A component detection method and system for apple tree special-purpose fertilizer

Abstract

The present application relates to the technical field of fertilizer component detection, in particular to a component detection method and system for special fertilizer for apple trees, comprising: sampling and multi-spectral detection of a package fertilizer for apple trees, and obtaining the theoretical component proportion of each stage fertilizer in the package fertilizer; obtaining the core component in the package fertilizer, combining the theoretical component proportion of the core component to reduce the dimension of the spectral data, using the dimension reduction result to cluster the samples, thereby calculating the uniformity coefficient of the stage fertilizer; using the distribution of the dimension reduction result and the difference between the actual proportion and the theoretical component proportion of the core component in the sample, calculating the formula accuracy of the stage fertilizer to which the sample belongs; combining the difference between the formula accuracy and the uniformity coefficient of the stage fertilizer in the adjacent period, obtaining the component skewness of the package fertilizer; using the component skewness to obtain the component detection result of the package fertilizer. The present application improves the detection effect of the core component in the package fertilizer for apple trees.

Description

Technical Field

[0001] This invention relates to the field of fertilizer component detection technology, specifically to a method and system for detecting the components of fertilizers specifically for apple trees. Background Technology

[0002] Apple trees have varying nutrient requirements at different growth stages. If the actual composition of the fertilizer does not match the required ratio, it can lead to malnutrition in the trees and decreased fruit quality. Therefore, to meet the different nutrient needs of apple trees at different growth stages, the industry has developed and produced specialized apple tree fertilizer packages. However, in actual production, the composition ratios of these packages sometimes deviate from production requirements. Therefore, it is necessary to test the composition of apple tree-specific fertilizers to allow for timely adjustments to the fertilizer composition ratios during production, thus avoiding insufficient nutrient supply to the apple trees due to imbalanced composition ratios caused by inadequate product quality.

[0003] In existing fertilizer component testing methods, multispectral technology is usually used to detect fertilizer components. However, the testing process is still carried out using the conventional testing methods for general-purpose fertilizers. This fails to fully reflect the heterogeneity of apple tree-specific fertilizer packages compared to general-purpose fertilizers in terms of the proportion of some components. As a result, conventional component testing methods are not specific enough for apple tree-specific fertilizer package component testing and cannot accurately and effectively detect whether there are deviations in the proportion of components in apple tree-specific fertilizers from production requirements. Summary of the Invention

[0004] This invention provides a method and system for detecting the components of fertilizers specifically for apple trees, in order to solve existing problems.

[0005] The present invention provides a method and system for detecting the components of a fertilizer specifically for apple trees, which adopts the following technical solution: One embodiment of the present invention provides a method for detecting the components of a fertilizer specifically for apple trees, the method comprising the following steps: Obtain a fertilizer package for apple trees, and perform sampling and multispectral detection to obtain spectral data of the samples; the fertilizer package contains several stage fertilizers for different stages of apple tree growth, with several samples corresponding to each stage fertilizer; obtain the theoretical component ratios of the components in each stage fertilizer of the fertilizer package. The core components of the fertilizer package are obtained, and the spectral data corresponding to the sample of the fertilizer stage are reduced in dimension by combining the theoretical component ratio of the core components in the fertilizer stage. The dimensionality reduction results are used to cluster all samples of the fertilizer stage, and the uniformity coefficient of the fertilizer stage is calculated based on the distribution of samples in the clustering results. By utilizing the distribution of the dimensionality reduction results and the difference between the actual and theoretical proportions of the core components in the sample, the formulation accuracy of the fertilizer of the sample's stage is calculated; by combining the differences in formulation accuracy and uniformity coefficient of the fertilizer of the next stage in adjacent periods, the component skewness of the package fertilizer is obtained. The component analysis results of the fertilizer package were obtained by utilizing component bias.

[0006] Optionally, the specific method for obtaining the core components of the fertilizer package, performing dimensionality reduction on the spectral data corresponding to the samples of the stage fertilizer based on the theoretical component ratio of the core components in any stage fertilizer, clustering all samples of the stage fertilizer using the dimensionality reduction result, and calculating the uniformity coefficient of the stage fertilizer based on the distribution of samples in the clustering result is as follows: Nitrogen, phosphorus, and potassium are collectively referred to as core components. The spectral data segments corresponding to the core components in the spectral data of any sample of fertilizer at any stage are obtained and principal component analysis is performed. In the process of principal component analysis, the theoretical component ratios between nitrogen, phosphorus, and potassium of the fertilizer at the stage are combined as weights to obtain the core component vector of the sample. Clustering is performed using the core vector group of all samples of the stage fertilizer, and the uniformity coefficient of the stage fertilizer is calculated based on the distribution of samples in the clustering results.

[0007] Optionally, the specific method for obtaining the spectral data segments corresponding to the core components contained in the spectral data of any sample of fertilizer at any stage and performing principal component analysis, combining the theoretical component ratios among nitrogen, phosphorus, and potassium of the fertilizer at any stage as weights during the principal component analysis to obtain the core component vector of the sample, includes: Obtain the corresponding spectral band intervals for nitrogen, phosphorus, and potassium in the spectral data, denoted as the nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval, respectively. For any stage fertilizer, obtain the data segments corresponding to the nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval in the spectral data of any sample of the stage fertilizer, denoted as the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, respectively. Combine the theoretical component ratio of the fertilizer package to which the sample belongs, and the covariance matrix obtained during the principal component analysis of the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, to perform dimensionality reduction and obtain the core component vector of the sample.

[0008] Optionally, the step of combining the theoretical component ratios of the fertilizer package to which the sample belongs, and the covariance matrix obtained during principal component analysis of the nitrogen, phosphorus, and potassium spectral data segments of the sample to perform dimensionality reduction and obtain the core component vector of the sample, includes the following specific methods: Based on the theoretical component ratios of the core components in the fertilizer at the sample's stage, the proportional weights of the core components are obtained. Principal component analysis is used to obtain the covariance matrices of the nitrogen, phosphorus, and potassium spectral data segments, denoted as the nitrogen covariance matrix, phosphorus covariance matrix, and potassium covariance matrix, respectively. The proportional weight of nitrogen is multiplied by the nitrogen covariance matrix to obtain the nitrogen weighted covariance matrix. Similarly, the weighted covariance matrix of each core component is obtained. All the weighted covariance matrices of the core components are processed using principal component analysis to obtain the eigenvectors of the corresponding core components. The eigenvectors of all the core components of the sample are then grouped into an array, denoted as the core vector group of the sample.

[0009] Optionally, the specific method for obtaining the proportion weight of the core component based on the theoretical proportion of the core component in the sample's stage of fertilizer is as follows: Obtain the corresponding proportion values ​​of nitrogen, phosphorus, and potassium in the theoretical component ratio, and record them as nitrogen ratio value, phosphorus ratio value, and potassium ratio value respectively. The nitrogen ratio value, phosphorus ratio value, and potassium ratio value are collectively referred to as component ratio value. The ratio between the component ratio value corresponding to any core component and the sum of the component ratio values ​​of all core components is used as the ratio weight of the core component.

[0010] Optionally, the specific method for clustering the core vector groups of all samples of the stage fertilizer and calculating the uniformity coefficient of the stage fertilizer based on the distribution of samples in the clustering results includes: Obtain the core vector set of all samples at any stage of fertilizer growth; set the parameters of the K-means clustering algorithm. The uniformity coefficient of the stage fertilizer is calculated based on the Euclidean distance between the core vector groups and the K-means clustering algorithm. The cluster center of the cluster is recorded as the first cluster center. The sample that is far from the first cluster center is recorded as the second cluster center. The uniformity coefficient of the stage fertilizer is calculated based on the distance difference between the sample in the cluster and the first cluster center and the second cluster center.

[0011] Optionally, the method for calculating the uniformity coefficient of the stage fertilizer based on the distance differences between samples in the cluster and the centers of the first and second clusters is as follows: Obtain the Euclidean distances between any sample in the cluster and the centers of the first and second clusters, and denot them as the first distance and the second distance, respectively. Use the absolute value of the difference between the first distance and the second distance as the deviation distance of the sample. Divide the deviation distance of the sample by the maximum value of the first distance and the second distance, and use the result as the deviation parameter of the sample. Obtain the average value of the deviation parameters of all samples in the stage fertilizer to which the sample belongs, and use it as the uniformity coefficient of the stage fertilizer.

[0012] Optionally, the method for calculating the formulation accuracy of the fertilizer to which the sample belongs by utilizing the distribution of the dimensionality reduction results and the difference between the actual and theoretical proportions of the core components in the sample; and combining the differences in formulation accuracy and uniformity coefficients of fertilizers in adjacent periods to obtain the component skewness of the fertilizer package, includes the following specific methods: By utilizing the eigenvalues ​​of the spectral data segments of the core components corresponding to the sample of fertilizer at any stage during the dimensionality reduction process, and establishing a rectangular coordinate system, the uniform influence coefficient of the fertilizer at that stage can be obtained based on the discreteness of all core components of the sample in the rectangular coordinate system. The content of core components is obtained from the spectral data of any sample. The actual component ratio of the core components in the sample is obtained by the ratio of the contents of the core components. The deviation factor of the core components is calculated based on the difference between the actual component ratio and the theoretical component ratio of any core component in the sample. The average value of the deviation factors of all core components in all samples of the stage fertilizer is recorded as the fertilizer formulation accuracy of the stage fertilizer. The quality coefficient of the stage fertilizer is obtained based on the uniformity coefficient, uniformity influence coefficient, and fertilizer formulation accuracy; the difference in quality coefficients between stage fertilizers used for adjacent periods of apple trees in the package fertilizer is obtained to obtain the component skewness of the package fertilizer.

[0013] Optionally, the uniformity influence coefficient of the stage fertilizer includes the following specific methods: The eigenvalues ​​of the spectral data segments corresponding to all core components of any sample are obtained during principal component analysis. The array formed by these eigenvalues ​​is denoted as the core feature array of the sample. A Cartesian coordinate system is constructed with the same number of dimensions as the number of core components. The eigenvalues ​​of each core component are used as the coordinate axes of the Cartesian coordinate system. A scatter plot of all samples of any stage fertilizer based on the corresponding core feature array is obtained in the Cartesian coordinate system. The standard deviation of the coordinates of all samples in the scatter plot is obtained and denoted as the component dispersion factor of the stage fertilizer. Based on the component dispersion factor, the uniform influence coefficient of the stage fertilizer is obtained. The component dispersion factor and the uniform influence coefficient are negatively correlated.

[0014] A component detection system for a fertilizer specifically for apple trees includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of any one of the component detection methods for a fertilizer specifically for apple trees.

[0015] The beneficial effects of the technical solution of this invention are as follows: By combining multispectral technology with an innovative quantitative model, the uniformity of the distribution of core components of fertilizers at each stage, the accuracy of the formulation, and the continuity of nutrient supply between stages can be accurately evaluated. Its overall beneficial effects are: significantly improving the scientific nature and reliability of fertilizer production quality control, ensuring that the nutrient supply of base fertilizer, budding fertilizer, and fruit-expanding fertilizer at their respective growth stages strictly matches the physiological needs of apple trees; effectively eliminating uneven fruit quality caused by local fluctuations in nutrient levels, such as abnormal sugar content distribution or decreased fruit set rate; and simultaneously, by quantifying the quality connection relationship between fertilizers at adjacent stages, ensuring a smooth transition and stability of nutrient supply throughout the apple tree's life cycle, avoiding growth inhibition caused by nutrient gaps or excesses, and overall improving the detection effect of core components in apple tree fertilizer packages. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the steps of a method for detecting the components of a fertilizer specifically for apple trees according to the present invention. Figure 2 This is a structural block diagram of a component detection system for apple tree-specific fertilizers according to the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method and system for detecting the components of a fertilizer specifically for apple trees proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details a specific scheme for a method and system for detecting the components of a fertilizer specifically for apple trees provided by this invention.

[0021] Please see Figure 1The diagram illustrates a flowchart of a method for detecting the components of a fertilizer specifically for apple trees, according to an embodiment of the present invention. The method includes the following steps: Step S001: Obtain the package fertilizer for apple trees, and perform sampling and multispectral detection to obtain the spectral data of the samples; the package fertilizer contains several stage fertilizers for different stages of apple trees, and each stage fertilizer corresponds to several samples; obtain the theoretical component ratio of each stage fertilizer in the package fertilizer.

[0022] It should be noted that, referring to the invention "A Special Fertilizer Package for Apple Trees," publication number CN115093293A, it can be seen that existing technology has developed a fertilizer package specifically for apple trees. This fertilizer package includes fertilizer formulas for apple trees at different growth stages, thereby utilizing multiple fertilizer formulas to match the nutrient requirements of apple trees at different growth cycles, in order to further increase yield and improve quality. In order to accurately detect the components of this fertilizer package in this embodiment of the invention, and to ensure that the component ratios during the production process conform as closely as possible to the theoretical requirements, so that the fertilizer package can maximize its ability to meet the nutrient needs of apple trees at different growth stages, it is first necessary to collect samples of the fertilizer package produced using this invention and perform multispectral analysis.

[0023] Specifically, in order to implement the component detection method for apple tree-specific fertilizers proposed in this embodiment, it is first necessary to collect samples of different fertilizers in the fertilizer package and their multispectral data. The specific process is as follows: First, optional crushing equipment and multispectral detection equipment are installed.

[0024] As an optional example, the specific method for selecting the pulverizing equipment and multispectral detection equipment includes: using a Spectral Devices multispectral imaging system with a spectral range of 400-900 nm; and simultaneously using a Retsch ZM 200 high-speed rotary pulverizer to pulverize and grind the collected sample to facilitate subsequent multispectral detection.

[0025] Then, the fertilizers used for different stages of apple trees in the package fertilizer are collectively referred to as stage fertilizers, and samples are collected for each stage fertilizer to obtain several samples corresponding to each stage fertilizer; the stage fertilizers include base fertilizer, budding fertilizer and fruit expansion fertilizer.

[0026] As an optional embodiment, the fertilizers used for different stages of apple tree growth in the fertilizer package are collectively referred to as stage fertilizers, and samples are collected from each stage fertilizer to obtain several samples corresponding to each stage fertilizer. The specific method includes: For any stage of fertilizer, a number of samples are collected from the fertilizer pile formed by the stage fertilizer using a five-point sampling method, with each sample having the same weight.

[0027] Finally, all samples were pulverized and ground using a high-speed rotary pulverizer. After pulverization and grinding, the samples were subjected to multispectral detection using a multispectral imaging system to obtain the corresponding spectral data.

[0028] In addition, the theoretical component ratios of each stage of the fertilizer package were obtained.

[0029] Thus, the spectral data of samples corresponding to each stage of the apple tree fertilizer package and the theoretical component ratios of the components in the stage fertilizer were obtained through the above method.

[0030] Step S002: Obtain the core components of the fertilizer package, reduce the dimensionality of the spectral data corresponding to the sample of the fertilizer stage by combining the theoretical component ratio of the core components in any stage fertilizer, cluster all samples of the fertilizer stage using the dimensionality reduction result, and calculate the uniformity coefficient of the fertilizer stage based on the distribution of samples in the clustering result.

[0031] It should be noted that because the apple tree fertilizer package contains various fertilizers in different proportions, i.e., staged fertilizers, and because the fertilizer production process may result in slight differences in the proportions of components at different locations within the fertilizer, this will also affect the component analysis of the sampled fertilizer. It may even affect the nutrient uptake of the apple tree at the targeted stage of the fertilizer application. Therefore, this embodiment of the invention uses multispectral technology to analyze different samples of each stage of the fertilizer to obtain corresponding spectral data, in order to analyze the uniformity of the components of the fertilizer at each stage. Furthermore, because the requirements of apple trees for nitrogen, phosphorus, and potassium vary significantly at different growth stages, and the nutrients in fertilizers at different stages have a continuity—that is, the nutrient residue of the previous stage fertilizer needs to be compatible with the nutrient requirements of the next growth stage—this embodiment of the invention further combines the component proportions of each stage fertilizer with the nutrient requirements of different growth stages to conduct a linkage evaluation between the components of the stage fertilizer corresponding to different stages of the apple tree and the growth stage.

[0032] It should be further explained that, for the spectral data of fertilizer at any stage, since the spectral data is acquired within a certain wavelength range during the acquisition process, it contains the spectral information of all components in the fertilizer at that stage. It includes not only the spectral information of the core components required for apple tree growth, but also the spectral information of other non-core components. Therefore, in order to perform spectral analysis on the core components required for apple tree growth by the package fertilizer, this embodiment of the invention selects to perform wavelength filtering and dimensionality reduction on the spectral data, so as to use the dimensionality-reduced data to further analyze the uniformity of the distribution of all samples of the fertilizer at the stage on the core components.

[0033] As a preferred embodiment, the specific method for obtaining the core components of the fertilizer package, performing dimensionality reduction on the spectral data corresponding to the samples of the stage fertilizer based on the theoretical component ratio of the core components in any stage fertilizer, clustering all samples of the stage fertilizer using the dimensionality reduction result, and calculating the uniformity coefficient of the stage fertilizer based on the distribution of samples in the clustering result is as follows: First, nitrogen, phosphorus, and potassium are collectively referred to as core components. The spectral data segments corresponding to the core components in the spectral data of any sample of fertilizer at any stage are obtained and principal component analysis is performed. In the process of principal component analysis, the theoretical component ratios between nitrogen, phosphorus, and potassium of the fertilizer at the stage are used as weights to obtain the core component vector of the sample.

[0034] As an optional embodiment, the step of obtaining the spectral data segments corresponding to the core components contained in the spectral data of any sample of fertilizer at any stage and performing principal component analysis, and combining the theoretical component ratios of nitrogen, phosphorus, and potassium in the fertilizer at any stage as weights to obtain the core component vector of the sample, includes the following specific calculation method: obtaining the band intervals corresponding to nitrogen, phosphorus, and potassium in the spectral data, denoted as nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval, respectively; for any fertilizer at any stage, obtaining the data segments corresponding to the nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval in the spectral data of any sample of the fertilizer at that stage, denoted as the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, respectively; combining the theoretical component ratios of the fertilizer package to which the sample belongs, and the covariance matrix obtained during the principal component analysis of the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, thereby performing dimensionality reduction to obtain the core component vector of the sample.

[0035] As an optional embodiment, the method of combining the theoretical component ratio of the fertilizer package to which the sample belongs, and the covariance matrix obtained during principal component analysis of the nitrogen, phosphorus, and potassium spectral data segments of the sample to perform dimensionality reduction and obtain the core component vector of the sample includes the following specific methods: obtaining the proportion weight of the core component based on the theoretical component ratio of the core component in the fertilizer package to which the sample belongs; obtaining the covariance matrices of the nitrogen, phosphorus, and potassium spectral data segments through principal component analysis, denoted as the nitrogen covariance matrix, phosphorus covariance matrix, and potassium covariance matrix, respectively; multiplying the nitrogen proportion weight by the nitrogen covariance matrix to obtain the nitrogen weighted covariance matrix; similarly, obtaining the weighted covariance matrix of each core component; processing the weighted covariance matrices of all core components through principal component analysis to obtain the feature vectors of the corresponding core components; forming an array of the feature vectors of all core components of the sample, denoted as the core vector group of the sample.

[0036] As an optional embodiment, the method for obtaining the proportion weight of the core component based on the theoretical component ratio of the core component in the fertilizer of the sample stage includes: obtaining the proportion values ​​of nitrogen, phosphorus and potassium in the theoretical component ratio, and recording them as nitrogen proportion value, phosphorus proportion value and potassium proportion value respectively; collectively referring to the nitrogen proportion value, phosphorus proportion value and potassium proportion value as component proportion value; and taking the ratio between the component proportion value corresponding to any core component and the sum of the component proportion values ​​of all core components as the proportion weight of the core component.

[0037] It should be noted that in apple tree-specific fertilizer packages, nitrogen is a fundamental nutrient for budding and growth, phosphorus promotes fruit formation and prevents fruit drop, and potassium promotes fruit enlargement and sweetness. Therefore, in this embodiment of the invention, nitrogen, phosphorus, and potassium are used as core components. Subsequent analysis of these core components ensures their proportions remain within theoretical ranges, effectively meeting the nutrient requirements of apples at different growth stages. Furthermore, since the spectral data segments corresponding to nitrogen, phosphorus, and potassium still contain considerable redundant information, principal component analysis is used to reduce the dimensionality of the spectral data segments to facilitate subsequent analysis of the key information in the spectra of nitrogen, phosphorus, and potassium.

[0038] It should also be noted that, since nitrogen, phosphorus, and potassium are the core components of the apple tree fertilizer package, and in order to analyze the proportion of the core components during the component testing process in this embodiment of the invention, the content ratio of the core components in fertilizers at different stages is directly obtained from the invention "A Special Fertilizer Package for Apple Trees" with publication number CN115093293A, as the theoretical component ratio between the core components. For example, the content ratio of nitrogen, phosphorus, and potassium in the base fertilizer is 8:4:3, with a total ratio of 15. Therefore, the specific content ratio (i.e., the proportional weight of the core components) is: nitrogen 8 / 15, phosphorus 4 / 15, and potassium 3 / 15. Thus, in this embodiment of the invention, the proportional weight is set based on the differentiated needs of apple trees at different growth stages for the core components (nitrogen, phosphorus, and potassium) to characterize the contribution ratio of each component to the growth function at a specific stage. For example, nitrogen demand is high during the basal fertilizer stage (8 / 15 of the theoretical ratio of 8:4:3), as it dominates cell division and leaf growth during budding; while potassium demand increases during the fruit expansion stage, as it regulates sugar transport and fruit enlargement. Therefore, by weighting the principal component analysis process with proportional weights, the dimensionality reduction can be focused on the spectral response of key nutrients, significantly improving the targeting of core component detection.

[0039] Then, clustering is performed using the core vector group of all samples of the stage fertilizer, and the uniformity coefficient of the stage fertilizer is calculated based on the distribution of samples in the clustering results.

[0040] As a preferred embodiment, the method of clustering using the core vector set of all samples of the stage fertilizer and calculating the uniformity coefficient of the stage fertilizer based on the distribution of samples in the clustering results includes: obtaining the core vector set of all samples of any stage fertilizer; setting the parameters of the K-means clustering algorithm. The uniformity coefficient of the stage fertilizer is calculated based on the Euclidean distance between the core vector groups and the K-means clustering algorithm. The cluster center of the cluster is recorded as the first cluster center. The sample that is far from the first cluster center is recorded as the second cluster center. The uniformity coefficient of the stage fertilizer is calculated based on the distance difference between the sample in the cluster and the first cluster center and the second cluster center.

[0041] It should be noted that the K-means clustering algorithm is an existing clustering algorithm, so it will not be described in detail in the embodiments of this invention.

[0042] As an optional embodiment, the method for calculating the uniformity coefficient of the stage fertilizer based on the distance differences between samples in the cluster and the centers of the first and second clusters is as follows: obtaining the Euclidean distances between any sample in the cluster and the centers of the first and second clusters, respectively, and recording them as the first distance and the second distance; using the absolute value of the difference between the first distance and the second distance as the deviation distance of the sample; dividing the deviation distance of the sample by the maximum value of the first distance and the second distance; using the result as the deviation parameter of the sample; and obtaining the average value of the deviation parameters corresponding to all samples in the stage fertilizer to which the sample belongs, as the uniformity coefficient of the stage fertilizer.

[0043] It should be noted that the cluster centers obtained by clustering can only reflect the difference between the sample and the ideal uniform state. However, when a certain basal fertilizer sample is at a certain distance from the uniform state but does not reach the degree of non-uniformity that causes abnormal fruit tree growth, it is difficult to judge based on a single uniform benchmark. Therefore, it is also necessary to construct a comparison benchmark that represents the non-uniform state.

[0044] It should be noted that the fertilizer uniformity quantified by the profile coefficient reflects the consistency of nutrient distribution among the various fertilizer types in the apple tree-specific fertilizer package. This provides a standardized quantitative indicator for subsequent calculations of fertilizer comprehensive suitability, clarifying both the basic quality baseline of individual fertilizer types and the evaluation criteria for nutrient synergy during the growing season of the fertilizer package.

[0045] Thus, the uniformity coefficient of fertilizer at each stage in the package fertilizer was obtained through the above method.

[0046] Step S003: Calculate the formulation accuracy of the fertilizer of the sample stage by using the distribution of the dimensionality reduction results and the difference between the actual proportion and the theoretical proportion of the core components in the sample; combine the difference in formulation accuracy and uniformity coefficient of the fertilizer of the next stage in adjacent periods to obtain the component skewness of the package fertilizer.

[0047] It should be noted that the proportions of the components in each stage of the fertilizer package must meet the theoretical requirements for the fertilizer package to be effective in promoting apple tree growth and fruiting. If the proportions of the components in each stage of the fertilizer package deviate from the theoretical requirements, the fertilizer in that stage will not be able to effectively exert its corresponding effect at the corresponding growth stage of the apple tree. Since the growth of apple trees is a continuous and progressive life cycle with strong connections between each stage, it is necessary to ensure the continuity of nutrient supply between fertilizers in order to avoid problems of nutrient gaps or excesses.

[0048] Specifically, as a preferred embodiment, the method for calculating the formulation accuracy of the fertilizer to which the sample belongs by utilizing the distribution of the dimensionality reduction results and the difference between the actual and theoretical proportions of the core components in the sample; and combining the differences in formulation accuracy and uniformity coefficients of fertilizers in adjacent periods to obtain the component skewness of the fertilizer package, includes the following specific methods: First, by utilizing the eigenvalues ​​of the spectral data segments of the core components corresponding to the sample of any stage fertilizer during the dimensionality reduction process, and establishing a Cartesian coordinate system, the uniform influence coefficient of the stage fertilizer is obtained based on the discreteness of all core components of the sample in the Cartesian coordinate system.

[0049] As a preferred embodiment, the specific method for obtaining the uniform influence coefficient of the stage fertilizer is as follows: obtain the feature values ​​of the spectral data segments corresponding to all core components of any sample during the principal component analysis process, and denote the array formed by the feature values ​​as the core feature array of the sample. Construct a rectangular coordinate system with the same number of dimensions as the number of core components, and use the feature value of each core component as the coordinate axis of the rectangular coordinate system. Obtain a scatter plot of all samples of any stage fertilizer based on the corresponding core feature array in the rectangular coordinate system. Obtain the standard deviation of the coordinates of all samples in the scatter plot, denote it as the component dispersion factor of the stage fertilizer, and obtain the uniform influence coefficient of the stage fertilizer based on the component dispersion factor. The component dispersion factor is negatively correlated with the uniform influence coefficient.

[0050] It should be noted that, due to process limitations, stage-specific fertilizers are prone to microscopic inhomogeneities during production (e.g., local potassium content fluctuations of ±15%). Although the uniformity coefficient meets the standard, the drastic dispersion of the spectral characteristics of key components (such as potassium during fruit expansion) can lead to uneven sugar content distribution in the fruit. Therefore, the uniformity coefficient may not be able to quantify the impact of nutrient fluctuations on functionality. Thus, this invention uses a uniformity influence coefficient to describe the reliability of stage-specific fertilizers in nutrient supply, transforming the testing from a simple uniformity evaluation to a functional assessment of the fertilizer, avoiding fertilizer ineffectiveness due to characteristic fluctuations, and ensuring the continuity of apple tree growth.

[0051] As an optional embodiment, the specific calculation method for the uniformity influence coefficient of the stage fertilizer for any stage fertilizer is as follows: In the formula, This represents the uniformity of the effect of staged fertilizer application; The component dispersion factor of stage fertilizer; This represents an exponential function with the natural constant as its base.

[0052] It should be noted that the uniformity influence coefficient is used to describe the constraint of uniformity on fertilizer function. When the CV is zero, the means of each principal component are consistent, the PCA features are not discrete, and the stability of fertilizer nutrient attack reaches its optimal level. The uniformity score ensures the uniformity of nutrient distribution within the fertilizer. On this basis, the fertilizer formulation accuracy score also needs to be included, because the formulation accuracy score quantifies the actual nutrient characteristics of the fertilizer and the degree of conformity with the nutrient requirements of apple trees at the corresponding growth stage. The combination of the two can better reflect the performance of the fertilizer from the aspects of nutrient supply stability and growth demand adaptability. Then, the content of the core component is obtained from the spectral data of any sample, and the actual component ratio of the core component in the sample is obtained by the ratio of the contents of the core components. Based on the difference between the actual component ratio and the theoretical component ratio of any core component in the sample, the deviation factor of the core component is calculated. The average value of the deviation factors of all core components in all samples of the stage fertilizer is recorded as the fertilizer formulation accuracy of the stage fertilizer.

[0053] As an optional embodiment, for any core component of any sample, the specific calculation method for the deviation factor of the core component is as follows: In the formula, The deviation factor representing the core component; This indicates the actual proportion of the core component in the sample. This indicates the theoretical proportion of the core components in the fertilizer at the corresponding stage; This indicates that the absolute value is obtained.

[0054] It should be noted that the deviation factor is used to describe the degree of difference between the actual component ratio of the corresponding core component and the theoretical component ratio of the core component in the fertilizer at the corresponding stage; in the formula... This means dividing the absolute deviation by the proportion of the theoretical component, which essentially normalizes the deviation.

[0055] Finally, based on the uniformity coefficient, uniformity influence coefficient, and fertilizer formulation accuracy of the stage fertilizer, the quality coefficient of the stage fertilizer is obtained; the difference in quality coefficients between stage fertilizers used for adjacent periods of apple trees in the package fertilizer is obtained, and the compositional bias of the package fertilizer is obtained.

[0056] As an optional embodiment, for any stage fertilizer, the specific calculation method for the quality coefficient of the stage fertilizer is as follows: ;in, This indicates the quality coefficient of the fertilizer at different stages; This represents the uniformity of the effect of staged fertilizer application; Indicates the uniformity coefficient of fertilizer applied in stages; This indicates the precision of the fertilizer formula for staged fertilizers.

[0057] It should be noted that a single stage fertilizer needs to simultaneously meet the requirements of uniform spatial distribution, nutrient formulation matching the growth stage of the apple tree, and stable underlying nutrient characteristics to avoid unstable nutrient supply due to characteristic fluctuations. Therefore, by obtaining the quality coefficient of the stage fertilizer, it is used to describe the comprehensive ability index of the stage fertilizer to meet the needs of the apple tree during its growth period. The larger the quality coefficient value, the stronger the comprehensive ability of the stage fertilizer to meet the needs of the apple tree during its growth period, and vice versa.

[0058] It should be noted that apple tree growth is a continuous life cycle, and there is a continuity of nutrient requirements between different growth stages. For example, the base fertilizer can provide the nitrogen needed for budding. The nitrogen residue in the soil at the later stage of this stage must be just enough to meet the initial nitrogen requirement of the budding stage. If the nitrogen residue at the base fertilizer stage is insufficient, even if the nitrogen content of the fertilizer at the budding stage meets the standard, poor budding will still occur due to insufficient reserves in the early stage.

[0059] As an optional embodiment, the specific calculation method for the component bias of the fertilizer package is as follows: ,in Indicates the degree of bias in the composition of the fertilizer package; This represents the average quality coefficient of all stages of fertilizer in the fertilizer package; Indicates the first [item] used for apple trees The quality coefficient of stage fertilizer in each period; Indicates the first [item] used for apple trees The quality coefficient of stage fertilizer in each period; This indicates obtaining the absolute value; Represents an exponential function with the natural constant as its base; This represents the first preset parameter.

[0060] It should be noted that, in this embodiment of the invention, the first parameter is preset to 0.1 based on experience to avoid the case where the denominator in the calculation formula is 0. In other embodiments, it can be adjusted according to the actual situation. This embodiment of the invention does not make specific limitations.

[0061] It should be noted that the key to fertilizer adaptation at each stage is the degree to which the residual supply of fertilizer in the previous stage matches the growth requirements of the next stage. Furthermore, the uniformity of fertilizer in the previous stage determines the stability of nutrient residue. If the uniformity of the previous fertilizer is low, even if the average nutrient residue perfectly matches the requirements of the next stage, actual planting may result in localized areas with excessive or insufficient residue, leading to nutrient transition failure. Therefore, this embodiment of the invention analyzes the differences in quality coefficients corresponding to fertilizers used in adjacent stages of apple tree growth within the fertilizer package, thereby describing whether the overall effect of the core components in the fertilizer package deviates from the effect to be achieved by the theoretical proportions. The closer the value is to 1, the more similar the quality coefficients of the two stages of fertilizer are. The larger the value, the higher the overall quality of the fertilizer composition, meaning the closer the composition is to the theoretical requirements. The smaller the component deviation value, the smaller the deviation of the core components; conversely, the larger the value, the lower the deviation.

[0062] Thus, the compositional bias of the fertilizer package was obtained through the above method.

[0063] Step S004: Obtain the component detection results of the fertilizer package using component bias.

[0064] Specifically, the component skewness of the fertilizer package is obtained and compared with a preset skewness threshold. If the component skewness is less than or equal to the skewness threshold, the fertilizer package is deemed to meet the growth requirements of apple trees, and a qualified test report is generated. The report includes the component skewness value and fertilizer quality coefficients for each stage. If the component skewness is greater than the skewness threshold, the component is deemed to have skewness, an unqualified report is output, and the stage of skewness exceeding the standard is marked.

[0065] It should be noted that in the embodiments of the present invention, the preset skewness threshold is 0.6 based on experience. In other embodiments, it can be adjusted according to the actual situation. The embodiments of the present invention do not impose specific limitations.

[0066] The above steps complete the component testing of the fertilizer package used for apple trees.

[0067] Please see Figure 2The diagram illustrates a structural block diagram of a component detection system for apple tree-specific fertilizers according to an embodiment of the present invention. The system includes a memory 202, a processor 201, and a computer program 2021 stored in the memory 202 and executable on the processor. When the processor 201 executes the computer program 2021, it implements steps S001 to S004 of the component detection method for apple tree-specific fertilizers.

[0068] Furthermore, in an optional embodiment, the memory 202 described above may include read-only memory and random access memory, and provide instructions and data to the processor. The memory 202 may also include non-volatile random access memory. For example, the memory may also store device type information.

[0069] The processor 201 mentioned above can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0070] It should be noted that the embodiments used in this example The model is only used to represent negative correlations and the results of the constraint model output are in Within this range, in specific implementations, other models with the same purpose can be substituted; this embodiment is merely an example. The description will be based on a model, without making specific limitations on it. This refers to the input of the model.

[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting the components of a fertilizer specifically for apple trees, characterized in that, The method includes the following steps: Obtain the packaged fertilizer for apple trees, and perform sampling and multispectral detection to obtain the spectral data of the samples; the packaged fertilizer contains several stage fertilizers for different stages of apple tree growth, with several samples corresponding to each stage fertilizer; obtain the theoretical component ratio of each stage fertilizer in the packaged fertilizer. The core components of the fertilizer package are obtained. The spectral data corresponding to the samples of the fertilizer at any stage are dimensionality-reduced by combining the theoretical component ratios of the core components in the fertilizer at any stage. The dimensionality-reduction results are then used to cluster all samples of the fertilizer at each stage. The uniformity coefficient of the fertilizer at each stage is calculated based on the distribution of samples in the clustering results. This process includes: collectively referring to nitrogen, phosphorus, and potassium as core components; obtaining the spectral data segments corresponding to the core components in the spectral data of any sample of any fertilizer at any stage and performing principal component analysis; combining the theoretical component ratios of nitrogen, phosphorus, and potassium in the fertilizer at each stage as weights during the principal component analysis to obtain the core component vector of the sample; and clustering the core vector groups of all samples of the fertilizer at each stage. The uniformity coefficient of the fertilizer at each stage is then calculated based on the distribution of samples in the clustering results. By utilizing the distribution of dimensionality reduction results and the difference between the actual and theoretical proportions of core components in the sample, the formulation accuracy of the fertilizer to which the sample belongs is calculated; by combining the differences in formulation accuracy and uniformity coefficient of fertilizer in adjacent periods, the component skewness of the package fertilizer is obtained. The component analysis results of the packaged fertilizer were obtained by utilizing component skewness. Specifically, the spectral band intervals corresponding to nitrogen, phosphorus, and potassium in the spectral data are obtained and denoted as the nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval, respectively. For any stage fertilizer, the data segments corresponding to the nitrogen spectral interval, phosphorus spectral interval, and potassium spectral interval in the spectral data of any sample of the stage fertilizer are obtained and denoted as the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, respectively. Combining the theoretical component ratio of the fertilizer package to which the sample belongs, and the covariance matrix obtained during the principal component analysis of the nitrogen spectral data segment, phosphorus spectral segment, and potassium spectral segment of the sample, dimensionality reduction is performed to obtain the core component vector of the sample. Specifically, based on the theoretical component proportions of the core components in the fertilizer at the sample's stage, the proportional weights of the core components are obtained; the covariance matrices of the nitrogen, phosphorus, and potassium spectral data segments are obtained through principal component analysis, denoted as the nitrogen covariance matrix, phosphorus covariance matrix, and potassium covariance matrix, respectively; the proportional weight of nitrogen is multiplied by the nitrogen covariance matrix to obtain the nitrogen weighted covariance matrix; similarly, the weighted covariance matrix of each core component is obtained; the weighted covariance matrices of all core components are processed by principal component analysis to obtain the eigenvectors of the corresponding core components; the eigenvectors of all core components in the sample are formed into an array, denoted as the sample's core vector group; Specifically, the process involves using the eigenvalues ​​of the spectral data segments corresponding to the core components of any stage fertilizer sample during dimensionality reduction and establishing a Cartesian coordinate system. Based on the dispersion of all core components in the Cartesian coordinate system, the uniformity influence coefficient of the stage fertilizer is obtained. The content of the core components is acquired from the spectral data of any sample, and the actual proportion of the core components in the sample is obtained by comparing the proportions of the contents between the core components. The deviation factor of the core components is calculated based on the difference between the actual and theoretical proportions of any core component in the sample. The average deviation factor of all core components in all samples of the stage fertilizer is recorded as the fertilizer formulation accuracy of the stage fertilizer. The quality coefficient of the stage fertilizer is obtained based on the uniformity coefficient, uniformity influence coefficient, and fertilizer formulation accuracy. Finally, the component skewness of the package fertilizer is obtained by acquiring the difference in the quality coefficients between stage fertilizers used for adjacent periods of apple trees.

2. The method for component detection of a fertilizer specifically for apple trees according to claim 1, characterized in that, The specific method for obtaining the proportion weight of the core component based on the theoretical proportion of the core component in the fertilizer of the sample stage is as follows: Obtain the corresponding proportion values ​​of nitrogen, phosphorus, and potassium in the theoretical component ratio, and record them as nitrogen ratio value, phosphorus ratio value, and potassium ratio value respectively. The nitrogen ratio value, phosphorus ratio value, and potassium ratio value are collectively referred to as component ratio value. The ratio between the component ratio value corresponding to any core component and the sum of the component ratio values ​​of all core components is used as the ratio weight of the core component.

3. The method for component detection of a fertilizer specifically for apple trees according to claim 1, characterized in that, The method for clustering the core vector groups of all samples of the stage fertilizer and calculating the uniformity coefficient of the stage fertilizer based on the distribution of samples in the clustering results includes the following specific methods: Obtain the core vector set of all samples at any stage of fertilizer growth; set the parameters of the K-means clustering algorithm. The uniformity coefficient of the stage fertilizer is calculated based on the Euclidean distance between the core vector groups and the K-means clustering algorithm. The cluster center of the cluster is recorded as the first cluster center. The sample that is far from the first cluster center is recorded as the second cluster center. The uniformity coefficient of the stage fertilizer is calculated based on the distance difference between the sample in the cluster and the first cluster center and the second cluster center.

4. The method for component detection of a fertilizer specifically for apple trees according to claim 3, characterized in that, The method for calculating the uniformity coefficient of the stage fertilizer based on the distance differences between samples in the cluster and the centers of the first and second clusters is as follows: Obtain the Euclidean distances between any sample in the cluster and the centers of the first and second clusters, and denot them as the first distance and the second distance, respectively. Use the absolute value of the difference between the first distance and the second distance as the deviation distance of the sample. Divide the deviation distance of the sample by the maximum value of the first distance and the second distance, and use the result as the deviation parameter of the sample. Obtain the average value of the deviation parameters of all samples in the stage fertilizer to which the sample belongs, and use it as the uniformity coefficient of the stage fertilizer.

5. The method for component detection of a fertilizer specifically for apple trees according to claim 1, characterized in that, The uniformity influence coefficient of the stage fertilizer includes the following specific methods: The eigenvalues ​​of the spectral data segments corresponding to all core components of any sample are obtained during principal component analysis. The array formed by these eigenvalues ​​is denoted as the core feature array of the sample. A Cartesian coordinate system is constructed with the same number of dimensions as the number of core components. The eigenvalues ​​of each core component are used as the coordinate axes of the Cartesian coordinate system. A scatter plot of all samples of any stage fertilizer based on the corresponding core feature array is obtained in the Cartesian coordinate system. The standard deviation of the coordinates of all samples in the scatter plot is obtained and denoted as the component dispersion factor of the stage fertilizer. Based on the component dispersion factor, the uniform influence coefficient of the stage fertilizer is obtained. The component dispersion factor and the uniform influence coefficient are negatively correlated.

6. A component detection system for a fertilizer specifically for apple trees, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for detecting the components of a fertilizer specifically for apple trees as described in any one of claims 1 to 5.

Images