Tobacco agronomic character acquisition and ideal plant type analysis and evaluation system and method

Abstract

The invention provides a tobacco agronomic character collection and ideal plant type analysis and evaluation system and method, and relates to the technical field of tobacco planting management, and the system comprises an Internet of Things device which is used for collecting tobacco agronomic character data; the mobile terminal is used for preprocessing the data to obtain a tobacco agronomic character sample; the server side is used for constructing a sample vector and a sample vector set according to the agronomic character sample of the tobacco, and calculating the mahalanobis distance of the sample vector by adopting a robust mahalanobis distance method according to a pre-constructed ideal plant type vector so as to obtain a first analysis evaluation result; and the server side is also used for calculating to obtain a robust mean value and a standard deviation by adopting a minimum covariance determinant method according to the sample vector set, and calculating to obtain a robust value of the sample vector according to the robust mean value, the standard deviation and the sample vector so as to obtain a second analysis evaluation result. The data acquisition efficiency, the data management capability and the difference analysis capability of the agronomic traits of the tobacco can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tobacco planting management, in particular to a tobacco agronomic character collection and ideal plant type analysis and evaluation system and method. BACKGROUND

[0002] Tobacco is an important economic crop, and its agronomic characters are key indicators for measuring tobacco yield, quality and variety breeding. Accurate and efficient collection of these agronomic character data and scientific analysis and evaluation can not only provide important reference for tobacco variety breeding, but also become an effective evaluation basis for the cultivation management level of planting bases.

[0003] The ideal plant type is a high-quality growth form set according to the breeding goals and production needs of tobacco, which has optimized agronomic character indicators and can reflect the excellent characteristics and adaptability of tobacco varieties. In the planting base, by comparing with the ideal plant type, the following can be achieved: (1) guiding variety breeding: by analyzing the differences in agronomic characters of different materials, high-quality varieties meeting the breeding goals can be selected. (2) evaluating the cultivation management level: by tracking the changes of the characters of a certain variety for many years, the cultivation management technology level of the planting base can be evaluated. (3) monitoring the stability of the variety: by comparing the character differences between the ideal plant type and the sample for a long time, the performance fluctuation of a specific variety under the cultivation management condition can be found and analyzed.

[0004] However, there are still the following main problems in the collection, management and analysis and evaluation of tobacco agronomic character data: (1) low data collection efficiency: traditional field measurement relies on manual operation, which is time-consuming and labor-intensive, and cannot meet the rapid collection needs under large-scale planting conditions; paper records are easily affected by subjective and environmental factors, resulting in inaccurate data and tedious sorting. (2) Lack of systematic data management: existing methods are mostly stored in scattered records, and data is easily lost and incomplete. (3) Single analysis method: existing analysis methods are often only for a single indicator, and it is difficult to comprehensively measure the differences between samples and ideal plant types from multiple dimensions, which cannot effectively guide tobacco breeding and base management.

[0005] Therefore, how to improve the data collection efficiency, data management capability and difference analysis capability of tobacco agronomic characters is a technical problem to be solved by those skilled in the art. SUMMARY

[0006] To solve the above technical problems, the present application provides a tobacco agronomic character collection and ideal plant type analysis and evaluation system, which can improve the data collection efficiency, data management capability and difference analysis capability of tobacco agronomic characters. The present application also provides a tobacco agronomic character collection and ideal plant type analysis and evaluation method, which has the same technical effect.

[0007] The first object of the present application is to provide a tobacco agronomic character collection and ideal plant type analysis and evaluation system.

[0008] The first object of the present application is achieved by the following technical solution: The tobacco agronomic character collection and ideal plant type analysis and evaluation system comprises an Internet of Things device, a mobile terminal and a server end, wherein: The Internet of Things device is configured to collect tobacco agronomic character data and transmit the data to the mobile terminal. The mobile terminal is configured to preprocess the tobacco agronomic character data to obtain tobacco agronomic character samples and transmit the samples to the server end. The server end is configured to construct a sample vector and a sample vector set based on the tobacco agronomic character samples, calculate the Mahalanobis distance of the sample vector based on the sample vector, the sample vector set and a pre-constructed ideal plant type vector using a robust Mahalanobis distance method, and obtain a first analysis and evaluation result based on the Mahalanobis distance of the sample vector. The server end is further configured to calculate a robust mean and a standard deviation based on the sample vector set using a minimum covariance determinant method, calculate a robust value of the sample vector based on the robust mean, the standard deviation and the sample vector, and obtain a second analysis and evaluation result based on the robust value of the sample vector.

[0009] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the server end is further configured to construct the ideal plant type vector, and when performing the construction of the ideal plant type vector, the server end is specifically configured to: calculate a sample mean based on all tobacco agronomic character samples to obtain a mean sample; construct the ideal plant type vector based on the mean sample.

[0010] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the server end is further configured to construct the ideal plant type vector, and when performing the construction of the ideal plant type vector, the server end is specifically configured to: obtain a prediction model corresponding to each preset agronomic character index in the sample vector, wherein the prediction model is specifically a Gaussian process regression model taking an environmental feature vector as input and a predicted value of the preset agronomic character index as output; obtain a real-time environmental feature vector, input the real-time environmental feature vector into the prediction model corresponding to each preset agronomic character index to obtain a real-time predicted value of each preset agronomic character index; and construct the ideal plant type vector based on the real-time predicted value of each preset agronomic character index.

[0011] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the formula for calculating the Mahalanobis distance is: ; wherein, denotes the Mahalanobis distance, denotes the sample vector, denotes the ideal plant type vector, denotes the transposition operation, denotes the inverse covariance matrix calculated from the sample vector set by the minimum covariance determinant.

[0012] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the formula for calculating the robust value is: ; wherein, denotes the robust value, denotes the robust mean value, denotes the standard deviation.

[0013] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the server, when performing the first analysis and evaluation result according to the Mahalanobis distance of the sample vector, is specifically configured to: compare the size of the Mahalanobis distance of the sample vector and a first preset threshold to obtain the first analysis and evaluation result; The server, when performing the second analysis and evaluation result according to the robust value of the sample vector, is specifically configured to: compare the size of the robust value of the sample vector and a second preset threshold to obtain the second analysis and evaluation result.

[0014] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the mobile terminal, when performing the preprocessing of the tobacco agronomic character data to obtain the tobacco agronomic character sample, is specifically configured to: perform outlier rejection processing on the tobacco agronomic character data, fill in missing values using an interpolation method, and then perform a predetermined format conversion to obtain the tobacco agronomic character sample.

[0015] Preferably, in the tobacco agronomic character collection and ideal plant type analysis and evaluation system, the server is further configured to perform statistical analysis on the tobacco agronomic character sample to obtain statistical indicators, wherein the statistical indicators include one or more of the mean value, the maximum value, the minimum value, the standard deviation, and the variance of each preset agronomic character indicator.

[0016] Preferably, in the tobacco agronomic trait collection and ideal plant type analysis and evaluation system, the server end is further configured to generate a table display according to the statistical index, the Mahalanobis distance of the sample vector, and the robust value of the sample vector. The server end is further configured to generate a radar chart and a column chart display according to the sample vector and the ideal plant type vector. The server end is further configured to extract a first principal component and a second principal component by using principal component analysis according to the tobacco agronomic trait sample, perform cluster analysis to obtain a category label, and visualize the cluster result in a scatter plot in different colors.

[0017] The second object of the present application is to provide a tobacco agronomic trait collection and ideal plant type analysis and evaluation method.

[0018] The above-mentioned second object of the present application is achieved by the following technical scheme. A tobacco agronomic trait collection and ideal plant type analysis and evaluation method is realized based on the tobacco agronomic trait collection and ideal plant type analysis and evaluation system as described above, the system comprising an Internet of Things device, a mobile terminal, and a server end, and the method comprising: Collecting tobacco agronomic trait data by using the Internet of Things device and transmitting the data to the mobile terminal; Preprocessing the tobacco agronomic trait data by using the mobile terminal to obtain a tobacco agronomic trait sample and transmitting the sample to the server end; Constructing a sample vector and a sample vector set according to the tobacco agronomic trait sample by using the server end, and calculating the Mahalanobis distance of the sample vector by using a robust Mahalanobis distance method according to the sample vector, the sample vector set, and a pre-constructed ideal plant type vector, obtaining a first analysis and evaluation result according to the Mahalanobis distance of the sample vector; Calculating a robust mean and a standard deviation by using a minimum covariance determinant method according to the sample vector set by using the server end, calculating the robust value of the sample vector according to the robust mean, the standard deviation, and the sample vector, and obtaining a second analysis and evaluation result according to the robust value of the sample vector.

[0019] The tobacco agronomic character data is collected by the Internet of Things device and transmitted to the mobile terminal, rapid and accurate measurement of field agronomic characters and real-time data transmission are realized, work efficiency is greatly improved and manual operation error is reduced; the tobacco agronomic character data is preprocessed by the mobile terminal, tobacco agronomic character samples are obtained, and the tobacco agronomic character samples are transmitted to the server end, automatic preprocessing of agronomic character data is realized, scattered sample data is sent to the unified server end, and digital management is realized; according to the tobacco agronomic character samples, a sample vector and a sample vector set are constructed by the server end, and according to the sample vector, the sample vector set and the pre-constructed ideal plant type vector, the Mahalanobis distance of the sample vector is calculated by using the robust Mahalanobis distance method, the first analysis and evaluation result is obtained according to the Mahalanobis distance of the sample vector, the correlation between characters, scale difference and abnormal value are comprehensively processed by introducing the robust Mahalanobis distance, and the multi-character comprehensive difference evaluation method is more stable, more real and more statistically significant; according to the sample vector set, the robust mean and standard deviation are calculated by using the minimum covariance determinant method by the server end, the robust value of the sample vector is calculated according to the robust mean, the standard deviation and the sample vector, the second analysis and evaluation result is obtained according to the robust value of the sample vector, the accuracy of difference identification is significantly improved by introducing the robust value index calculation, the difference between the sample and the ideal plant type is accurately evaluated, and quantitative basis is provided for variety improvement.

[0020] In summary, the technical scheme can improve the data collection efficiency, data management capability and difference analysis capability of tobacco agronomic characters. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0022] Figure 1 An interactive schematic diagram of a tobacco agronomic character collection and ideal plant type analysis and evaluation system provided in an embodiment of the present application; Figure 2 A radar chart provided in an embodiment of the present application; Figure 3 A column chart provided in an embodiment of the present application; Figure 4 A scatter plot provided in an embodiment of the present application; Figure 5 A flowchart of a tobacco agronomic character collection and ideal plant type analysis and evaluation method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described as follows. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0024] In the embodiments provided in the present application, it should be understood that the disclosed method and system can be implemented in other manners. The embodiments described below are only schematic. For example, the division of the units and modules is only a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling, direct coupling or communication connection between the components can be indirect coupling or communication connection through some interfaces, devices or modules, and can be electrical, mechanical or other forms.

[0025] In addition, each functional unit in each embodiment of the present application can be integrated into one processor, or each unit can be a separate device, or two or more units can be integrated into one device. Each functional unit in each embodiment of the present application can be implemented in the form of hardware, or in the form of hardware plus software functional units.

[0026] Those skilled in the art can understand that all or part of the steps of the method embodiments described below can be completed by program instructions and related hardware. The aforementioned program instructions can be stored in a computer readable storage medium, and when the program instructions are executed, the steps of the method embodiments are executed. The aforementioned storage medium includes mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, and various media that can store program codes.

[0027] It should be understood that if "system", "device", "unit" and / or "module" are used in the present application, it is only a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0028] In addition, the terms "first", "second", etc. are used only for descriptive purposes and do not connote or imply relative importance or a quantity of the indicated technical features. Thus, a feature defined with "first", "second", etc. can include one or more of the features explicitly or implicitly. In the description of the present application, the meaning of "a plurality of" or "several" is two or more, unless otherwise expressly specified.

[0029] If flowcharts are used in the present application, the flowcharts are used to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or subsequent operations are not necessarily performed in sequence. Instead, each step can be processed in reverse order or simultaneously. Meanwhile, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0030] It should also be noted that in the present document, terms such as "comprise", "comprising", or any other variant thereof are intended to cover non-exclusive inclusions, so that an item or device comprising a list of elements does not only include those elements, but also includes other elements not explicitly listed, or inherent to such item or device. Without more limitations, an element defined by the phrase "comprising a" does not exclude the presence of additional identical elements in the item or device comprising the above element.

[0031] Embodiments of the present application are written in a progressive manner.

[0032] As shown in Figure 1 The present application provides a tobacco agronomic trait collection and ideal plant type analysis and evaluation system, comprising: an Internet of Things device, a mobile terminal and a server end, wherein: The Internet of Things device is used to collect tobacco agronomic trait data and transmit it to the mobile terminal; Specifically, the Internet of Things device can include a Bluetooth vernier caliper and a Bluetooth tape measure. By using the Internet of Things device, tobacco agronomic trait data in the field can be quickly and accurately collected, and specific indicators include one or more of the following: topping plant height, leaf number, pitch, stem circumference, middle leaf length, middle leaf width, upper leaf length, and upper leaf width. The Internet of Things device can be connected to the mobile terminal through Bluetooth, and the Bluetooth text data transmission is adopted to transmit the tobacco agronomic trait data to the mobile terminal in real time, so as to realize automatic recording of data, thereby greatly improving work efficiency and reducing manual operation errors. The Internet of Things device and the mobile terminal can also be connected through Wi-Fi or other communication methods, which is not limited by the present application. In order to ensure the accuracy and standardization of measurement, data collection is strictly performed according to the standard of "Tobacco Agronomic Trait Survey and Measurement Method" (YC / T 142-2010). During the measurement process, the measurer can measure the agronomic trait indicators of tobacco according to the prompts of the mobile terminal.

[0033] The mobile terminal is configured to preprocess tobacco agronomic trait data, obtain a tobacco agronomic trait sample, and transmit the tobacco agronomic trait sample to a server end. Specifically, the mobile terminal can use a Web service page to collect and manage data. A measurer can select a project name, a field, and a sample on the Web service page, and enter tobacco agronomic trait data item by item. The mobile terminal supports item-by-item measurement and automatically jumps to the next index entry, reducing the operation steps. To improve data quality, the mobile terminal also preprocesses the tobacco agronomic trait data sent by the Internet of Things device to obtain a tobacco agronomic trait sample. The scattered tobacco agronomic trait samples are transmitted to the server end through a wireless network for unified data management and analysis. In some embodiments, when the mobile terminal performs preprocessing on the tobacco agronomic trait data to obtain the tobacco agronomic trait sample, it is specifically configured to: perform outlier rejection processing on the tobacco agronomic trait data, and use an interpolation method to fill in missing values, and then perform predetermined format conversion to obtain the tobacco agronomic trait sample. In this way, the data quality is improved through outlier processing and missing value filling, and the data processing efficiency of the server end is improved through unified data format.

[0034] The server end is configured to construct a sample vector and a sample vector set according to the tobacco agronomic trait sample, and calculate the Mahalanobis distance of the sample vector according to the sample vector, the sample vector set, and a pre-constructed ideal plant type vector using a robust Mahalanobis distance method, and obtain a first analysis and evaluation result according to the Mahalanobis distance of the sample vector. Specifically, the server end can store and manage all tobacco agronomic trait samples uploaded by the mobile terminal. The tobacco agronomic trait samples can be stored in the central database of the server end in units of projects, supporting regular backup to ensure data security and long-term effectiveness. The server end uses a robust Mahalanobis distance (Robust Mahalanobis Distance) method to measure differences. Specifically, the server end can determine a tobacco agronomic trait sample to be analyzed and evaluated from multiple stored tobacco agronomic trait samples, construct a sample vector according to the sample to be analyzed and evaluated, denoted as ; construct a sample vector set from the multiple stored tobacco agronomic trait samples. The sample vector set can include multiple sample vectors. A sample matrix is constructed from the sample vector set, denoted as X. Then, a covariance matrix is estimated from the sample matrix X using a minimum covariance determinant (MCD) method. The MCD method only uses a data subset with the smallest volume to eliminate extreme outliers, obtaining a covariance matrix that is not sensitive to abnormal data. Then, based on the covariance matrix , the inverse covariance matrix is calculated The MCD method is prior art, and the specific calculation process thereof will not be described herein. Then, an ideal plant type vector is obtained, denoted as wherein the preset agronomic trait indexes in the ideal plant type vector are the same as the preset agronomic trait indexes of the sample vector, and the values of the preset agronomic trait indexes in the ideal plant type vector can be set artificially according to actual needs or can be automatically generated by the server side. For example, in some embodiments, the server side is further configured to construct the ideal plant type vector, and when constructing the ideal plant type vector, the server side is specifically configured to: calculate a sample mean value according to all tobacco agronomic trait samples to obtain a mean sample; and construct the ideal plant type vector according to the mean sample. The reference value is set based on the sample mean value, which is used for subsequent sample difference evaluation. After obtaining the sample vector , the inverse covariance matrix and the ideal plant type vector , the Mahalanobis distance of the sample vector can be calculated, and the specific calculation formula is as follows: ; wherein denotes the Mahalanobis distance, denotes the transposition operation, and the inverse covariance matrix is calculated from the sample vector set by the minimum covariance determinant. The smaller the Mahalanobis distance is, the closer the tobacco agronomic trait sample to be analyzed and evaluated is to the ideal plant type, and the larger the Mahalanobis distance is, the more serious the deviation of the tobacco agronomic trait sample to be analyzed and evaluated from the ideal plant type is. The server side can obtain the first analysis and evaluation result according to the size of the Mahalanobis distance of the sample vector and a preset evaluation rule.

[0035] In some embodiments, when the server side performs the first analysis and evaluation result according to the Mahalanobis distance of the sample vector, the server side is specifically configured to: compare the size of the Mahalanobis distance of the sample vector and a first preset threshold to obtain the first analysis and evaluation result. Specifically, the first preset threshold can be set based on actual needs, and when the Mahalanobis distance of the sample vector is greater than the first preset threshold, the tobacco agronomic trait sample to be analyzed and evaluated can be determined as an abnormal sample as the first analysis and evaluation result, and when the Mahalanobis distance of the sample vector is less than or equal to the first preset threshold, the tobacco agronomic trait sample to be analyzed and evaluated can be determined as a normal sample as the first analysis and evaluation result, and the application is not limited thereto.

[0036] ​​The traditional Euclidean distance assumes that each trait is independently and identically distributed, and cannot process the correlation between traits and adapt to different dimensions. The robust Mahalanobis distance is used as a difference total metric measurement method in the application, which comprehensively processes the correlation between traits, scale difference and outliers, so that the multi-trait comprehensive difference evaluation method is more stable, more real and more statistically significant.

[0037] The server end is also configured to calculate the robust mean and standard deviation based on the sample vector set using the minimum covariance determinant method, calculate the robust value of the sample vector based on the robust mean, the standard deviation and the sample vector, and obtain the second analysis evaluation result based on the robust value of the sample vector.

[0038] Specifically, the server end can also calculate the robust mean and standard deviation based on the sample vector set using the minimum covariance determinant method, and denote them as and respectively. The specific calculation process of the MCD method is a prior art. After obtaining the sample vector , the robust mean and the standard deviation , the robust value of the sample vector can be calculated. The specific calculation formula is: ; wherein represents the robust value. The smaller the absolute value of is, the smaller the significant trait difference between the tobacco agronomic trait sample to be analyzed and evaluated and the ideal plant type is, and the larger the absolute value of is, the larger the significant trait difference between the tobacco agronomic trait sample to be analyzed and evaluated and the ideal plant type is. The server end can obtain the second analysis evaluation result based on the size of the robust value of the sample vector and a preset evaluation rule, thereby significantly improving the accuracy of difference identification and accurately evaluating the difference between the sample and the ideal plant type by introducing the robust value index calculation, and providing a quantitative basis for variety improvement.

[0039] In some embodiments, when the server end performs the second analysis evaluation result based on the robust value of the sample vector, it is specifically configured to compare the size of the robust value of the sample vector and a second preset threshold to obtain the first analysis evaluation result. Specifically, the second preset threshold can be set based on actual needs. For example, it can be set to 2.5. When the absolute value of the robust value of the sample vector is greater than the second preset threshold, it can be determined that there is a significant trait difference between the tobacco agronomic trait sample to be analyzed and evaluated and the ideal plant type, which is taken as the second analysis evaluation result. When the absolute value of the robust value of the sample vector When the absolute value is less than or equal to the second preset threshold, it can be determined that the tobacco agronomic trait sample to be analyzed and evaluated has little difference from the significant trait of the ideal plant type, and this can be used as the second analysis and evaluation result. This application is not limited to this.

[0040] Currently, the following major problems still exist in the collection, management, analysis and evaluation of tobacco agronomic traits data: (1) Low data collection efficiency: Traditional field measurements rely on manual operation, which is time-consuming and labor-intensive, making it difficult to meet the rapid collection needs under large-scale planting conditions; paper records are easily affected by subjectivity and environment, resulting in inaccurate data and cumbersome data processing. (2) Lack of systematic data management: Existing methods mostly store data in a scattered record form, making data easy to lose or incomplete. (3) Single analytical methods: Existing analytical methods often only target a single indicator, making it difficult to comprehensively measure the difference between the sample and the ideal plant type from multiple dimensions, and thus failing to effectively guide tobacco breeding and base management.

[0041] The above embodiments utilize IoT devices to collect tobacco agronomic trait data and transmit it to mobile terminals, enabling rapid and accurate measurement and real-time data transmission of agronomic traits in the field, significantly improving work efficiency and reducing human error. The mobile terminals preprocess the tobacco agronomic trait data to obtain tobacco agronomic trait samples, which are then transmitted to the server, achieving automatic preprocessing of agronomic trait data and sending scattered sample data to a unified server for digital management. On the server side, based on the tobacco agronomic trait samples, sample vectors and sample vector sets are constructed. Using the robust Mahalanobis distance method, the sample vectors, sample vector sets, and pre-constructed ideal plant type vectors are used to calculate the sample... The Mahalanobis distance of the vectors is used to obtain the first analytical evaluation result. By introducing robust Mahalanobis distance, a comprehensive handling of correlations, scale differences, and outliers among traits is achieved, making the multi-trait comprehensive difference assessment method more stable, realistic, and statistically significant. On the server side, based on the sample vector set, the robust mean and standard deviation are calculated using the minimum covariance determinant method. Based on the robust mean, standard deviation, and sample vectors, the robust value of the sample vectors is calculated, and the second analytical evaluation result is obtained based on the robust value of the sample vectors. By introducing the robust value index, the accuracy of difference identification is significantly improved, and the difference between the sample and the ideal plant type is accurately assessed, providing a quantitative basis for variety improvement. In summary, the above embodiments can improve the data collection efficiency, data management capabilities, and difference analysis capabilities of tobacco agronomic traits.

[0042] In other embodiments of this application, when the server performs the construction of the ideal plant type vector, it is specifically used for: Obtain the prediction model corresponding to each preset agronomic trait index in the sample vector. Specifically, the prediction model is a Gaussian process regression model that takes the environmental feature vector as input and the predicted value of the preset agronomic trait index as output. Obtain real-time environmental feature vectors, input the real-time environmental feature vectors into the prediction model corresponding to each preset agronomic trait index, and obtain the real-time predicted value of each preset agronomic trait index. An ideal plant type vector is constructed based on the real-time predicted values ​​of each preset agronomic trait index.

[0043] Specifically, each preset agronomic trait index in the sample vector can be flexibly set based on actual needs. For example, preset agronomic trait indexes include one or more of the following: topped plant height, number of leaves, internode distance, stem circumference, middle leaf length, middle leaf width, upper leaf length, and upper leaf width. For each preset agronomic trait index, a corresponding prediction model can be pre-constructed. This prediction model is specifically a Gaussian Process Regression (GPR) model that takes an environmental feature vector as input and outputs the predicted value of the preset agronomic trait index. The environmental feature vector can include multiple indicators such as temperature, rainfall, accumulated temperature, available potassium in the soil, and organic matter content. Each prediction model can be independently trained based on historical observation data of the corresponding preset agronomic trait index. Then, the real-time environmental feature vector is obtained and input into the prediction model corresponding to each preset agronomic trait index to obtain the real-time predicted value of each preset agronomic trait index. This serves as a basic reference value for constructing an ideal plant type vector. Used for subsequent sample difference assessment.

[0044] In some embodiments, the expression for a Gaussian process regression model is: ; ; in, Represents a continuous function to be predicted. It follows a Gaussian process distribution. Represents the environmental feature vector. Represents the mean function, This refers to kernel functions, such as the commonly used radial basis function (RBF). Indicates the length scale parameter. This represents the relative weights controlling small-scale and large-scale features.

[0045] For each preset agronomic trait index The server side uses existing historical observation data. Train independent Gaussian process models, inputting real-time environmental feature vectors. The index value for predicting the ideal plant type, that is, the real-time predicted value of each preset agronomic trait index. .

[0046] In this embodiment, an environment-dependent prediction model is introduced, abandoning the traditional fixed vector ideal plant type. This makes the ideal plant type vector no longer a fixed constant, but a "dynamic vector" that changes with the environment. It can fit the nonlinear and complex relationship between traits and the environment, automatically adapt to different production areas and different years, improve the accuracy of subsequent difference analysis, and significantly enhance the scientific nature and intelligence of the system.

[0047] In other embodiments of this application, the server is further configured to perform statistical analysis based on tobacco agronomic trait samples to obtain statistical indicators. These statistical indicators include one or more of the mean, maximum, minimum, standard deviation, and variance of each preset agronomic trait indicator. By calculating these statistical indicators, the overall growth trend and data distribution of the samples can be quickly analyzed.

[0048] In other embodiments of this application, the server-side is further configured to generate a table display based on statistical indicators, the Mahalanobis distance of the sample vectors, and the robustness value of the sample vectors. Specifically, the server-side can generate a table containing statistical indicators, the Mahalanobis distance of the sample vectors, and the robustness value of the sample vectors, supporting sorting by column name, downloading, and table display.

[0049] In other embodiments of this application, the server-side is further configured to generate radar charts and bar charts based on the sample vectors and the ideal plant type vector. Specifically, the radar chart is used to display the overall performance of each sample across all agronomic traits, and the bar chart is used to identify specific differences between the sample and the ideal plant type, combined with... Figure 2 and Figure 3 As shown, radar charts and bar charts are used to display the differences between the sample (number BN1115) and the ideal plant type in various traits, marking significant traits and intuitively analyzing key traits.

[0050] In other embodiments of this application, the server is further configured to perform principal component analysis to extract the first principal component and the second principal component based on the tobacco agronomic trait sample, and to perform cluster analysis to obtain category labels, and to visualize the clustering results in a scatter plot with different colors.

[0051] Specifically, the server performs principal component analysis (PCA) on tobacco agronomic trait samples to extract the first principal component (PC1) and the second principal component (PC2). Then, it obtains the scores for each sample on PC1 and PC2, performs cluster analysis on the PCA scores, selects a suitable clustering algorithm (such as K-means based on Euclidean distance), and clusters the two-dimensional PCA scores. The clustering results are visualized by plotting a scatter plot, with the horizontal axis representing the first principal component (PC1) and the vertical axis representing the second principal component (PC2). Each point represents a tobacco agronomic trait sample. Based on the category labels obtained from clustering, samples in different clusters are marked with different colors, such as... Figure 4 As shown in the figure. In this embodiment, PCA analysis is combined to perform dimensionality reduction on agronomic trait data and extract the most representative principal components, which can intuitively reflect the overall distribution and differences of the samples, providing scientific and efficient technical support for tobacco variety breeding and planting management.

[0052] like Figure 5 As shown, in another embodiment of this application, a method for collecting tobacco agronomic traits and analyzing and evaluating ideal plant type is also provided. This method is implemented based on the aforementioned tobacco agronomic trait collection and ideal plant type analysis and evaluation system. The tobacco agronomic trait collection and ideal plant type analysis and evaluation system includes: an Internet of Things (IoT) device, a mobile terminal, and a server. The method includes: S101. Use Internet of Things (IoT) devices to collect tobacco agronomic trait data and transmit them to mobile terminals; S102. Using a mobile terminal, preprocess the tobacco agronomic trait data to obtain tobacco agronomic trait samples, and transmit the tobacco agronomic trait samples to the server. S103. Using the server, construct sample vectors and sample vector sets based on tobacco agronomic trait samples. Then, using the robust Mahalanobis distance method, calculate the Mahalanobis distance of the sample vectors based on the sample vectors, sample vector sets, and pre-constructed ideal plant type vectors. Based on the Mahalanobis distance of the sample vectors, obtain the first analysis and evaluation results. S104. Using the server, based on the sample vector set, the robust mean and standard deviation are calculated using the minimum covariance determinant method. Based on the robust mean, standard deviation, and sample vectors, the robust value of the sample vectors is calculated. Based on the robust value of the sample vectors, the second analysis and evaluation results are obtained.

[0053] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A tobacco agronomic trait collection and ideal plant type analysis evaluation system, characterized by, The application relates to an Internet of Things device, a mobile terminal and a server end, and comprises the following steps: The Internet of Things device is used for collecting tobacco agronomic character data and transmitting the data to the mobile terminal; The mobile terminal is used for preprocessing the tobacco agronomic character data to obtain tobacco agronomic character samples and transmitting the tobacco agronomic character samples to the server end; The server end is used for constructing a sample vector and a sample vector set according to the tobacco agronomic character samples, and calculating the Mahalanobis distance of the sample vector according to the sample vector, the sample vector set and a pre-constructed ideal plant type vector by using a robust Mahalanobis distance method, and obtaining a first analysis and evaluation result according to the Mahalanobis distance of the sample vector; The server end is also used for calculating a robust mean and a standard deviation according to the sample vector set by using a minimum covariance determinant method, calculating a robust value of the sample vector according to the robust mean, the standard deviation and the sample vector, and obtaining a second analysis and evaluation result according to the robust value of the sample vector. The server end is also used for constructing the ideal plant type vector, and when the server end executes the construction of the ideal plant type vector, the server end is specifically used for:

2. The system as recited in claim 1, wherein, calculating a sample mean according to all the tobacco agronomic character samples to obtain a mean sample; constructing the ideal plant type vector according to the mean sample. The server end is also used for constructing the ideal plant type vector, and when the server end executes the construction of the ideal plant type vector, the server end is specifically used for:

3. The system as recited in claim 1, wherein, obtaining a prediction model corresponding to each preset agronomic character index in the sample vector, wherein the prediction model is specifically a Gaussian process regression model taking an environmental feature vector as input and taking a predicted value of the preset agronomic character index as output; obtaining a real-time environmental feature vector, inputting the real-time environmental feature vector into the prediction model corresponding to each preset agronomic character index to obtain a real-time predicted value of each preset agronomic character index; and constructing the ideal plant type vector according to the real-time predicted value of each preset agronomic character index. The calculation formula of the Mahalanobis distance is as follows:

4. The system as recited in claim 1, wherein, The calculation formula of the robust value is as follows: ； wherein, denotes the Mahalanobis distance, denotes the sample vector, denotes the ideal plant type vector, denotes the transpose operation, denotes the inverse covariance matrix calculated from the set of sample vectors by the minimum covariance determinant.

5. The system as recited in claim 4, wherein, When the server end executes the first analysis and evaluation result according to the Mahalanobis distance of the sample vector, the server end is specifically used for: ； wherein, denotes the robust value, denotes the robust mean value, denotes the standard deviation.

6. The system as recited in claim 1, wherein, comparing the Mahalanobis distance of the sample vector with a first preset threshold to obtain the first analysis and evaluation result; When the server end executes the second analysis and evaluation result according to the robust value of the sample vector, the server end is specifically used for: comparing the robust value of the sample vector with a second preset threshold to obtain the second analysis and evaluation result. When the mobile terminal executes the preprocessing of the tobacco agronomic character data to obtain the tobacco agronomic character samples, the mobile terminal is specifically used for:

7. The system as recited in claim 1, wherein, performing outlier rejection processing on the tobacco agronomic character data, filling in missing values by using an interpolation method, and then performing predetermined format conversion to obtain the tobacco agronomic character samples. ​ 8. The system as recited in claim 1, wherein, The server end is further configured to perform statistical analysis according to the tobacco agronomic trait sample, and obtain statistical indexes, wherein the statistical indexes include one or more of a mean value, a maximum value, a minimum value, a standard deviation, and a variance of each preset agronomic trait index.

9. The system of claim 8, wherein, The server end is further configured to generate a table display according to the statistical indexes, the Mahalanobis distance of the sample vector, and the robust value of the sample vector. The server end is further configured to generate a radar chart and a column chart display according to the sample vector and the ideal plant type vector. The server end is further configured to extract a first principal component and a second principal component by using principal component analysis according to the tobacco agronomic trait sample, perform cluster analysis to obtain a category label, and visualize the cluster result in a scatter plot with different colors.

10. A method for collecting and evaluating agronomic traits of tobacco plants and ideal plant type, characterized by, The tobacco agronomic trait collection and ideal plant type analysis and evaluation system according to any one of claims 1-9 is implemented, and the system comprises an Internet of Things device, a mobile terminal, and a server end, and the method comprises: The tobacco agronomic trait data is collected by using the Internet of Things device and transmitted to the mobile terminal. The tobacco agronomic trait data is preprocessed by using the mobile terminal to obtain a tobacco agronomic trait sample, and the tobacco agronomic trait sample is transmitted to the server end. The server end is configured to construct a sample vector and a sample vector set according to the tobacco agronomic trait sample, and calculate the Mahalanobis distance of the sample vector by using a robust Mahalanobis distance method according to the sample vector, the sample vector set, and a pre-constructed ideal plant type vector, and obtain a first analysis and evaluation result according to the Mahalanobis distance of the sample vector. The server end is configured to calculate a robust mean value and a standard deviation by using a minimum covariance determinant method according to the sample vector set, calculate the robust value of the sample vector according to the robust mean value, the standard deviation, and the sample vector, and obtain a second analysis and evaluation result according to the robust value of the sample vector.

Images