;

[0041] The positioning phenotype response lag unit compares the phenotype response modes between different sub-domains by using a dynamic time warping algorithm, and when the response mode of a certain sub-domain deviates significantly from the expectation, it is marked as a potential phenotype response lag unit;

[0042] The implicit stress source analysis includes: integrating multi-source sensor data, identifying the main environmental factors affecting the plant phenotype response by using multivariate regression analysis and principal component analysis, and automatically triggering the root impedance scanning module to obtain more detailed underground part physiological information.

[0043] As a further selection of the method, the inverse ellipsoid mapping back-propagates the optimal environmental parameter combination includes: an environmental parameter vector , a phenotype feature vector , an inverse matrix of a joint covariance matrix The formula of the inverse ellipsoid mapping is as follows:

[0044] ;

[0045] The variety-environment adaptation matrix construction includes: based on the inverse ellipsoid mapping result, evaluating the adaptation degree of different plant varieties under different environmental conditions, and the higher the adaptation degree value, the more adaptive the variety is in the environment.

[0046] The technical scheme provided by the embodiments of the present application has at least the following beneficial effects:

[0047] The technical scheme provides a systematic and multi-modal fusion plant phenotype digital diversity analysis method, which realizes heterogeneous perception of plant canopy multispectral radiation, leaf oscillation frequency and root electrical signal by constructing a plant digital phenotype acquisition and analysis system, and comprehensively captures dynamic response characteristics of plants under different environmental conditions. The method further combines a historical variety database, introduces three-dimensional ellipsoid domain modeling, non-uniform sub-domain decomposition and phenotype response fingerprint library construction, and breaks through the limitations of traditional phenotype analysis in spatial resolution, time dynamics and environmental interaction analysis capability, and significantly improves the identification accuracy and interpretability of plant phenotype response modes.

[0048] On this basis, the technical scheme realizes real-time tracking of plant stress response and back-propagation of the optimal environmental parameter combination by using an organ cascade feedback mechanism and an inverse ellipsoid mapping algorithm, constructs a variety-environment adaptation matrix, and forms a scientific diversity maintenance scheme. The system not only has high-precision and high-robustness environmental stress recognition and genetic advantage positioning capability, but also can provide data-driven decision support for crop breeding, precision agriculture and intelligent cultivation, and has wide application prospect and industrial value. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1A schematic diagram of the overall process of the index analysis method for plant phenotype digital diversity;

[0050] Figure 2 A detailed flowchart of the S100 step of the index analysis method for plant phenotype digital diversity;

[0051] Figure 3 A detailed flowchart of the S200 step of the index analysis method for plant phenotype digital diversity;

[0052] Figure 4 A detailed flowchart of the S300 step of the index analysis method for plant phenotype digital diversity;

[0053] Figure 5 A detailed flowchart of the S400 step of the index analysis method for plant phenotype digital diversity;

[0054] Figure 6 A detailed flowchart of the S500 step of the index analysis method for plant phenotype digital diversity;

[0055] Figure 7 A detailed flowchart of the S600 step of the index analysis method for plant phenotype digital diversity;

[0056] Figure 8 A detailed flowchart of the S700 step of the index analysis method for plant phenotype digital diversity. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present application will be clearly and completely described 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, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0058] Plant phenotype is the external manifestation of the interaction between plant genotype and environment, and its diversity is an important basis for species to adapt to the environment, maintain survival and evolution. Traditional plant phenotype analysis methods mostly rely on manual observation or single sensor data acquisition, and the existing problems include: single data dimension, discrete analysis limitation, lack of dynamic nature and lack of system maintenance scheme. To solve the above problems, please refer to Figure 1 which shows a plant phenotype digital diversity index analysis method provided by an embodiment of the present application, which comprises:

[0059] S100: The plant phenotype analysis constructs a heterogeneous monitoring network through a plant digital phenotype collection and analysis system, cooperatively collects multispectral radiation data above the plant canopy, leaf oscillation frequency and root system electrical signal topology, and generates an environment-phenotype correlation matrix.

[0060] S200: Extract genotype and environmental stress parameters based on the environment-phenotype dynamic matrix, combined with the historical variety database, and construct a three-dimensional ellipsoid domain model.

[0061] S300: According to the three-dimensional ellipsoid domain model, implement non-uniform sub-domain decomposition along the principal axis direction: high-resolution subdivision in areas with severe environmental stress, and low-resolution subdivision in stable areas, so that each sub-domain corresponds to a unique phenotype response mode feature, forming a phenotype response fingerprint library.

[0062] S400: For the divided sub-domain, configure a dedicated plant digital phenotype acquisition and analysis system tracking unit, and trigger root impedance scanning when detecting leaf oscillation abnormalities, and dynamically determine the sub-domain response boundary based on the organ cascade feedback mechanism.

[0063] S500: Based on the three-dimensional ellipsoid domain model, label the sub-domain and the corresponding sub-domain response boundary, and generate a three-dimensional heat map.

[0064] S600: In response to the three-dimensional heat map result, start deep tracking in the red high-risk area: control the plant digital phenotype acquisition and analysis system to capture hidden stress sources along the crown vorticity, and locate the phenotype response lag unit.

[0065] S700: Based on the phenotype characteristics of the blue stable area and the diagnosis results of the lag unit, reverse the optimal environmental parameter combination through inverse ellipsoid mapping, generate a variety-environment adaptation matrix, and output a diversity maintenance scheme.

[0066] The specific scheme is as follows:

[0067] In an index analysis method for plant phenotype digital diversity, the plant digital phenotype acquisition and analysis system (PDPAS) in S100 is an intelligent system integrating multi-source sensors, data processing algorithms, and environmental monitoring modules, aiming to accurately obtain the phenotype characteristics of plants at different growth stages, and dynamically analyze them combined with environmental factors. The core functions of PDPAS include real-time collection of multi-spectral radiation data above the plant canopy, leaf oscillation frequency, and root electrical signal topology. It provides basic data support for subsequent digital diversity analysis.

[0068] Please refer to Figure 2 , which shows a flowchart of an exemplary plant phenotype digital diversity index analysis method S100 of the present application, the content of which includes:

[0069] S110: Multi-spectral radiation data collection.

[0070] In terms of multispectral radiation data acquisition, PDPAS utilizes multispectral imaging technology by mounting multiple narrowband wavelength sensors to obtain the reflectance information of plant canopy in the visible to near-infrared band. Multispectral radiation data can reflect key physiological parameters such as chlorophyll content, water status, and photosynthetic efficiency of plant leaves.

[0071] S120: Leaf oscillation frequency measurement.

[0072] Leaf oscillation frequency acquisition relies on laser ranging technology and acceleration sensors. When plants are subjected to wind or mechanical disturbance, leaves will produce specific vibration patterns, and changes in vibration frequency can reflect the physical properties of plant tissues such as elastic modulus, cell wall strength, and stomatal opening. PDPAS captures leaf motion trajectories through high-speed cameras and inertial measurement units, and analyzes vibration frequency spectrum using Fourier transform. By analyzing the main vibration frequency, the structural stability of plants and their adaptability to environmental changes can be identified.

[0073] S130: Root electrical signal topology analysis.

[0074] Root electrical signal topology acquisition involves electrophysiological measurement technology, including rhizosphere impedance detection and ion flow analysis. PDPAS uses microelectrode arrays implanted in the soil to measure changes in electrical conductivity around the roots to reflect the ability of roots to absorb water and nutrients. In addition, PDPAS records the action potential of roots under different environmental conditions, which is the basic unit of plant electrical signal conduction.

[0075] In summary, S110-S130, PDPAS integrates multispectral imaging, leaf vibration analysis, and root electrophysiological measurement to achieve multi-dimensional perception of plant phenotypic characteristics.

[0076] S140: Constructing environment-phenotype correlation matrix.

[0077] Due to the different time resolution, spatial resolution, and physical dimension of multispectral radiation data, leaf oscillation frequency, and root electrical signal collected by different sensors, data alignment and standardization are performed to ensure data consistency and comparability. Dimensionality reduction is also performed to extract key features and improve the efficiency of subsequent analysis.

[0078] After completing data alignment, standardization, and dimensionality reduction, the environment-phenotype correlation matrix can be constructed. The environment-phenotype correlation matrix integrates environmental variables (temperature, humidity, light intensity, soil conductivity) and plant phenotypic characteristics (multispectral radiation data, leaf oscillation frequency, root impedance) into a correlation matrix, which is used to represent the response trajectory of plants under different environmental conditions. The form of this matrix can be represented as:

[0079] ;

[0080] wherein, represents an environment-phenotype association matrix, represents the measured value of the th sample on the th environment-phenotype feature, is the number of samples, is the feature dimension.

[0081] In an index analysis method of plant phenotype digital diversity, S200 extracts genotype and environment parameters from a historical database, screens influencing factors, constructs an ellipsoid model, and finally converts to a probability space.

[0082] Please refer to Figure 3 , which shows a flowchart of an example of an index analysis method S200 of plant phenotype digital diversity of the present application, the content of which includes:

[0083] S210: Extract genotype and environment stress parameters.

[0084] In plant phenotype digital diversity analysis, the extraction of genotype and environment stress parameters is the basis for constructing a three-dimensional ellipsoid domain model. S210 screens the genotype information closely related to plant phenotype response from the historical variety database, and extracts the key environment stress parameters in combination with the environment variables in EPDM.

[0085] The extraction of genotype information depends on the support of the historical variety database. The historical variety database usually contains the genomic sequences, single nucleotide polymorphism data and known phenotype-genotype association information of different plant varieties. In a possible implementation, PDPAS uses BLAST algorithm to align the genomic sequence of the target plant variety to identify the genetic loci related to specific phenotypes. Based on the above implementation, in an alternative implementation, PDPAS uses principal component analysis to reduce the dimension of genotype data to reduce redundant information and highlight major genetic variation.

[0086] The extraction of environment stress parameters is based on the environment variables in EPDM. EPDM contains various environmental factors such as temperature, humidity, light intensity, soil conductivity, and environmental variables constitute the environmental background of plant growth. In order to improve the comparability of data, in a possible implementation, PDPAS uses Z-score normalization method to standardize the environment variables to eliminate the dimensional difference between different variables.

[0087] S220: Construct a three-dimensional ellipsoid domain model based on genotype and environment stress parameters.

[0088] A three-dimensional ellipsoidal model was constructed based on genotype and environmental stress parameters to quantify the environmental stress gradient and the strength of genotype-environment interactions. The three-dimensional ellipsoidal model maps the joint distribution of genotype traits and environmental stress parameters onto a continuous ellipsoidal space, enabling a geometric visualization of plant phenotypic responses under different environmental conditions. The principal axis of the three-dimensional ellipsoidal model reflects the gradient of environmental stress, while the curvature of the ellipsoid characterizes the strength of the interaction between genotype and environmental factors.

[0089] In a possible implementation, the steps of constructing the three-dimensional ellipsoid domain model include:

[0090] S221: Construction of a three-dimensional ellipsoidal domain model based on the genotype data matrix and environmental stress parameter matrix The joint covariance matrix of .

[0091] The linear correlation between genotypic characteristics and environmental stress parameters is described by the mathematical expression:

[0092] ;

[0093] in, is the genotype data matrix, is the environmental stress parameter matrix, is the sample size, is the joint covariance matrix of genotype and environmental stress parameters. The eigenvalues ​​and eigenvectors of the ellipsoid are used to determine the direction of the principal axes of the ellipsoid and its corresponding scale. Specifically, the eigenvectors determine the direction of the principal axes of the ellipsoid, while the eigenvalues ​​determine the length of each principal axis, that is, the strength of the environmental stress gradient.

[0094] S222: The shape of the ellipsoid in the three-dimensional ellipsoidal domain model is determined by the ratio of the eigenvalues, reflecting the strength of the genotype-environment interaction.

[0095] set up are the three largest eigenvalues ​​of the covariance matrix, corresponding to the three main axes of the ellipsoid. The curvature characteristics of the ellipsoid are quantified by the ratio of the eigenvalues:

[0096] ;

[0097] in, represents the curvature characteristics of the ellipsoid, and are the eigenvalues ​​of any two principal axes.

[0098] Larger The value indicates that the degree of stretching of the ellipsoid in this direction is high, meaning that the environmental stress has a strong impact on the phenotypic response in this direction. Conversely, a smaller The value indicates that the curvature of the ellipsoid in this direction is low, meaning that the interaction between the genotype and the environmental factor is strong, resulting in a higher stability of the phenotypic response.

[0099] S223: The three-dimensional ellipsoid domain model maps the joint distribution of genotypes and environmental stress parameters to a continuous probability space through the ellipsoid equation. The standard equation of the ellipsoid is:

[0100] ;

[0101] wherein, is the genotype feature vector, is the environmental stress parameter vector, is the inverse matrix of the joint covariance matrix.

[0102] The equation of the three-dimensional ellipsoid domain model defines a three-dimensional ellipsoid domain, so that all points satisfying the equation are located on the surface of the ellipsoid, and the points inside the ellipsoid represent the phenotypic response mode with a higher probability. Through this model, the trend of plant phenotypic response under different environmental conditions can be intuitively analyzed, and the genetic advantage expression region and the diversity fluctuation risk site can be identified.

[0103] In an index analysis method of plant phenotypic digital diversity, S300 dynamically divides the non-uniform sub-domain of the environmental stress severe area and the stable area based on the principal axis gradient and curvature characteristics of the three-dimensional ellipsoid domain model, extracts the feature vector of the phenotypic response mode of each sub-domain and constructs an iteratively optimized phenotypic response fingerprint library, to realize the fine characterization and storage of the plant multi-environmental adaptability characteristics.

[0104] Please refer to Figure 4 , which shows a flowchart of an example of a plant phenotypic digital diversity index analysis method S300 of the present application, the content of which includes:

[0105] S310: Design a non-uniform sub-domain decomposition strategy.

[0106] The non-uniform sub-domain decomposition strategy is to adapt to the plant phenotypic response mode under different environmental stress intensities.

[0107] In one possible implementation, the non-uniform sub-domain decomposition strategy is to dynamically adjust the resolution of the sub-domain according to the principal axis direction and curvature characteristics of the ellipsoid domain, so that the environmental stress severe area adopts high-resolution subdivision, and the environmental stress stable area adopts low-resolution subdivision, thereby ensuring that each sub-domain can accurately capture specific phenotypic response characteristics.

[0108] In particular, the non-uniform sub-domain decomposition strategy relies on the principal axis directions and curvature information of the three-dimensional ellipsoidal domain model. The principal axis directions of the three-dimensional ellipsoidal domain model are determined by the eigenvectors of the joint covariance matrix , while the principal axis lengths are represented by the corresponding eigenvalues . The variation of environmental stress gradient is quantified by the gradient function in the principal axis direction, where is the eigenvector of the th principal axis direction. For each sample point , its projection in the principal axis direction is represented as:

[0109] ;

[0110] where denotes the projection value of the sample point in the th principal axis direction. By calculating the projection distribution of different sample points in the principal axis direction, the regions with large variation of environmental stress gradient are identified, and the decomposition density of the sub-domain is adjusted accordingly.

[0111] The curvature characteristics of the ellipsoid determine the strength of genotype-environment interaction, which in turn affects the resolution of the sub-domain. Based on the principal axis gradient and curvature information, PDPAS adopts an adaptive grid partitioning technique to dynamically adjust the decomposition density of the sub-domain. High-resolution subdivision is performed in regions with intense environmental stress, while low-resolution subdivision is performed in stable regions.

[0112] In one possible implementation, the partitioning step based on the non-uniform sub-domain decomposition strategy includes:

[0113] S311: Set the initial sub-domain partitioning in the three-dimensional ellipsoidal domain model space, usually using equal-interval partitioning to ensure coverage of the entire ellipsoidal domain.

[0114] S312: Calculate the gradient and curvature in the principal axis direction for each sub-domain to evaluate its environmental stress intensity and genotype-environment interaction level.

[0115] S313: For sub-domains with large gradients or high curvatures, use quadtree or octree algorithms for recursive subdivision to improve resolution; for sub-domains with small gradients or low curvatures, maintain lower resolution.

[0116] S314: After sub-domain refinement, adjust the boundaries of adjacent sub-domains to ensure data consistency and continuity, avoiding analysis errors caused by uneven partitioning.

[0117] S320: Extract phenotype response pattern features from each sub-domain.

[0118] Phenotypic response pattern features are extracted from each sub-domain to construct a highly discriminative phenotypic response fingerprint.

[0119] In one possible implementation, the step of extracting phenotypic response pattern features from each sub-domain comprises:

[0120] PDPAS employs principal component analysis to reduce the dimensionality of data within each sub-domain to extract the main phenotypic response features. Specifically, principal component analysis projects high-dimensional data into a lower-dimensional space by computing the eigenvectors of the data covariance matrix, making the new variables mutually independent and preserving the maximum variance information.

[0121] PDPAS combines time series analysis methods to dynamically model the phenotypic response patterns of each sub-domain. Specifically, since plant phenotypic responses have time dependence, PDPAS uses Fourier transform and wavelet transform to extract frequency domain features and time-frequency features of phenotypic responses, respectively.

[0122] PDPAS employs clustering analysis methods to identify patterns in reduced phenotypic response features. Specifically, the system uses the K-means clustering algorithm to classify the phenotypic response patterns of each sub-domain to identify similar response patterns and form a phenotypic response fingerprint.

[0123] In one possible implementation, the step of extracting phenotypic response pattern features from each sub-domain comprises: spectral oscillation ratio, vibration damping coefficient, and root system conductance entropy.

[0124] S330: Construct a phenotypic response fingerprint library.

[0125] The phenotypic response fingerprint library is used to store and manage phenotypic response pattern features within different sub-domains. The phenotypic response fingerprint library not only needs to have efficient data storage and retrieval capabilities, but also supports dynamic updates and pattern matching to quickly identify similar phenotypic response patterns in subsequent analysis and assist in genetic optimization and environmental adaptability assessment.

[0126] In one possible implementation, the step of constructing a phenotypic response fingerprint library comprises:

[0127] S331: The construction of PRFD is based on the encoding and indexing of feature vectors. Specifically, PDPAS uses autoencoder technology to encode the extracted phenotypic response features to generate compact and representative fingerprint vectors.

[0128] S332: PRFD employs approximate nearest neighbor search technology to achieve efficient pattern matching. Specifically, PDPAS introduces a locality-sensitive hashing algorithm to map feature vectors to a hash space, making similar feature vectors have a higher collision probability.

[0129] S333: PRFD adopts an incremental learning strategy to support dynamic updates to the database. Specifically, the system uses a mini-batch stochastic gradient descent algorithm to periodically update the parameters of the autoencoder, ensuring that the encoder can adapt to the latest phenotypic response features.

[0130] S334: PRFD employs knowledge graph technology to enhance the semantic association capabilities of the fingerprint library. Specifically, PDPAS combines phenotypic response patterns with known genotype-environment interaction relationships to construct a semantic network, where nodes represent phenotypic response patterns and edges represent similarities or causal relationships between patterns.

[0131] In one plant phenotypic digital diversity index analysis method, S400 tracks the unit of the plant digital phenotypic acquisition and analysis system by configuring the sub-domain-specific plant digital phenotypic acquisition and analysis system tracking unit, combines the cross-organ feedback mechanism of abnormal leaf oscillation triggering root impedance scanning, dynamically calculates the sub-domain response boundary, realizes the multi-dimensional physiological signal collaborative driving of environmental stress self-adaptive regulation and control, and significantly improves the dynamic analysis accuracy of plant phenotypic response and the environmental adaptability optimization capability.

[0132] Please refer to Figure 5 which shows a flowchart of one exemplary plant phenotypic digital diversity index analysis method S400 of the present application, the content of which includes:

[0133] S410: Configure the unit of the dedicated plant digital phenotypic acquisition and analysis system tracking unit.

[0134] The unit of the dedicated plant digital phenotypic acquisition and analysis system tracking unit is used to realize real-time monitoring and dynamic feedback of key physiological parameters of plants. The core functions of the tracking unit include: real-time monitoring of leaf oscillation frequency, triggering root impedance scanning, coordinating multi-source sensor data flow, and dynamically adjusting the sub-domain response boundary based on organ cascade feedback mechanism.

[0135] In one possible implementation, the tracking unit configuration includes:

[0136] The PDPAS tracking unit adopts a distributed sensor network architecture, with each sub-domain equipped with an independent multispectral imaging module, laser rangefinder, inertial measurement unit, and root conductivity sensor.

[0137] The PDPAS tracking unit integrates an edge computing module for real-time processing of sensor data and performing preliminary analysis.

[0138] The PDPAS tracking unit uses an event-driven mechanism to control the data acquisition frequency. Under normal conditions, the sensors operate at a lower frequency to reduce energy consumption; when abnormal leaf oscillation is detected, the system automatically switches to a high-frequency sampling mode and triggers root impedance scanning.

[0139] The PDPAS tracking unit is connected to the central control system through a wireless communication protocol, ensuring real-time data transmission and remote control.

[0140] S420: Establish a leaf oscillation anomaly detection mechanism and trigger root impedance scanning when an anomaly is detected.

[0141] After configuring the PDPAS tracking unit specific to the sub-domain, the next step is to establish a leaf oscillation anomaly detection mechanism and trigger root impedance scanning when an anomaly signal is detected. Changes in leaf oscillation frequency can reflect the structural stability and environmental adaptability of plants, providing physiological response information for the underground part of the plant.

[0142] In one possible implementation, the leaf oscillation anomaly detection mechanism is a threshold-based leaf oscillation anomaly detection. That is, a reference value is set, and when the leaf oscillation frequency exceeds the threshold, it is determined to be abnormal.

[0143] As an alternative to another embodiment, in order to improve the accuracy of anomaly detection, a machine learning classifier is used for anomaly detection.

[0144] When a leaf oscillation anomaly is detected, the PDPAS automatically triggers the root impedance scanning module to obtain the changes in root electrical signals. In one possible implementation, the root impedance scanning module uses the AC impedance spectroscopy technique to measure the dielectric properties of root tissues, which follows the basic principles of Ohm's law and the complex impedance model.

[0145] S430: Dynamically determine the sub-domain response boundary based on the organ cascade feedback mechanism.

[0146] Based on the organ cascade feedback mechanism, the sub-domain response boundary is dynamically determined to realize real-time correction of the stomatal conductance safety threshold.

[0147] The organ cascade feedback mechanism is to establish a cross-organ feedback network by integrating the physiological signals of leaves and roots, so that the system can dynamically adjust the response boundary of environmental stress according to the overall state of the plant.

[0148] Based on the organ cascade feedback mechanism, the response boundary of the sub-domain is dynamically adjusted to adapt to changes in plant phenotypes. The calculation of the dynamic response boundary is usually based on the principal axis direction and curvature characteristics of the ellipsoidal domain model, and is updated in combination with real-time feedback data. Its calculation formula is as follows:

[0149] ;

[0150] wherein, represents the updated sub-domain boundary, is the original boundary, is the learning rate, The gradient vector field of the ellipsoid surface. The formula can adjust the sub-domain boundary according to the real-time feedback data, so that it can more accurately reflect the phenotypic response pattern of plants.

[0151] In an index analysis method of plant phenotypic digital diversity, S500 generates a three-dimensional heat map by dynamic color mapping and spatial modeling technology based on a three-dimensional ellipsoid domain model, and realizes accurate labeling of risk sites and genetic advantage areas by combining statistical analysis and knowledge graph. The advantage is that the spatial heterogeneity of plant phenotypic response is revealed in an intuitive visual form, supporting real-time interactive adjustment and multi-source data fusion, and providing high-precision decision-making basis for subsequent stress tracking and environmental optimization.

[0152] PDPAS adopts a joint analysis method of ellipsoid domain model and phenotype response fingerprint library, maps multi-dimensional data to three-dimensional space, and reflects environmental stress intensity and genotype-environment interaction level through the color gradient of the heat map.

[0153] In a possible implementation, the three-dimensional heat map generation step is as shown in Figure 6 , including:

[0154] S510: Data mapping and visualization of three-dimensional heat map:

[0155] Specifically, the construction of the three-dimensional heat map depends on the principal axis direction and curvature information of the ellipsoid domain model, as well as the feature vectors in the phenotype response fingerprint library. First, PDPAS takes the principal axis direction of the ellipsoid domain model as the basis of the three-dimensional coordinate system, in which the three principal axes correspond to the three main directions of environmental stress gradient. Then, the system maps the sub-domain and its response boundary to the three-dimensional space, so that the phenotypic response pattern of each sub-domain can be visualized on the ellipsoid domain model.

[0156] S520: PDPAS adopts a color coding strategy, in which red represents areas with high environmental stress intensity, and blue represents stable areas with strong genotype-environment interaction.

[0157] Specifically, the red hot spot area corresponds to the high gradient area in the principal axis direction of the ellipsoid domain model, indicating that the phenotypic response in this area is significantly affected by environmental stress; while the blue stable area corresponds to the high curvature area of the ellipsoid domain model, indicating that the phenotypic response in this area is regulated by genotype-environment interaction, and has high stability.

[0158] In the plant phenotype digital diversity analysis, the three-dimensional heat map generated by S500 can intuitively show the environmental stress intensity and genotype-environment interaction level of different subdomains. Among them, the red hotspot area represents the area with higher environmental stress intensity, and the plant phenotype response mode in the red hotspot area is greatly affected by environmental factors, which may exist potential stress sources or phenotype response lag phenomenon. Therefore, S600 is to track the red high-risk area in depth to identify the hidden stress source and locate the phenotype response lag unit.

[0159] Please refer to Figure 7 which shows a flowchart of an exemplary plant phenotype digital diversity index analysis method S600 of the present application, the content of which includes:

[0160] S610: Crown vortex captures hidden stress sources.

[0161] The red hotspot area usually corresponds to the area with higher environmental stress gradient, and the source of environmental stress may not be directly visible. Therefore, the crown vortex capture technology is adopted, combined with high-precision meteorological sensors and infrared imaging equipment, to monitor the air flow pattern on the crown surface in real time.

[0162] The basic principle of crown vortex capture is to use laser radar and infrared thermal imaging technology to analyze the air flow trajectory on the plant canopy surface. When the plant is subjected to environmental stress, its stomatal opening and transpiration rate will change, thereby affecting the air flow pattern on the canopy surface. By comparing the crown vortex pattern under normal environmental conditions, PDPAS can identify abnormal air flow areas and further analyze the potential hidden stress sources in these areas.

[0163] S620: Locate the phenotype response lag unit.

[0164] In addition to identifying hidden stress sources, it is also necessary to locate the phenotype response lag unit. In some cases, the plant's phenotype response does not immediately appear, but there is a certain lag effect. For example, in the case of root system under the condition of soil nutrient deficiency or saline-alkali stress, the aboveground part of the plant may still remain normal in a short time, but the photosynthetic efficiency and structural stability of the leaves will gradually decrease as the stress accumulates.

[0165] In order to capture this lag effect, PDPAS uses time series analysis method to monitor the multi-spectral radiation data, leaf oscillation frequency and root electrical signal for a long time.

[0166] Specifically, PDPAS records the phenotype data of each sample at different time points, and uses dynamic time warping algorithm to compare the phenotype response patterns between different subdomains. When the phenotype response pattern of a certain subdomain deviates significantly from the expected model, the system will mark this area as a potential phenotype response lag unit.

[0167] S630: Data-driven stress source analysis.

[0168] To further improve the accuracy of deep tracking, PDPAS adopts a data-driven approach to analyze the stress sources in the red high-risk area. PDPAS integrates multi-source sensor data, including meteorological data, soil conductivity, leaf temperature, photosynthetic efficiency, etc., and uses multivariate regression analysis and principal component analysis methods to identify the main environmental factors affecting plant phenotypic responses.

[0169] For example, when analyzing drought stress, PDPAS focuses on the trends of soil moisture, transpiration rate, and leaf temperature. When soil moisture continues to decline while leaf temperature rises, the system determines that the area may be affected by water stress and automatically triggers the root impedance scanning module to obtain more detailed underground physiological information.

[0170] In the analysis of plant phenotypic digital diversity, the three-dimensional heat map generated by S500 not only identifies red hotspots with high environmental stress intensity, but also identifies blue stable zones with strong genotype-environment interaction. Blue stable zones represent areas where plant phenotypic responses are relatively stable, and plants in blue stable zones exhibit high genetic stability under specific environmental conditions, making them suitable for genetic optimization and environmental adaptability research. Therefore, S700 is based on the phenotypic characteristics of blue stable zones and the diagnosis results of lag units to reverse the optimal environmental parameter combination through inverse ellipsoid mapping and generate a variety-environment adaptation matrix to develop a diversity maintenance plan.

[0171] Please refer to Figure 8 , which shows a flowchart of an exemplary plant phenotypic digital diversity index analysis method S700 of the present application, the content of which includes:

[0172] S710: Reverse ellipsoid mapping to reverse the optimal environmental parameter combination.

[0173] The phenotypic response pattern of plants in blue stable zones is regulated by genotype-environment interaction, so PDPAS uses inverse ellipsoid mapping technology to reverse the optimal environmental parameter combination from the three-dimensional ellipsoid domain model. Inverse ellipsoid mapping uses the covariance matrix of the three-dimensional ellipsoid domain model to map the phenotypic response pattern back to the environmental parameter space, thereby determining the environmental conditions most conducive to the stable growth of plants.

[0174] In the specific implementation process, PDPAS first extracts the phenotypic feature vector of the blue stable zone from the three-dimensional ellipsoid domain model and calculates its corresponding covariance matrix. This covariance matrix reflects the correlation between genotype characteristics and environmental stress parameters, and PDPAS maps the phenotypic feature vector back to the environmental parameter space by solving the inverse matrix of this matrix. The formula for inverse ellipsoid mapping is as follows:

[0175] ;

[0176] where, represents the environmental parameter vector, represents the phenotypic feature vector, is the inverse matrix of the joint covariance matrix. Through this formula, PDPAS calculates the optimal environmental parameter combination, so that the phenotypic response of plants reaches the maximum stability under the environment.

[0177] S720: Generate variety-environment adaptation matrix.

[0178] After determining the optimal environmental parameter combination, PDPAS further generates a variety-environment adaptation matrix to evaluate the adaptability of different plant varieties under different environmental conditions. The construction of this matrix is based on the principal axis direction and curvature information of the three-dimensional ellipsoidal domain model, combined with the feature vectors in the phenotypic response fingerprint library.

[0179] The mathematical form of the variety-environment adaptation matrix is as follows:

[0180] ;

[0181] where, represents the adaptation degree of the th plant variety under the th environmental condition, the higher the value, the stronger the adaptability of the variety under the environment. The calculation of the adaptation degree is based on the environmental parameter combination obtained by the inverse ellipsoidal mapping, and combined with the feature vectors in the phenotypic response fingerprint library for weighted calculation.

[0182] S730: Develop a diversity maintenance plan.

[0183] After generating the variety-environment adaptation matrix, PDPAS further develops a diversity maintenance plan to ensure the long-term adaptability of plant populations under different environmental conditions. By optimizing the environmental parameter combination and variety selection, the stability of plant phenotypes is maintained, while promoting the development of genetic diversity.

[0184] First, PDPAS adopts a genetic optimization strategy based on the adaptation matrix to screen the most adaptable plant varieties, and combines genomics data to identify gene sites related to environmental adaptability.

[0185] Secondly, PDPAS adopts dynamic environmental regulation technology. According to the results of the adaptation matrix, the environmental parameters of the greenhouse or field are adjusted to ensure the growth of plants in the best environment. For example, in the blue stable zone, the system automatically adjusts the light intensity, temperature, humidity and other parameters to maintain the stable growth state of plants. In addition, PDPAS can also combine with intelligent irrigation and fertilization system, dynamically adjust water and fertilizer supply strategy according to the physiological needs of plants, to improve resource utilization efficiency.

[0186] Finally, PDPAS adopts knowledge graph-based decision support system, integrates plant phenotype, genotype, environmental data, and combines historical experimental data to provide scientific diversity maintenance suggestions. For example, PDPAS analyzes the plant growth trend under different environmental conditions and predicts the impact of future climate change on plant adaptability, so as to develop long-term genetic optimization and environmental regulation strategies.

[0187] Plant phenotype digital diversity aims to comprehensively analyze the plant phenotype response mode under different environmental conditions through high-precision sensors, data analysis and modeling methods. The core framework of this method consists of multiple key steps, covering data collection, modeling analysis, pattern recognition, dynamic feedback and optimization decision-making, forming a complete analysis system.

[0188] The whole plant phenotype digital diversity analysis framework realizes the comprehensive analysis and accurate control of plant phenotype response through multi-source data fusion, mathematical modeling, pattern recognition and intelligent optimization. This method not only improves the efficiency of plant scientific research, but also provides strong technical support for agricultural production and ecological protection.

[0189] The application also provides a computer device, which comprises a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to realize the index analysis method of plant phenotype digital diversity as one of the technical solutions.

[0190] The basic principles of the application are described above in combination with specific embodiments, but it should be noted that the advantages, advantages, effects and the like mentioned in the application are only examples and are not limited, and these advantages, advantages, effects and the like cannot be considered as each embodiment of the application must have. In addition, the above specific details are only for the purpose of example and for the purpose of understanding, and are not limited to the above specific details, which do not limit the application to the above specific details.

[0191] The block diagrams of the devices, apparatuses, and methods involved in the present application are only illustrative examples and are not intended to require or imply that the connection, arrangement, configuration must be as shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, and methods can be connected, arranged, and configured in any manner. Words such as "include," "contain," "have," and the like are open-ended words that mean "including but not limited to," and can be used interchangeably with each other. The words "or" and "and" used herein mean the word "and / or," and can be used interchangeably with each other, unless the context clearly indicates otherwise. The word "such as" used herein means the phrase "such as but not limited to," and can be used interchangeably with each other.

[0192] It should also be noted that in the devices, apparatuses, and methods of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalents of the present application.

[0193] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects without departing from the scope of the present application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0194] The above description is only the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for analyzing plant phenotypic digital diversity indicators, characterized in that: include: Plant phenotyping analysis uses a plant digital phenotyping collection and analysis system to build a heterogeneous monitoring network, collaboratively collecting multispectral radiation data above the plant canopy, leaf oscillation frequency, and root electrical signal topology to generate an environment-phenotype dynamic matrix; Based on the environment-phenotype dynamic matrix, combined with the historical variety database, genotype and environmental stress parameters were extracted to construct a three-dimensional ellipsoid domain model; Based on the three-dimensional ellipsoid domain model, non-uniform subdomain decomposition is implemented along the main axis direction: high-resolution subdivision is performed in areas with severe environmental stress, and low-resolution subdivision is performed in stable areas, so that each subdomain corresponds to a unique phenotypic response pattern feature, forming a phenotypic response fingerprint library; For each subdomain, a tracking unit of the plant digital phenotyping acquisition and analysis system is configured. When abnormal leaf oscillation is detected, a root impedance scan is triggered, and the subdomain response boundary is dynamically determined based on the organ cascade feedback mechanism. Based on the three-dimensional ellipsoid domain model, the subdomains and the corresponding subdomain response boundaries are marked to generate a three-dimensional heat map; In response to the three-dimensional heat map results, deep tracking is initiated for the red high-risk areas. The deep tracking method is to control the plant digital phenotyping acquisition and analysis system to capture hidden stress sources along the canopy vorticity and locate the phenotypic response lag units; Based on the phenotypic characteristics of the blue stable zone and the diagnostic results of the hysteresis unit, the optimal environmental parameter combination is inferred through inverse ellipsoid mapping, the variety-environment adaptation matrix is ​​generated, and the diversity maintenance plan is output.

2. The method for analyzing plant phenotypic digital diversity according to claim 1, characterized in that: The multispectral radiation data acquisition uses a multispectral imaging system equipped with multiple narrow-band wavelength sensors to obtain plant canopy reflectance information in the visible to near-infrared band, and realizes digital representation of key physiological parameters through modeling of chlorophyll content, water status and photosynthetic efficiency; The blade oscillation frequency measurement uses a high-speed camera and an inertial measurement unit to capture the blade motion trajectory, and analyzes the main vibration spectrum through Fourier transform; The root electrical signal topology analysis includes: using a microelectrode array implanted in the soil to measure changes in rhizosphere conductivity, combining ion flow analysis with action potential recording to evaluate the root system's ability to absorb water and nutrients.

3. The method for analyzing plant phenotypic digital diversity according to claim 1, wherein: The construction of the environment-phenotype dynamic matrix includes time alignment, spatial standardization and dimensionality reduction of multi-source data, wherein the dimensionality reduction method uses principal component analysis or wavelet transform to extract key phenotypic features; The environment-phenotype dynamic matrix contains the joint distribution of environmental variables and phenotypic characteristics, which is expressed as: ； in, represents the environment-phenotype dynamic matrix, Indicates the The sample in measurements of environmental-phenotypic traits, is the sample size, is the characteristic dimension.

4. The method for analyzing plant phenotypic digital diversity according to claim 1, wherein: The construction of the three-dimensional ellipsoid domain model includes: Based on the joint covariance matrix of the genotype data matrix and the environmental stress parameter matrix, the eigenvalues ​​and eigenvectors were calculated to determine the direction and scale of the ellipsoid's major axis. The curvature of the ellipsoid is quantified by the ratio of the eigenvalues ​​of any two principal axes and is used to characterize the strength of the genotype-environment interaction; The standard equation of the three-dimensional ellipsoid domain model is: ； in, is the genotype feature vector, is the environmental stress parameter vector, is the inverse matrix of the joint covariance matrix; Among them, the three-dimensional ellipsoid domain model distinguishes different types of phenotypic response patterns through the ellipsoid curvature and principal axis gradient. The ellipsoid stretching direction represents the area dominated by environmental stress, while the high curvature direction reflects the area with strong genetic stability.

5. The method for analyzing plant phenotypic digital diversity according to claim 1, characterized in that: The non-uniform subdomain decomposition strategy is based on the ellipsoid principal axis gradient function and local curvature evaluation, and uses an adaptive meshing algorithm to dynamically adjust the subdomain resolution; Implementing quadtree / octree recursive subdivision in high-gradient regions to achieve refined modeling and classification of stress response patterns; The non-uniform subdomain division includes: calculating the gradient and curvature in the main axis direction for each subdomain, identifying areas with severe environmental stress and improving the resolution, while retaining the low-resolution division of stable areas to balance computational efficiency and modeling accuracy.

6. The method for analyzing plant phenotypic digital diversity indicators according to claim 1, characterized in that: The phenotypic response fingerprint library construction includes: PDPAS uses autoencoder technology to encode the extracted phenotypic response features to generate compact and representative fingerprint vectors; PDPAS introduces a locality-sensitive hashing algorithm to map feature vectors into a hash space, so that similar feature vectors have a higher probability of collision. PRFD uses a mini-batch stochastic gradient descent algorithm to periodically update the parameters of the autoencoder to ensure that the encoder can adapt to the latest phenotypic response characteristics; PDPAS combines phenotypic response patterns with known genotype-environment interactions to construct a semantic network, where nodes represent phenotypic response patterns and edges represent similarities or causal relationships between patterns. The key features stored in the phenotypic response fingerprint library include: spectral oscillation ratio, vibration damping coefficient, and root conductance entropy.

7. The method for analyzing plant phenotypic digital diversity indicators according to claim 1, characterized in that: The organ cascade feedback mechanism integrates the real-time physiological signals of leaf oscillation and root impedance to construct a cross-organ coupling feedback model, and dynamically adjusts the subdomain response boundary based on the ellipsoid surface gradient field.

8. The method for analyzing plant phenotypic digital diversity indicators according to claim 1, characterized in that: The three-dimensional heat map generation includes: using the three main axes of the ellipsoid domain model as a three-dimensional coordinate system, mapping each subdomain and its response boundary to the space, and using a color coding strategy, with red hot spots marking diversity fluctuation risk sites and blue stable areas marking genetic advantage expression sites; The three-dimensional heat map introduces a knowledge graph to enhance the semantic annotation function, allowing users to interactively query the genotype background, historical response trends and potential stress types of the subdomain.

9. The method for analyzing plant phenotypic digital diversity indicators according to claim 1, characterized in that: Deep tracking of hidden stress sources includes: using lidar and infrared thermal imaging technology to capture the trajectory of canopy vorticity changes, combining meteorological sensor data to identify areas of abnormal air flow, and thus locate potential hidden stress sources; The positioning phenotypic response hysteresis unit uses a dynamic time warping algorithm to compare the phenotypic response patterns between different subdomains. When the response pattern of a subdomain deviates significantly from the expected one, it is marked as a potential phenotypic response hysteresis unit. The hidden stress source analysis includes: integrating multi-source sensor data, using multiple regression analysis and principal component analysis to identify the main environmental factors affecting plant phenotypic responses, and automatically triggering the root impedance scanning module to obtain more detailed underground physiological information.

10. The method for analyzing plant phenotypic digital diversity indicators according to claim 1, characterized in that: The inverse ellipsoid mapping inversely estimates the optimal environmental parameter combination, including: environmental parameter vector , phenotypic feature vector , the inverse matrix of the joint covariance matrix , the formula for inverse ellipsoid mapping is as follows: ； The variety-environment adaptation matrix is ​​constructed by evaluating the adaptability of different plant varieties under different environmental conditions based on the inverse ellipsoid mapping results. The higher the adaptability value, the more adaptable the variety is in the environment.

Images