Soil nutrient cyclic utilization method and system based on prediction driving

Abstract

The invention discloses a soil nutrient recycling method and system based on prediction driving, and belongs to the technical field of agricultural soil management.The method specifically comprises the steps that soil nutrient content data are obtained through sampling, and soil nutrient abnormal areas are preliminarily screened in combination with a geographic information system and a remote sensing technology; soil physicochemical properties are measured in the soil nutrient abnormal area, soil nutrient form distribution is measured through X-ray diffraction and near infrared spectroscopic analysis, and nutrient stability and release potential are evaluated; constructing a soil nutrient cyclic dynamic model based on multi-source data, and simulating nutrient evolution; according to simulation results, crop requirements, soil states and environmental factors, an accurate fertilization scheme is formulated; intelligent management and precise fertilization of soil nutrients are realized, and the fertilizer utilization rate and the agricultural production efficiency are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural soil management, and specifically relates to a soil nutrient recycling method and system based on prediction drive. Background Art

[0002] In agricultural production, the rational management of soil nutrients is crucial to the growth and yield of crops. Traditional soil nutrient management methods often rely on experience and regular soil testing, and there are problems such as inaccurate fertilization timing and low nutrient utilization efficiency. On the one hand, excessive fertilization not only causes waste of resources, but may also lead to soil pollution and eutrophication of water bodies; on the other hand, insufficient fertilization will affect the growth and development of crops and reduce yield and quality. With the development of information technology and data analysis technology, how to use these technologies to achieve precise management and recycling of soil nutrients has become a research hotspot in the current agricultural field. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention proposes a prediction-driven soil nutrient recycling method and system, which obtains soil nutrient content data through sampling, and preliminarily screens nutrient abnormal areas in combination with geographic information systems and remote sensing technology; measures soil physical and chemical properties in abnormal areas, and uses X-ray diffraction and near-infrared spectroscopy to measure soil nutrient form distribution, and evaluates nutrient stability and release potential; constructs a soil nutrient cycle dynamic model based on multi-source data to simulate nutrient evolution; formulates a precise fertilization plan based on simulation results and crop requirements, soil conditions and environmental factors; realizes intelligent management of soil nutrients and precise fertilization, and improves fertilizer utilization and agricultural production efficiency.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A soil nutrient recycling method based on prediction drive, comprising:

[0006] Step S1: obtaining soil nutrient content data by sampling, combining geographic information system and remote sensing technology, preliminarily screening out abnormal soil nutrient areas, and marking them;

[0007] Step S2: measuring the physical and chemical properties of the soil in the screened soil nutrient abnormality area; the physical and chemical properties of the soil include soil pH value, redox potential, organic matter content, and microbial activity;

[0008] Step S3: Determine the morphological distribution of nutrients in the soil, analyze the occurrence state of mineral-bound nutrients in the soil using an X-ray diffraction method, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients by combining a near-infrared spectroscopy analysis method, and draw a soil nutrient morphological distribution map;

[0009] Step S4: construct a soil nutrient cycle dynamic model to simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies;

[0010] Step S5: According to the simulation results of the soil nutrient cycle dynamic model, multi-source data is integrated to build an intelligent fertilization decision support system, and according to the output results of the intelligent fertilization decision support system, a precise fertilization plan is formulated, including the type, amount, time and method of fertilization.

[0011] Specifically, the specific steps of step S1 include:

[0012] S1.1: Conduct soil sampling in the study area, conduct laboratory analysis on the collected soil samples, and determine the nutrient content of the soil samples;

[0013] S1.2: Soil nutrient content data of sampling points Import the geographic information system, establish a soil nutrient database, and perform mean statistics on the soil nutrient content data in the soil nutrient database. , set the normal range threshold of nutrient content ,in, represents the soil nutrient content data of the i-th sampling point, and , n represents the total number of sampling points, represents the average value of soil nutrient content. represents the standard deviation of soil nutrient content data;

[0014] S1.3: Obtain remote sensing image data of the study area from the remote sensing data platform, use a polynomial model to perform geometric correction, and extract spatial distribution features of the geometrically corrected remote sensing image data; the spatial distribution features include statistical features and spatial autocorrelation features; the statistical features refer to mean and variance; the spatial autocorrelation features refer to coefficient of variation;

[0015] S1.4: In the geographic information system, the spatial distribution characteristics of soil nutrient content data and remote sensing images are superimposed and analyzed to identify areas where nutrient content deviates from the normal range;

[0016] S1.5: Perform Kriging interpolation on the identified areas where the nutrient content deviates from the normal range to generate refined abnormal areas;

[0017] S1.6: Use the map-making function of the geographic information system to geo-tag the refined abnormal areas and generate a distribution map of soil nutrient abnormal areas.

[0018] Specifically, the specific steps of S1.3 include:

[0019] S1.31: Obtain remote sensing image data of the study area from the remote sensing data platform, select corresponding feature points on the remote sensing image and reference map of the study area, and record the coordinates of the feature points on the remote sensing image and reference map of the study area to form coordinate pairs of control points; the feature points refer to points where the grayscale value of the image changes dramatically;

[0020] S1.32: Based on the coordinate pairs of the control points, use the polynomial correction model to correct the coordinates of the original remote sensing image data of the study area. Perform correction to obtain the corrected remote sensing image coordinates ;

[0021] S1.33: Determine the size and position of the target regular grid based on the corrected image range and resolution.

[0022] Specifically, the specific steps of S1.3 also include:

[0023] S1.34: For each target grid point, select the original 4×4 remote sensing image pixels of the study area around it as the interpolation window and use the cubic convolution weight function Calculate the weight of each corrected remote sensing image pixel of the study area, where d represents the distance from the target grid point to the original remote sensing image pixel of the study area within the interpolation window;

[0024] S1.35: Combined with corrected remote sensing image coordinates And the pixel value of the remote sensing image after correction , calculate the interpolation value of each target grid point, and combine the interpolation values ​​of all target grid points into a resampled remote sensing image, where, and Respectively represent the horizontal and vertical index values;

[0025] S1.36: Calculate the spatial distribution characteristics of the resampled remote sensing image data, where the spatial distribution characteristics include mean, variance, and coefficient of variation.

[0026] Specifically, the specific steps of S1.5 include:

[0027] S1.51: Collect soil sampling point data of the area where the nutrient content deviates from the normal range, wherein the soil sampling point data includes the coordinates of the sampling points and soil nutrient content data;

[0028] S1.52: According to the spatial autocorrelation and semivariogram analysis of soil sampling point data, set the Kriging interpolation parameters, input the soil sampling point data and Kriging interpolation parameters into the geographic information system software, perform the Kriging interpolation operation, and obtain the predicted nutrient content of the unsampled point p ;

[0029] S1.53: Based on the results of Kriging interpolation, a spatial distribution map of nutrient content is generated, and areas where the nutrient content exceeds the normal range are marked as abnormal areas, thereby obtaining refined abnormal areas.

[0030] Specifically, the specific steps of step S3 include:

[0031] S3.1: Collect soil samples at regular grid locations within the study area and record the geographical location, soil type, and environmental conditions of the sampling points;

[0032] S3.2: Air-dry, grind and sieve the soil samples, and use chemical extraction to determine the different forms of nutrients in the treated soil samples;

[0033] S3.3: Divide the measured soil samples into diffraction analysis soil samples and spectral analysis soil samples, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients based on the diffraction analysis soil samples and spectral analysis soil samples;

[0034] S3.4: Organize the results of X-ray diffraction and near-infrared spectroscopy analysis, and use the geographic information system to import the organized data and draw a soil nutrient form distribution map.

[0035] Specifically, the specific steps of S3.3 include:

[0036] S3.31: Analyze soil samples for diffraction;

[0037] Setting the working parameters of the X-ray diffractometer and conducting an X-ray diffraction experiment; the working parameters of the X-ray diffractometer include the type of X-ray source, tube voltage and tube current, scanning range, scanning speed and step size;

[0038] Collect diffraction pattern data and identify the mineral types in the diffraction analysis soil samples through the position and intensity of the diffraction peaks;

[0039] Combining X-ray diffraction results with chemical extraction data, the occurrence state of mineral-bound nutrients was determined, and at the same time, a release kinetic model of mineral-bound nutrients was constructed;

[0040] The release potential of mineral-bound nutrients was evaluated by the release kinetic model parameters of mineral-bound nutrients;

[0041] S3.32: Spectroscopic analysis of soil samples;

[0042] Use a near infrared spectrometer to collect NIR spectral data of soil samples for spectral analysis;

[0043] The partial least squares method was used to establish a quantitative model of NIR spectrum and organic matter-bound nitrogen content;

[0044] The stability of organic matter-bound nitrogen was evaluated based on its content and spectral characteristics.

[0045] Specifically, the soil nutrient cycle dynamic model in step S4 is constructed based on nutrient abnormality area data, soil physical and chemical properties data, nutrient form and release potential data, specifically soil nutrient content data, soil physical and chemical properties data, combined with real-time collected meteorological data, land use data and fertilization data as input variable data, and nutrient dynamic changes as output variables.

[0046] A soil nutrient recycling system based on prediction drive, including: a data acquisition module, a soil physical and chemical property determination module, a nutrient form distribution module, a model building module, and a fertilization plan formulation module;

[0047] The data acquisition module is used to obtain soil nutrient content data, soil physical and chemical property data and environmental data;

[0048] The soil physical and chemical property measurement module is used to measure the physical and chemical properties of the soil in the nutrient abnormality area;

[0049] The nutrient form distribution module is used to measure the form distribution of nutrients in the soil and evaluate the stability and release potential of nutrients;

[0050] The model building module is used to build a soil nutrient cycle dynamic model based on multi-source data to simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies;

[0051] The fertilization plan formulation module is used to formulate a precise fertilization plan based on the simulation results of the soil nutrient cycle dynamic model, taking into account crop needs, soil nutrient status and environmental factors.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The present invention proposes a soil nutrient recycling system based on prediction drive, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.

[0054] 2. The present invention proposes a prediction-driven soil nutrient recycling method, which can efficiently and accurately identify areas with abnormal soil nutrients by combining geographic information systems, remote sensing technology and physical and chemical property measurements, and deeply analyze soil nutrient morphology and stability, which not only improves the accuracy of soil nutrient assessment, but also optimizes the spatial allocation of nutrient resources.

[0055] 3. The present invention proposes a soil nutrient recycling method based on prediction-driven, constructs a soil nutrient cycle dynamic model and an intelligent fertilization decision support system, which can simulate the nutrient evolution under different conditions and intelligently formulate precise fertilization plans, thereby promoting the effective utilization of soil nutrients and environmental protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of a soil nutrient recycling method based on prediction drive of the present invention;

[0057] Figure 2 This is a principle flow chart of a soil nutrient recycling method based on prediction drive of the present invention;

[0058] Figure 3 This is a diagram of the architecture of a prediction-driven soil nutrient recycling system according to the present invention. DETAILED DESCRIPTION

[0059] Example 1

[0060] See also Figure 1 and Figure 2 The present invention provides an embodiment: a soil nutrient recycling method based on prediction drive, comprising the following steps:

[0061] Step S1: Obtain soil nutrient content data through sampling, combine Geographic Information System (GIS) and remote sensing technology to preliminarily screen out abnormal soil nutrient areas and mark them;

[0062] Step S2: measuring the physical and chemical properties of the soil in the screened soil nutrient abnormality area; the physical and chemical properties of the soil include soil pH value, redox potential, organic matter content, and microbial activity;

[0063] Step S3: Determine the morphological distribution of nutrients in the soil, analyze the occurrence state of mineral-bound nutrients in the soil using an X-ray diffraction method, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients by combining a near-infrared spectroscopy analysis method, and draw a soil nutrient morphological distribution map;

[0064] It is important to understand that organic-bound nitrogen and mineral-bound nutrients are two important forms of nutrients in the soil, and they play different roles in soil nutrient cycling and plant absorption. Organic-bound nitrogen refers to nitrogen elements in the soil that are combined with organic matter, such as humus, plant residues, microbial residues, etc. This nitrogen mainly exists in the form of organic compounds, such as proteins, amino acids, and nucleic acids; while mineral-bound nutrients refer to nutrients in the soil that are combined with minerals, such as clay minerals and silicate minerals, mainly including phosphorus, potassium, calcium, and magnesium. These nutrients exist in the soil in the form of minerals, such as apatite and potassium feldspar, as shown in Table 1.

[0065] Table 1 Comparison of organic matter-bound nitrogen and mineral-bound nutrients

[0066]

[0067] As can be seen from Table 1, organic-bound nitrogen is nitrogen elements combined with soil organic matter, which needs to be converted into inorganic nitrogen absorbable by plants through microbial mineralization; mineral-bound nutrients are nutrient elements combined with soil minerals, which need to be released into effective nutrients absorbable by plants through chemical weathering or dissolution processes; both are important nutrient reserves in the soil, which can achieve slow release and sustainable utilization of nutrients by adjusting soil environmental conditions and microbial activity.

[0068] Step S4: construct a soil nutrient cycle dynamic model to simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies;

[0069] The soil nutrient cycle dynamic model in step S4 is constructed based on nutrient abnormality area data, soil physical and chemical property data, nutrient form and release potential data, specifically, soil nutrient content data, soil physical and chemical property data, combined with real-time collected meteorological data, land use data and fertilization data as input variable data, and nutrient dynamic changes as output variables.

[0070] Furthermore, the specific steps of step S4 include:

[0071] (1) Obtain soil nutrient content data and soil physical and chemical property data. At the same time, collect meteorological data, land use data and fertilization data, and perform cleaning and normalization processing;

[0072] (2) The goal of the model is to simulate the dynamic evolution of soil nutrients, including the input, transformation, output, and circulation of nutrients. According to the goal of the model, the pre-built denitrification-decomposition (DNDC) model is loaded;

[0073] (3) Constructing a soil nutrient cycle dynamic model, which is based on the pre-constructed DNDC model and includes:

[0074] 1) Nutrient input module:

[0075] The simulation of nutrient input process, including fertilization, atmospheric deposition, and biological nitrogen fixation, is as follows: ,in, represents the nutrient input, represents the fertilization rate, C represents the fertilization rate;

[0076] 2) Nutrient conversion module:

[0077] Simulate the transformation process of nutrients in the soil, including mineralization, nitrification, denitrification, and fixation. The formula is: ,in, Represents the amount of mineralization, represents the mineralization rate constant, Indicates the organic nitrogen content;

[0078] 3) Nutrient output module:

[0079] The simulation of nutrient output process, including crop absorption, leaching, and volatilization, is as follows: ,in, Indicates the crop absorption, is the absorption coefficient, Indicates the available nitrogen content;

[0080] 4) Nutrient Cycle Module:

[0081] The simulation of nutrient circulation in the soil-plant system, including residue decomposition and root exudates, is as follows: ,in, represents the amount of decomposition, is the decomposition rate constant, Indicates the residual nitrogen content;

[0082] Among them, when constructing the dynamic model of soil nutrient cycle, the soil nutrient cycle process is decomposed into multiple modules, such as nutrient input module, nutrient conversion module, nutrient output module and nutrient cycle module, and a mathematical equation is established for each module to describe the nutrient change process to complete the construction of the dynamic model of soil nutrient cycle.

[0083] (4) Obtaining soil nutrient cycle dynamic model parameters such as mineralization rate constant, absorption coefficient, and decomposition rate constant based on data fitting, obtaining a constructed soil nutrient cycle dynamic model, and verifying the constructed soil nutrient cycle dynamic model using an independent data set; the independent data set is data that is not used for calibration;

[0084] (5) Designing different environmental conditions and fertilization strategy scenarios, wherein the fertilization strategy includes fertilizer amount and fertilization time;

[0085] (6) Run the model under each scenario to simulate the dynamic evolution of soil nutrients.

[0086] Step S5: According to the simulation results of the soil nutrient cycle dynamic model, an intelligent fertilization decision support system is constructed, and according to the output results of the intelligent fertilization decision support system, a precise fertilization plan is formulated, including the type, amount, time and method of fertilization.

[0087] Furthermore, the construction process of the intelligent fertilization decision support system includes:

[0088] (1) The user inputs soil nutrient content data, soil physical and chemical property data and crop growth environment data through the user interface; the crop growth environment data includes meteorological data, land use data and fertilization data;

[0089] (2) The system cleans, integrates, and standardizes the input data to generate processed model input data;

[0090] (3) Loading the soil nutrient cycle dynamic model, and using the soil nutrient cycle dynamic model to analyze and predict the processed model input data;

[0091] (4) Based on the analysis and prediction results, design the overall architecture of the system, including input module, processing module, and output module, and develop a user interface to ensure that users can easily input data and view results;

[0092] (5) Test the system and optimize and improve it based on the test results to obtain an intelligent fertilization decision support system;

[0093] (6) Deploy the intelligent fertilization decision support system into the target user’s environment.

[0094] Furthermore, the process of formulating a precision fertilization plan includes:

[0095] (1) Input the current soil nutrient content data, soil physical and chemical properties data, and crop growth environment data into the intelligent fertilization decision support system;

[0096] (2) The intelligent fertilization decision support system automatically generates fertilization recommendations based on input conditions, including fertilizer type, fertilizer amount, and fertilization time;

[0097] (3) Optimize the generated fertilization suggestions according to actual needs to form the final precise fertilization plan;

[0098] (4) Apply fertilizers according to the precise fertilization plan, regularly monitor crop growth and soil nutrient changes, and adjust the fertilization plan in a timely manner based on the feedback results.

[0099] The soil nutrient recycling method based on prediction driving also includes:

[0100] Step S6: Regularly monitor soil nutrient status, crop growth status and environmental indicators to evaluate the effectiveness of fertilization programs;

[0101] Step S7: According to the evaluation results, the soil nutrient cycle dynamic model and fertilization plan are optimized to improve the accuracy and adaptability of the system.

[0102] The specific steps of step S1 include:

[0103] S1.1: Conduct soil sampling in the study area, conduct laboratory analysis on the collected soil samples, and determine the nutrient content of the soil samples;

[0104] S1.2: Soil nutrient content data of sampling points Import the geographic information system, establish a soil nutrient database, and perform mean statistics on the soil nutrient content data in the soil nutrient database. , set the normal range threshold of nutrient content ,in, represents the soil nutrient content data of the i-th sampling point, and , n represents the total number of sampling points, represents the average value of soil nutrient content. represents the standard deviation of soil nutrient content data;

[0105] Furthermore, the specific steps of S1.2 include:

[0106] (1) Collect soil nutrient content data from all sampling points to ensure the integrity and accuracy of the data;

[0107] (2) Open the GIS software and select the file import method to import the soil nutrient content data of the sampling points into the GIS;

[0108] (3) In GIS, according to the characteristics and requirements of soil nutrient content data, design the database structure, including table structure, field type, index, create a database, and store the imported soil nutrient content data in the corresponding table;

[0109] (4) In GIS, use SQL query statements to select the nutrient fields that need to be counted and calculate the average value of each field. ;

[0110] (5) Generate a mean statistical report based on the statistical results to intuitively understand the overall distribution of soil nutrients;

[0111] (6) Based on the mean statistical results, combined with historical data and industry standards, set the normal range threshold of nutrient content ,in, It reflects the fluctuation of data points relative to the mean;

[0112] (7) Record the set threshold into the database.

[0113] It should be noted that according to normal distribution theory, 95% of the data points will fall within the range of the mean plus or minus two standard deviations. As a normal range threshold, it can cover the vast majority of data points, thus providing a relatively reliable nutrient content range, and can fully reflect the overall distribution of soil nutrients, reducing misjudgments caused by data fluctuations.

[0114] S1.3: Obtain remote sensing image data of the study area from the remote sensing data platform, use a polynomial model to perform geometric correction, and extract spatial distribution features of the geometrically corrected remote sensing image data; the spatial distribution features include statistical features and spatial autocorrelation features; the statistical features refer to mean and variance; the spatial autocorrelation features refer to coefficient of variation;

[0115] S1.4: In the geographic information system, the spatial distribution characteristics of soil nutrient content data and remote sensing images are superimposed and analyzed to identify areas where nutrient content deviates from the normal range;

[0116] Furthermore, the specific steps of S1.4 include:

[0117] (1) Load soil nutrient content data and remote sensing image spatial distribution characteristic data into GIS software to ensure that the spatial range, resolution and projection of the two are consistent. If they are inconsistent, resampling is required;

[0118] (2) Use GIS spatial analysis tools, such as overlay analysis tools, to overlay soil nutrient content data with remote sensing image spatial distribution characteristic data;

[0119] (3) Combined with the normal range threshold of nutrient content , extracting areas where nutrient content deviates from the normal range. For example, if the soil nitrogen content is lower than the normal range threshold, it is marked as a nitrogen-deficient area, and if the soil nitrogen content is higher than the normal range threshold, it is marked as a nitrogen-excess area.

[0120] S1.5: Perform Kriging interpolation on the identified areas where the nutrient content deviates from the normal range to generate refined abnormal areas;

[0121] S1.6: Use the map-making function of the geographic information system to geo-tag the refined abnormal areas and generate a distribution map of soil nutrient abnormal areas.

[0122] The specific steps of S1.3 include:

[0123] S1.31: Obtain remote sensing image data of the study area from the remote sensing data platform, select corresponding feature points on the remote sensing image and reference map of the study area, and record the coordinates of the feature points on the remote sensing image and reference map of the study area to form coordinate pairs of control points; the feature points refer to points where the grayscale value of the image changes dramatically;

[0124] S1.32: Use polynomial correction model based on coordinate pairs of control points The original remote sensing image data coordinates of the study area Perform correction to obtain the corrected remote sensing image coordinates ,in, and represents polynomial coefficients;

[0125] S1.33: Determine the size and position of the target regular grid according to the corrected image range and resolution;

[0126] S1.34: For each target grid point, select the original 4×4 remote sensing image pixels of the study area around it as the interpolation window and use the cubic convolution weight function Calculate the weight of each corrected remote sensing image pixel of the study area, where d represents the distance from the target grid point to the original remote sensing image pixel of the study area within the interpolation window, and c represents a constant;

[0127] S1.35: Combined with corrected remote sensing image coordinates And the pixel value of the remote sensing image after correction , calculate the interpolation value of each target grid point, and the calculation formula of the interpolation value of the target grid point is: , and combine the interpolation values ​​of all target grid points into a resampled remote sensing image, where and Respectively represent the horizontal and vertical index values, represents the interpolated pixel value of the remote sensing image, represents the weight function;

[0128] S1.36: Calculate the spatial distribution characteristics of the resampled remote sensing image data, the spatial distribution characteristics including mean, variance, and coefficient of variation, wherein the mean, variance, and coefficient of variation are prior art contents in this field, are not the inventive solutions of this application, and are not elaborated here.

[0129] The specific steps of S1.5 include:

[0130] S1.51: Collect soil sampling point data of the area where the nutrient content deviates from the normal range, wherein the soil sampling point data includes the coordinates of the sampling points and soil nutrient content data;

[0131] S1.52: According to the spatial autocorrelation and semivariogram analysis of soil sampling point data, set the Kriging interpolation parameters, input the soil sampling point data and Kriging interpolation parameters into the geographic information system software, perform the Kriging interpolation operation, and obtain the predicted nutrient content of the unsampled point p ,in, represents the predicted value of soil nutrient content at unsampled points, Indicates sampling point The weight coefficient of Indicates sampling point The soil nutrient content value of the sampling point is n, because the total number of sampling points is n. The number of is n, which can also be understood as the name corresponding to the i-th sampling point is , i is the sampling point index value;

[0132] Furthermore, the sampling point The weight coefficient The solution formula is: ,in, Indicates sampling point and The semivariogram value between , represents the Lagrange multiplier, represents the unsampled point p and the sampled point The semivariogram values ​​between .

[0133] And the semivariogram formula is: ,in, represents the semivariogram, represents the nugget value, A represents the base value, It represents the distance between two sampling points, and r represents the range.

[0134] S1.53: Based on the results of Kriging interpolation, a spatial distribution map of nutrient content is generated, and areas where the nutrient content exceeds the normal range are marked as abnormal areas, thereby obtaining refined abnormal areas.

[0135] The specific steps of step S3 include:

[0136] S3.1: Collect soil samples at regular grid locations within the study area and record the geographical location, soil type, and environmental conditions of the sampling points;

[0137] S3.2: Air-dry, grind and sieve the soil samples, and use chemical extraction to determine the different forms of nutrients in the treated soil samples, including inorganic, organic and mineral-bound forms;

[0138] Among them, the chemical extraction method adopts a continuous extraction method, and commonly used extractants include water, dilute acid, and chelating agents. The continuous extraction method is the existing technical content in this field and is not an inventive solution of this application, so it will not be described in detail here;

[0139] S3.3: Divide the measured soil samples into diffraction analysis soil samples and spectral analysis soil samples, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients based on the diffraction analysis soil samples and spectral analysis soil samples;

[0140] S3.4: Organize the results of X-ray diffraction and near-infrared spectroscopy analysis, and use the geographic information system to import the organized data and draw a soil nutrient form distribution map.

[0141] The specific steps of S3.3 include:

[0142] S3.31: Analyze soil samples for diffraction;

[0143] Setting the working parameters of the X-ray diffractometer and performing an X-ray diffraction experiment; the working parameters of the X-ray diffractometer include the type of X-ray source, tube voltage and tube current, scanning range, scanning speed and step length, wherein the X-ray diffraction experiment is a prior art content in this field, is not an inventive solution of the present application, and is not described in detail here;

[0144] Collect diffraction pattern data, and identify the mineral types in the diffraction analysis soil sample through the position and intensity of the diffraction peak. The present invention mainly matches the identified peak value with the diffraction peak of the known mineral to confirm the mineral types that may exist in the sample;

[0145] Combining X-ray diffraction results with chemical extraction data, we determined the occurrence state of mineral-bound nutrients and constructed a release kinetic model for mineral-bound nutrients. ,in, represents the nutrient release at time t, represents the maximum release amount, l represents the release rate constant, and e represents the exponent;

[0146] The release potential of mineral-bound nutrients was evaluated by the release kinetic model parameters of mineral-bound nutrients;

[0147] S3.32: Spectroscopic analysis of soil samples;

[0148] Use a near infrared spectrometer to collect NIR spectral data of soil samples for spectral analysis;

[0149] Establishment of a quantitative model for NIR spectra and organic matter-bound nitrogen content using partial least squares method , where o represents the organic matter-bound nitrogen content, and D represents the regression coefficient. represents NIR spectral data, represents the error term, and the partial least squares method is the prior art content in this field, which is not the inventive solution of the present application and will not be described in detail here;

[0150] The stability of organic matter-bound nitrogen was evaluated based on its content and spectral characteristics.

[0151] Example 2

[0152] See also Figure 3 Another embodiment provided by the present invention is a soil nutrient recycling system based on prediction drive, comprising:

[0153] Data collection module, soil physical and chemical properties determination module, nutrient form distribution module, model building module, and fertilization plan formulation module;

[0154] Data acquisition module, used to obtain soil nutrient content data, soil physical and chemical properties data and environmental data;

[0155] The soil physical and chemical properties measurement module is used to measure the physical and chemical properties of the soil in the nutrient abnormality area, providing a basis for evaluating soil fertility;

[0156] Nutrient form distribution module, used to determine the form distribution of nutrients in the soil and evaluate the stability and release potential of nutrients;

[0157] Model building module, which is used to build a soil nutrient cycle dynamic model based on multi-source data and simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies;

[0158] The fertilization plan formulation module is used to formulate precise fertilization plans based on the simulation results of the soil nutrient cycle dynamic model, comprehensive crop needs, soil nutrient status and environmental factors.

[0159] The data acquisition module includes: soil sampling unit, nutrient abnormal area identification unit, and pre-processing unit;

[0160] Soil sampling unit, which obtains soil samples through field sampling for determination of nutrient content;

[0161] The nutrient anomaly area identification unit is used to combine geographic information system and remote sensing technology to preliminarily screen out nutrient anomaly areas and mark them;

[0162] The pre-processing unit is used to clean, integrate and format the collected data for subsequent analysis.

[0163] The soil physical and chemical property measurement module includes: pH value measurement unit, oxidation-reduction potential measurement unit, organic matter content measurement unit, and microbial activity measurement unit;

[0164] pH measurement unit, used to measure the acidity and alkalinity of the soil;

[0165] Redox potential measurement unit, used to measure the redox state of the soil;

[0166] Organic matter content determination unit, used to determine the content of organic matter in soil;

[0167] Microbial activity assay unit, used to assess the activity level of microorganisms in soil.

[0168] The nutrient form distribution module includes: X-ray diffraction analysis unit and near infrared spectrum analysis unit;

[0169] X-ray diffraction analysis unit, used to analyze the occurrence state of mineral-bound nutrients in the soil;

[0170] Near infrared spectroscopy unit to assess the stability of organically bound nitrogen and the release potential of mineral bound nutrients.

[0171] The model building module includes: a data integration unit and a model building unit;

[0172] Data integration unit, used to integrate data on nutrient anomaly areas, soil physical and chemical properties, nutrient forms and release potential;

[0173] The model building unit is used to build a soil nutrient cycle dynamic model using mathematical models and algorithms.

[0174] The fertilization plan formulation module includes: demand analysis unit, fertilization strategy formulation unit, and fertilization plan optimization unit;

[0175] Demand analysis unit, used to analyze the nutrient requirements of crops;

[0176] Fertilization strategy formulation unit, used to formulate the type, amount, time and method of fertilization based on simulation results and demand analysis;

[0177] The fertilization scheme optimization unit is used to further optimize the fertilization scheme to ensure environmental protection.

[0178] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.

Claims

1. A soil nutrient recycling method based on prediction drive, characterized in that: include: Step S1: obtaining soil nutrient content data by sampling, combining geographic information system and remote sensing technology, preliminarily screening out abnormal soil nutrient areas, and marking them; Step S2: measuring the physical and chemical properties of the soil in the screened soil nutrient abnormality area; the physical and chemical properties of the soil include soil pH value, redox potential, organic matter content, and microbial activity; Step S3: Determine the morphological distribution of nutrients in the soil, analyze the occurrence state of mineral-bound nutrients in the soil using an X-ray diffraction method, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients by combining a near-infrared spectroscopy analysis method, and draw a soil nutrient morphological distribution map; Step S4: construct a soil nutrient cycle dynamic model to simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies; Step S5: According to the simulation results of the soil nutrient cycle dynamic model, multi-source data is integrated to build an intelligent fertilization decision support system, and according to the output results of the intelligent fertilization decision support system, a precise fertilization plan is formulated, including the type, amount, time and method of fertilization.

2. The method for soil nutrient recycling based on prediction drive according to claim 1, characterized in that: The specific steps of step S1 include: S1.1: Conduct soil sampling in the study area, conduct laboratory analysis on the collected soil samples, and determine the nutrient content of the soil samples; S1.2: Soil nutrient content data of sampling points Import the geographic information system, establish a soil nutrient database, and perform mean statistics on the soil nutrient content data in the soil nutrient database. , set the normal range threshold of nutrient content ,in, represents the soil nutrient content data of the i-th sampling point, and , n represents the total number of sampling points, represents the average value of soil nutrient content. represents the standard deviation of soil nutrient content data; S1.3: Obtain remote sensing image data of the study area from the remote sensing data platform, use a polynomial model to perform geometric correction, and extract spatial distribution features of the geometrically corrected remote sensing image data; the spatial distribution features include statistical features and spatial autocorrelation features; the statistical features refer to mean and variance; the spatial autocorrelation features refer to coefficient of variation; S1.4: In the geographic information system, the spatial distribution characteristics of soil nutrient content data and remote sensing images are superimposed and analyzed to identify areas where nutrient content deviates from the normal range; S1.5: Perform Kriging interpolation on the identified areas where the nutrient content deviates from the normal range to generate refined abnormal areas; S1.6: Use the map-making function of the geographic information system to geo-tag the refined abnormal areas and generate a distribution map of soil nutrient abnormal areas.

3. A method for soil nutrient recycling based on prediction drive as claimed in claim 2, characterized in that: The specific steps of S1.3 include: S1.31: Obtain remote sensing image data of the study area from the remote sensing data platform, select corresponding feature points on the remote sensing image and reference map of the study area, and record the coordinates of the feature points on the remote sensing image and reference map of the study area to form coordinate pairs of control points; the feature points refer to points where the grayscale value of the image changes dramatically; S1.32: Based on the coordinate pairs of the control points, use the polynomial correction model to correct the coordinates of the original remote sensing image data of the study area. Perform correction to obtain the corrected remote sensing image coordinates ; S1.33: Determine the size and position of the target regular grid based on the corrected image range and resolution.

4. The method for soil nutrient recycling based on prediction drive according to claim 3, characterized in that: The specific steps of S1.3 also include: S1.34: For each target grid point, select the original 4×4 remote sensing image pixels of the study area around it as the interpolation window and use the cubic convolution weight function Calculate the weight of each corrected remote sensing image pixel of the study area, where d represents the distance from the target grid point to the original remote sensing image pixel of the study area within the interpolation window; S1.35: Combined with corrected remote sensing image coordinates And the pixel value of the remote sensing image after correction , calculate the interpolation value of each target grid point, and combine the interpolation values ​​of all target grid points into a resampled remote sensing image, where, and Respectively represent the horizontal and vertical index values; S1.36: Calculate the spatial distribution characteristics of the resampled remote sensing image data, where the spatial distribution characteristics include mean, variance, and coefficient of variation.

5. A method for soil nutrient recycling based on prediction drive as claimed in claim 4, characterized in that: The specific steps of S1.5 include: S1.51: Collect soil sampling point data of the area where the nutrient content deviates from the normal range, wherein the soil sampling point data includes the coordinates of the sampling points and soil nutrient content data; S1.52: According to the spatial autocorrelation and semivariogram analysis of soil sampling point data, set the Kriging interpolation parameters, input the soil sampling point data and Kriging interpolation parameters into the geographic information system software, perform the Kriging interpolation operation, and obtain the predicted nutrient content of the unsampled point p ; S1.53: Based on the results of Kriging interpolation, a spatial distribution map of nutrient content is generated, and areas where the nutrient content exceeds the normal range are marked as abnormal areas, thereby obtaining refined abnormal areas.

6. A method for soil nutrient recycling based on prediction drive as claimed in claim 5, characterized in that: The specific steps of step S3 include: S3.1: Collect soil samples at regular grid locations within the study area and record the geographical location, soil type, and environmental conditions of the sampling points; S3.2: Air-dry, grind and sieve the soil samples, and use chemical extraction to determine the different forms of nutrients in the treated soil samples; S3.3: Divide the measured soil samples into diffraction analysis soil samples and spectral analysis soil samples, and evaluate the stability of organic matter-bound nitrogen and the release potential of mineral-bound nutrients based on the diffraction analysis soil samples and spectral analysis soil samples; S3.4: Organize the results of X-ray diffraction and near-infrared spectroscopy analysis, and use the geographic information system to import the organized data and draw a soil nutrient form distribution map.

7. The method for soil nutrient recycling based on prediction drive according to claim 6, characterized in that: The specific steps of S3.3 include: S3.31: Analyze soil samples for diffraction; Setting the working parameters of the X-ray diffractometer and conducting an X-ray diffraction experiment; the working parameters of the X-ray diffractometer include the type of X-ray source, tube voltage and tube current, scanning range, scanning speed and step size; Collect diffraction pattern data and identify the mineral types in the diffraction analysis soil samples through the position and intensity of the diffraction peaks; Combining X-ray diffraction results with chemical extraction data, the occurrence state of mineral-bound nutrients was determined, and at the same time, a release kinetic model of mineral-bound nutrients was constructed; The release potential of mineral-bound nutrients was evaluated by the release kinetic model parameters of mineral-bound nutrients; S3.32: Spectroscopic analysis of soil samples; Use a near infrared spectrometer to collect NIR spectral data of soil samples for spectral analysis; The partial least squares method was used to establish a quantitative model of NIR spectrum and organic matter-bound nitrogen content; The stability of organic matter-bound nitrogen was evaluated based on its content and spectral characteristics.

8. The method for soil nutrient recycling based on prediction drive according to claim 7, characterized in that: The soil nutrient cycle dynamic model in step S4 is constructed based on nutrient abnormality area data, soil physical and chemical property data, nutrient form and release potential data, specifically, soil nutrient content data, soil physical and chemical property data, combined with real-time collected meteorological data, land use data and fertilization data as input variable data, and nutrient dynamic changes as output variables.

9. A prediction-driven soil nutrient recycling system, which is used to implement a prediction-driven soil nutrient recycling method according to any one of claims 1 to 8, characterized in that: include: Data collection module, soil physical and chemical properties determination module, nutrient form distribution module, model building module, and fertilization plan formulation module; The data acquisition module is used to obtain soil nutrient content data, soil physical and chemical property data and environmental data; The soil physical and chemical property measurement module is used to measure the physical and chemical properties of the soil in the nutrient abnormality area; The nutrient form distribution module is used to measure the form distribution of nutrients in the soil and evaluate the stability and release potential of nutrients; The model building module is used to build a soil nutrient cycle dynamic model based on multi-source data to simulate the dynamic evolution of soil nutrients under different environmental conditions and fertilization strategies; The fertilization plan formulation module is used to formulate a precise fertilization plan based on the simulation results of the soil nutrient cycle dynamic model, taking into account crop needs, soil nutrient status and environmental factors.

Images