Scientific fertilization method based on detection of soil condition and microbial action

Abstract

The invention relates to the technical field of precision agriculture, in particular to a scientific fertilization method based on detection of soil conditions and microbial action, which comprises the following steps: collecting soil impedance and microbial indexes, correcting microbial distribution by using impedance depth, and executing environmental inhibition attenuation on nitrogen fixation and phosphorus solubilization rate by combining a carbon source and acid-base buffer capacity. And quantifying biological available nutrient supply, and generating a fertilizer discharge signal based on a crop demand gap and mechanical parameters. According to the method, the vertical cone index and the microbial gene abundance are collected, physical impedance data are introduced to correct the microbial spatial distribution weight, the metabolic activation model is constructed in combination with active organic carbon and the acid-base buffer capacity, and the actual nutrient conversion efficiency and the effective fertilizer supply total amount of nitrogen-fixing bacteria and phosphate solubilizing bacteria are quantified; biological fertilizer supply capacity is brought into a nutrient income and expenditure balance system, a discrete fertilizer discharge control instruction is generated according to actual gaps of crops and mechanical operation parameters, and the utilization rate of chemical fertilizer and agricultural ecological benefits are improved.

Description

Technical Field

[0001] This invention relates to the field of precision agriculture technology, and in particular to a scientific fertilization method based on detecting soil conditions and microbial activity. Background Technology

[0002] Precision agriculture technology involves comprehensive agricultural management strategies that combine modern information technology with agronomic principles. Its core components include the collection of farmland geographic information data, monitoring of soil physicochemical properties, diagnosis of crop growth status, and precise management of agricultural inputs. The aim is to acquire spatiotemporal variation information of farmland through GPS, GIS, remote sensing technology, and field sensor networks, thereby enabling refined control of agricultural operations such as tillage, sowing, fertilization, irrigation, and pest and disease control. Traditional scientific fertilization methods rely on calculating... In agricultural production, fertilizer application mainly involves developing fertilization plans based on the nutritional needs of crops for nitrogen, phosphorus, potassium, and micronutrients at different growth stages. Generally, soil samples are collected using the five-point field sampling method. The content of total nitrogen, available phosphorus, available potassium, and organic matter in the soil is determined by chemical titration or colorimetric methods. Based on the nutrient balance method or field fertilizer efficiency test results, the fertilizer application limit standard table is found. The specific dosage of single-element fertilizers or compound fertilizers such as urea, superphosphate, and potassium chloride is calculated by combining the fertilizer nutrient content. The fertilizer is then applied to the topsoil layer of the farmland by manual spreading or mechanical quantitative fertilizer application.

[0003] Current scientific fertilization methods mainly rely on discrete chemical sampling and static nutrient balance calculations, focusing on assessing soil chemical nutrient reserves while neglecting the hindering effect of soil physical compaction on root penetration and nutrient transport. They fail to effectively quantify the potential of soil microbial ecosystems to dynamically provide bioavailable nutrients, and formulate plans based solely on chemical measurements, ignoring the contribution of biological nutrient supply, resulting in redundant fertilizer inputs. Furthermore, the lack of consideration for physical impedance limitations leads to low nutrient absorption and conversion rates, making it difficult to achieve precise matching of nutrient supply and demand in complex soil environments. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a scientific fertilization method based on detecting soil conditions and microbial activity, comprising the following steps: S1: Using a hydraulic cone penetrometer and stratified soil sampling equipment, vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes are obtained. Line integrals and exponential weighting of the vertical cone index records are used to generate an impedance environment dataset. S2: Call the impedance environment dataset, calculate the cumulative impedance path based on the vertical bioimpedance depth, construct the anti-impedance weighting factor, spatially interpolate the absolute abundance of nitrogen-fixing bacteria genes and the absolute abundance of phosphate-solubilizing bacteria genes, and generate a microbial distribution matrix. S3: Call the microbial distribution matrix and the impedance environment dataset, calculate the ratio of soil active organic carbon content to absolute abundance of nitrogen-fixing bacteria genes, compare it with the microbial metabolic activation benchmark, and attenuate the standard nitrogen fixation conversion coefficient to obtain nitrogen fixation correction data and generate nitrogen fixation correction records. S4: Call the microbial distribution matrix and the impedance environment dataset, correct the standard phosphorus solubilization rate and integrate to obtain the total effective biological phosphorus solubilization, combine the nitrogen fixation correction record to obtain the total effective biological nitrogen fixation, and compile them to generate a biological fertilizer supply dataset. S5: Call the biological fertilizer supply dataset, calculate the difference between crop nutrient requirements and the total effective biological nitrogen fixation and total effective biological phosphorus solubilization, obtain the net fertilizer supply requirement, and map it into a fertilizer discharge control signal based on the operating width parameters and driving speed parameters of the fertilizer application machinery.

[0005] As a further aspect of the present invention, the impedance environment dataset includes vertical bioimpedance depth, active organic carbon content, soil pH buffer capacity, and absolute abundance of microbial genes; the microbial distribution matrix includes spatial distribution voxels of nitrogen-fixing bacteria, spatial distribution voxels of phosphate-solubilizing bacteria, and a spatial anti-impedance weight field; the nitrogen fixation correction record includes calculated carbon-bacterial ratio, nitrogen fixation conversion attenuation coefficient, and nitrogen fixation amount correction value; the bio-fertilizer supply dataset includes total effective biological nitrogen fixation, total effective biological phosphorus solubilization, and phosphorus solubilization efficiency damping coefficient; and the fertilizer discharge control signal includes net fertilizer supply demand, fertilizer discharge valve opening command, and variable fertilizer application execution parameters.

[0006] As a further aspect of the present invention, the step of obtaining the impedance environment dataset specifically includes: S101: In-situ testing and sampling analysis are performed using a hydraulic cone penetrometer and stratified soil sampling equipment to obtain vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes. Combined with sampling coordinates, full data alignment and standardization in the spatiotemporal dimensions are performed to generate a multidimensional set of basic soil detection elements. S102: Call the multidimensional soil foundation detection element set, perform continuous numerical line integration operation along the vertical depth direction on the vertical conic index record, identify the stress mutation interval in the integration path where the value exceeds the soil compaction resistance benchmark, perform nonlinear exponential penalty weight calculation for the stress mutation interval, map the linear physical geometric depth to the bioeffective impedance depth, and generate the vertical bioimpedance depth. S103: Call the multidimensional soil basic detection element set, associate and encapsulate the vertical bioimpedance depth with active organic carbon content, soil acid-base buffer capacity and absolute abundance of microbial genes, and generate an impedance environment dataset.

[0007] As a further aspect of the present invention, the process of obtaining the soil compaction resistance benchmark is specifically as follows: A negative correlation response model of the root elongation rate of the target crop to soil mechanical resistance is constructed. The critical resistance threshold when the root elongation rate decays to zero in the negative correlation response model is identified. The soil particle size distribution parameters of the test area are obtained. The internal friction angle correction coefficient based on the soil particle size distribution parameters is calculated. The mechanical strength correction is performed on the critical resistance threshold using the internal friction angle correction coefficient. The corrected value is set as the soil compaction resistance benchmark. The process of performing nonlinear exponential penalty weight calculation for stress mutation intervals is as follows: The difference between the average strength value recorded by the vertical conic index within the stress abrupt change interval and the soil compaction resistance benchmark is calculated to generate the resistance over-limit amplitude. An impedance enhancement index function with the natural constant as the base is constructed. The resistance over-limit amplitude is input into the impedance enhancement index function to solve the bioenergy consumption ratio of root penetration per unit soil layer. The product of the physical thickness of the stress abrupt change interval and the bioenergy consumption ratio is calculated to obtain the equivalent bioimpedance distance. The equivalent bioimpedance distances at all depth levels are accumulated to generate the vertical bioimpedance depth.

[0008] As a further aspect of the present invention, the step of obtaining the microbial distribution matrix specifically includes: S201: Call the impedance environment dataset, construct a three-dimensional voxelized spatial grid covering the target fertilization area, retrieve all voxel center points to be estimated and known sampling points within the grid, establish a straight vector path connecting each pair of points, perform discrete line integral accumulation operation on the spatial distribution field perpendicular to the bioimpedance depth along the straight vector path, and generate a cumulative impedance path tensor. S202: Call the accumulated impedance path tensor, perform a nonlinear reciprocal transformation on the accumulated path values, construct an inverse proportional mapping model between physical impedance and spatial correlation strength, calculate the weight contribution value of the sampling point to be estimated grid point, and perform normalization processing on the weight contribution value to generate an anti-impedance weight factor field. S203: Extract the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes based on the absolute abundance of microbial genes. Use the aforementioned anti-impedance weighting factor field to perform a weighted linear combination estimation of the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes based on impedance distance, reconstruct the global continuous field, and generate a microbial distribution matrix.

[0009] As a further aspect of the present invention, the step of obtaining the nitrogen fixation correction record specifically includes: S301: Call the microbial distribution matrix and the impedance environment dataset, extract the absolute abundance of nitrogen-fixing bacteria genes and the content of active organic carbon in the soil according to the global spatial grid index alignment, calculate the ratio of active organic carbon in the soil to the absolute abundance of nitrogen-fixing bacteria genes, and generate the carbon bacteria metabolic energy supply ratio. S302: Obtain the microbial metabolic activation benchmark, calculate the energy threshold difference between the carbon bacteria metabolic energy supply ratio and the microbial metabolic activation benchmark, construct a nonlinear negative exponential decay model based on the energy threshold difference, use the nonlinear negative exponential decay model to perform dynamic weighting operation on the preset standard nitrogen fixation conversion coefficient, obtain the actual biological nitrogen fixation rate, and generate nitrogen fixation correction data. S303: Analyze the spatial distribution characteristics of the nitrogen fixation correction data, establish a mapping table between the correction data and the original carbon source constraints, perform serialization encoding on the nitrogen fixation correction data of each grid point according to the spatial topological order, perform structured encapsulation with added metadata tags on the encoded data, and generate nitrogen fixation correction records.

[0010] As a further aspect of the present invention, the process for obtaining the microbial metabolic activation benchmark is specifically as follows: The standard ATP stoichiometry of nitrogenase-catalyzed reactions was obtained. Combined with the oxidative phosphorylation efficiency of heterotrophic bacteria, the theoretical carbon source mass required to maintain the minimum effective nitrogen fixation activity was calculated and set as the basic threshold for carbon source consumption per unit gene abundance. A negative correlation response function between soil mechanical resistance and effective porosity is constructed. The vertical bioimpedance depth is substituted into the negative correlation response function as the independent variable to solve the current effective soil porosity. The porosity tortuosity coefficient, which is inversely proportional to the effective soil porosity, is calculated to generate an effective diffusion resistance coefficient characterizing the substrate diffusion path extension effect. A metabolic energy consumption compensation model based on the effective diffusion resistance coefficient is constructed to quantify the additional active transport energy consumption required by microorganisms to maintain intracellular substrate flux and overcome the porosity coefficient. An environmental stress gain factor is generated, and the environmental stress gain factor is used to perform nonlinear multiplication correction on the carbon source consumption baseline threshold to generate the microbial metabolic activation benchmark.

[0011] As a further aspect of the present invention, the steps for obtaining the bio-fertilizer dataset are specifically as follows: S401: Call the impedance environment dataset and the microbial distribution matrix, calculate the phosphorus solubilization efficiency damping coefficient that characterizes the environmental inhibition effect based on the soil acid-base buffer capacity, combine the absolute abundance of phosphorus solubilization bacteria genes, and use the phosphorus solubilization efficiency damping coefficient to perform a correction calculation on the standard phosphorus solubilization rate to generate a corrected phosphorus solubilization rate. S402: Call the nitrogen fixation correction data in the nitrogen fixation correction record, combine it with the corrected phosphorus solubilization rate, determine the time span parameter of the crop growth cycle, perform time-domain integration operation on both along the time span parameter, accumulate and calculate the total effective nitrogen and total effective phosphorus released by microbial metabolism and transformation during the crop growth cycle, and generate effective total amount settlement data. S403: Extract the effective total settlement data, perform data alignment and structured assembly based on spatial grid indexing on the effective biological nitrogen fixation total and the effective biological phosphorus solubilization total to generate a biological fertilizer dataset.

[0012] As a further aspect of the present invention, the process for obtaining the standard phosphorus dissociation rate is specifically as follows: The metagenomic information contained in the microbial distribution matrix was analyzed to identify the inorganic phosphorus dissolution gene cluster encoding glucose dehydrogenase and the organic phosphorus mineralization gene cluster encoding phosphatase, and the gene copy number density of each gene cluster per unit soil volume was extracted. The standard catalytic kinetic constants of various phosphorus-solubilizing enzymes were obtained, ideal reaction boundary conditions of substrate saturation and no inhibitory factors were set, the maximum phosphorus conversion efficiency of monomeric gene products was calculated, and set as the theoretical phosphorus-solubilizing constant of monomers. The theoretical phosphorus solubilization constant of the monomer is used to perform a product operation on the gene copy number density to obtain the theoretical phosphorus production flux of different functional bacterial groups. The fluxes corresponding to the inorganic phosphorus dissolution gene cluster and the organic phosphorus mineralization gene cluster are accumulated to generate the standard phosphorus solubilization rate.

[0013] As a further aspect of the present invention, the step of acquiring the fertilizer discharge control signal specifically includes: S501: Call the biological fertilizer dataset, extract the total effective biological nitrogen fixation and the total effective biological phosphorus solubilization, and retrieve the target nutrient requirements for the entire growth period according to the crop type. Calculate the difference between the target nutrient requirements standard and the total effective biological nitrogen fixation and the total effective biological phosphorus solubilization, and generate the net fertilizer replenishment requirement. S502: Call the net fertilizer supply demand, read the operating width parameter of the fertilization machinery, use the operating width parameter of the fertilization machinery to define the effective coverage cross section of a single operation of the machinery, perform discretization grid reconstruction along the operating path direction on the spatial distribution field of the net fertilizer supply demand, divide it into independent fertilization control blocks aligned with the operating width of the machinery, and generate a width discretization fertilization mapping table. S503: Call the width discretized fertilization mapping table, monitor the machine's travel speed parameters in real time, calculate the dynamic fertilizer discharge rate at the current grid position, construct a time-domain coupled control model of fertilizer application amount, working width and travel speed, and generate fertilizer discharge control signals.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by collecting the vertical cone index and microbial gene abundance, physical impedance data is introduced to correct the spatial distribution weight of microorganisms. A metabolic activation model is constructed by combining active organic carbon and acid-base buffer capacity. The actual nutrient conversion efficiency and effective fertilizer supply of nitrogen-fixing bacteria and phosphate-solubilizing bacteria under specific environments are quantified. The biological fertilizer supply capacity is incorporated into the nutrient balance system. Discrete fertilizer discharge control instructions are generated based on the actual crop deficit and mechanical operation parameters to improve fertilizer utilization and agricultural ecological benefits. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0019] Please see Figure 1 This invention provides a scientific fertilization method based on detecting soil conditions and microbial activity, comprising the following steps: S1: Using a hydraulic cone penetrometer and stratified soil sampling equipment, vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes are obtained. Line integrals and exponential weighting of the vertical cone index records are used to generate an impedance environment dataset. S2: Call the impedance environment dataset, calculate the cumulative impedance path based on the vertical bioimpedance depth, construct the anti-impedance weighting factor, perform spatial interpolation on the absolute abundance of nitrogen-fixing bacteria genes and the absolute abundance of phosphate-solubilizing bacteria genes, and generate a microbial distribution matrix. S3: Call the microbial distribution matrix and impedance environment dataset, calculate the ratio of soil active organic carbon content to absolute abundance of nitrogen-fixing bacteria genes, compare it with the microbial metabolic activation benchmark and attenuate the standard nitrogen fixation conversion coefficient, obtain nitrogen fixation correction data, and generate nitrogen fixation correction records. S4: Call the microbial distribution matrix and impedance environment dataset, correct the standard phosphorus solubility rate and integrate to obtain the total effective biological phosphorus solubility, combine the nitrogen fixation correction record to obtain the total effective biological nitrogen fixation, and compile them to generate a biological fertilizer supply dataset. S5: Call the biological fertilizer supply dataset, calculate the difference between crop nutrient requirements and the total effective biological nitrogen fixation and total effective biological phosphorus solubilization, obtain the net fertilizer supply requirement, and map it into fertilizer discharge control signals based on the operating width and driving speed parameters of the fertilizer application machinery.

[0020] The impedance environment dataset includes vertical bioimpedance depth, active organic carbon content, soil pH buffer capacity, and absolute abundance of microbial genes. The microbial distribution matrix includes spatial distribution voxels of nitrogen-fixing bacteria, spatial distribution voxels of phosphate-solubilizing bacteria, and a spatial anti-impedance weight field. The nitrogen fixation correction records include calculated carbon-bacterial ratios, nitrogen fixation conversion attenuation coefficients, and corrected nitrogen fixation values. The bio-fertilizer supply dataset includes total effective biological nitrogen fixation, total effective biological phosphorus solubilization, and phosphorus solubilization efficiency damping coefficients. The fertilizer discharge control signals include net fertilizer supply demand, fertilizer discharge valve opening commands, and variable fertilization execution parameters.

[0021] Please see Figure 2 The specific steps for obtaining the impedance environment dataset are as follows: S101: In-situ testing and sampling analysis are performed using a hydraulic cone penetrometer and stratified soil sampling equipment to obtain vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes. Combined with sampling coordinates, full data alignment and standardization in the spatiotemporal dimensions are performed to generate a multidimensional set of basic soil detection elements. Using an all-terrain off-road chassis equipped with a GPS system as the vehicle, a gridded sampling path was planned within the target farmland area. A hydraulic cone penetrometer was vertically inserted into the soil at a constant penetration speed of 2 cm / s. Equipped with a high-precision pressure sensor, soil mechanical resistance data from 0 to 100 cm depth were continuously collected at a sampling frequency of 50 Hz, generating a vertical cone index record. Simultaneously, soil samples were extracted at 10 cm depths using a stratified soil sampling device. The soil samples were oxidized at 900°C using a dry-burning method to release carbon dioxide, and the active organic carbon content was determined using a non-dispersive infrared detector. A potentiometric titration method was used to add 0.05% oxidative stress solution dropwise to the soil sample suspension. A standard solution of 0.1 mol / L hydrochloric acid or sodium hydroxide was used to record pH change curves and calculate soil acid-base buffering capacity. Total soil DNA was extracted and real-time quantitative PCR was used to amplify the nifH and phoD genes using specific primers to determine the absolute abundance of microbial genes. Using sampling coordinates obtained through differential mapping technology, the four types of heterogeneous data were aligned spatiotemporally to construct a multidimensional data cube containing spatial three-dimensional coordinates and timestamps. Standardization was performed on the detection data for different dimensions using the Z-score standardization method to calculate the arithmetic mean and standard deviation of each detection index, and the original observed values ​​were then standardized. The arithmetic mean is divided by the standard deviation to eliminate dimensional differences. For example, if the original value of the vertical conic index at a certain measuring point is 2.5 MPa, the global average is 1.5 MPa, and the standard deviation is 0.5 MPa, the difference is 1.0 MPa after subtraction, and then the standardized value is 2.0 after division. Similarly, the active organic carbon content, soil pH buffer capacity, and absolute abundance of microbial genes are processed. All standardized data are structured and encapsulated with corresponding three-dimensional spatial coordinates and timestamps to generate a multidimensional soil basic detection element set.

[0022] S102: Call the multidimensional soil basic detection element set, perform continuous numerical line integration operation along the vertical depth direction on the vertical conic index record, identify the stress mutation interval in the integration path where the value exceeds the soil compaction resistance benchmark, perform nonlinear exponential penalty weight calculation for the stress mutation interval, map the linear physical geometric depth to the bioeffective impedance depth, and generate the vertical bioimpedance depth. The vertical conic index records from the multidimensional soil foundation testing feature set are retrieved. The integration step in the vertical depth direction is set to 1 mm. Numerical line integration is then performed on the standardized vertical conic index, i.e., the resistance value at each millimeter depth is multiplied by the sum. The depth increment quantifies the accumulated mechanical work during the vertical penetration of roots. During integration, the current depth's cone index value is compared in real-time with the soil compaction resistance benchmark. Depth intervals where the value consistently exceeds the benchmark are identified and marked as stress abrupt change intervals. The process for obtaining the soil compaction resistance benchmark is as follows: First, a negative correlation response model of the target crop's (e.g., maize) root elongation rate to soil mechanical resistance is constructed. This model describes the linear or non-linear decrease in root growth rate with increasing resistance. A controlled soil column experiment is used to determine the critical resistance threshold (e.g., 2.5 MPa) when the root elongation rate decays to zero. Then, the particle size distribution parameters of the soil in the test area are obtained (e.g., 25% clay content, 40% silt content, 35% sand content). The internal friction angle is calculated based on the Mohr-Coulomb failure criterion. The critical resistance threshold is then corrected for mechanical strength using the internal friction angle correction coefficient in the Terzaghi bearing capacity formula (e.g., correction coefficient = 1.2, then benchmark adjustment = 3.0 MPa). For stress abrupt change intervals, the arithmetic mean strength value recorded by the vertical cone index within that interval is calculated. A soil compaction resistance benchmark is established to generate the resistance overshoot value. An impedance enhancement exponential function is constructed with the natural constant e (approximately 2.718) as the base, and an empirical coefficient k (e.g., 0.5) is set. The resistance overshoot value is used as input to the exponential term to calculate the bioenergy consumption ratio of root penetration per unit soil layer. For example, when the interval average strength = 4.0 MPa and the benchmark = 3.0 MPa, the resistance overshoot value = 1.0 MPa, and the bioenergy consumption ratio = e. 0.5 That is, approximately 1.648; finally, the physical thickness of the stress abrupt change interval (e.g., 10 cm) × biological energy consumption ratio (16.48 cm) is calculated and defined as the equivalent bioimpedance distance. The equivalent bioimpedance distances of all depth levels in the entire profile are accumulated to map the physical linear geometric depth to the vertical bioimpedance depth reflecting biological physiological consumption, thus generating the vertical bioimpedance depth data of the measuring point.

[0023] S103: Call the multidimensional soil basic detection element set, associate and encapsulate the vertical bioimpedance depth with active organic carbon content, soil acid-base buffer capacity and microbial gene abundance to generate an impedance environment dataset. A composite data object is created containing five dimensions: spatial coordinates, physical properties, chemical properties, biological properties, and impedance characteristics. Active organic carbon content is mapped to the "chemical properties" dimension, soil pH buffering capacity is mapped to the "chemical properties" dimension, absolute abundance of microbial genes is mapped to the "biological properties" dimension, and vertical conic index records and their derived statistical characteristics are mapped to the "physical properties" dimension. Crucially, the calculated vertical bioimpedance depth is written as a core parameter into the "impedance characteristics" dimension. During this process, data integrity checks are performed to ensure that each sampling point contains all the above elements. For points with missing data, the neighborhood averaging method is used to fill in the gaps. Table 1 shows an example of the structure of a partially encapsulated impedance environment dataset. Finally, the composite data objects from all sampling points across the field are compiled into an impedance environment dataset, which is stored in HDF5 hierarchical data format, supporting efficient concurrent access and spatial analysis in subsequent steps.

[0024] Table 1. Example of Impedance Environment Dataset Structure

[0025] As shown in Table 1, the dataset integrates spatial location, physical impedance, and biochemical indicators, providing a basis for subsequent interpolation.

[0026] Please see Figure 3 The specific steps for obtaining the microbial distribution matrix are as follows: S201: Call the impedance environment dataset, construct a three-dimensional voxelized spatial grid covering the target fertilization area, retrieve all the center points of the voxels to be estimated and the known sampling points in the grid, establish a straight vector path connecting each pair of points, perform discrete line integral accumulation operation on the spatial distribution field of the vertical bioimpedance depth along the straight vector path, and generate the cumulative impedance path tensor. Using the impedance environment dataset, a three-dimensional voxelized spatial grid covering the target fertilization area is constructed based on the farmland boundary coordinates. The voxel resolution is set to 1 m × 1 m × 0.1 m, forming a globally discretized model. The center point of each voxel to be estimated within the grid is used as the target point, and all known sampling points in the dataset are retrieved as source points. For each pair of "source point-target point" combinations, a three-dimensional straight-line vector path connecting the two points is established. A ray tracing algorithm is used to determine all voxel units traversed by this straight-line vector path. The vertical bioresistance within each voxel unit is extracted along this straight-line vector path. For depth-resistance calculations, a discrete linear integral accumulation operation is performed. Specifically, the intercept length of the path within each voxel is calculated, multiplied by the corresponding vertical bioimpedance depth value for that voxel, and the product of all intercept segments is summed to generate the cumulative impedance path value connecting the source and target points. This value represents the total physical energy barrier that microorganisms or nutrients need to overcome to migrate and diffuse between two points. For example, if the path passes through 3 voxels with an intercept length of 1 meter each, and the impedance depth gradients of each voxel are 1.2, 1.5, and 1.8 respectively, then the cumulative impedance path value is 1.2 + 1.5 + 1.8 = 4.5 meters. All paired cumulative impedance path values ​​are organized into a high-dimensional tensor structure to generate the cumulative impedance path tensor.

[0027] S202: Call the cumulative impedance path tensor, perform a nonlinear reciprocal transformation on the cumulative path values, construct an inverse mapping model between physical impedance and spatial correlation strength, calculate the weight contribution value of the sampling point to be estimated grid point, and perform normalization processing on the weight contribution value to generate the inverse impedance weight factor field. The cumulative impedance path tensor is invoked, and a nonlinear reciprocal transformation is performed on each cumulative impedance path value to construct an inverse mapping model between physical impedance and spatial correlation strength. The specific calculation logic is as follows: A distance attenuation exponent p (e.g., 2.0) is set, and the cumulative impedance path value is used as the denominator. The exponentiation is then performed, and the reciprocal is taken to obtain the initial weight value. For example, when the cumulative impedance path value = 4.5 meters, the square of 4.5 equals 20.25, and the reciprocal is approximately 0.049. This calculation process reflects the physical law that the greater the impedance, the weaker the spatial correlation. To eliminate the influence of sample density differences, for each voxel center point to be estimated, the initial weight values ​​calculated with all known sampling points are collected, and the sum of these initial weight values ​​is calculated. This sum is then used to perform division normalization on each initial weight value, i.e., dividing the single initial weight value by the total weights to obtain the normalized inverse impedance weight factor. This factor ranges from 0 to 1, and the sum of all weight factors for the same estimated point is strictly equal to 1. 1; For example, if a point to be estimated is associated with only two sampling points, with initial weights of 0.049 and 0.151 respectively, and a total of 0.2, then the reverse impedance weight factor of the first sampling point is 0.049 / 0.2 = 0.245; finally, a reverse impedance weight factor field covering the entire grid is generated.

[0028] S203: Extract the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes based on the absolute abundance of microbial genes. Perform a weighted linear combination estimation based on impedance distance on the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes using an anti-impedance weighting factor field to reconstruct a global continuous field and generate a microbial distribution matrix. Based on the absolute abundance data of microbial genes, the absolute abundance of nitrogen-fixing bacteria genes (such as nifH) and phosphate-solubilizing bacteria genes (such as phoD and phoX) are extracted through gene sequence tag classification and used as attribute values ​​for interpolation. For each voxel to be estimated, a set of inverse impedance weighting factors corresponding to that voxel is retrieved from the inverse impedance weighting factor field, and the observed abundance values ​​of nitrogen-fixing and phosphate-solubilizing bacteria genes at the corresponding sampling points are retrieved from the impedance environment dataset. Weighted linear combination estimation based on impedance distance is performed: the observed gene abundance value at each sampling point is multiplied by its corresponding inverse impedance weighting factor to obtain a weighted component. Then, the weighted components of all sampling points are summed, and the result is the predicted gene abundance value of the voxel to be estimated. For example, if the abundance of nitrogen-fixing bacteria at sampling point A is 1000 units and the weight is 0.245, and the abundance at sampling point B is 2000 units and the weight is 0.755, then the predicted value for the voxel to be estimated is 1000 × 0.245 + 2000. × 0.755 = 245 + 1510 = 1755 units; This process reconstructs the global continuous distribution field, generating a microbial distribution matrix containing three-dimensional spatial coordinates, predicted abundance of nitrogen-fixing bacteria, and predicted abundance of phosphate-solubilizing bacteria, effectively solving the problem of accuracy deviation caused by neglecting soil physical barriers in traditional Euclidean distance interpolation.

[0029] Please see Figure 4 The specific steps for obtaining nitrogen fixation correction records are as follows: S301: Call the microbial distribution matrix and impedance environment dataset, extract the absolute abundance of nitrogen-fixing bacteria genes and soil active organic carbon content based on the global spatial grid index alignment, calculate the ratio of soil active organic carbon content to absolute abundance of nitrogen-fixing bacteria genes, and generate carbon bacteria metabolic energy supply ratio. The microbial distribution matrix and impedance environment dataset are used, and the two datasets are precisely aligned spatially using a global spatial grid index. For each voxel unit, the absolute abundance of nitrogen-fixing bacteria genes (unit: copies per gram of soil) and the soil active organic carbon content (unit: grams per kilogram) are extracted. A division operation is performed, and the carbon-bacterial metabolic energy supply ratio within each voxel is calculated by dividing the soil active organic carbon content by the absolute abundance of nitrogen-fixing bacteria genes. This ratio characterizes the level of carbon source energy supply that can be allocated to a unit number of nitrogen-fixing microorganisms. For example, when the active organic carbon content in a voxel is 12 grams per kilogram, the nitrogen-fixing bacteria gene abundance is 6 × 10⁻⁶. 6 When the copy number is per gram, for ease of calculation, the gene abundance can be converted to millions of copies (6 million), then the ratio = 12 / 6 = 2 micrograms of carbon per million copies; this value directly reflects the abundance of material basis for microbial-driven high-energy nitrogen fixation reaction, generating a data layer of carbon-based metabolic energy supply ratios across the entire domain.

[0030] S302: Obtain the microbial metabolic activation benchmark, calculate the energy threshold difference between the carbon bacteria metabolic energy supply ratio and the microbial metabolic activation benchmark, construct a nonlinear negative exponential decay model based on the energy threshold difference, use the nonlinear negative exponential decay model to perform dynamic weighting calculation on the preset standard nitrogen fixation conversion coefficient, obtain the actual biological nitrogen fixation rate, and generate nitrogen fixation correction data. Obtain the standard ATP stoichiometry for nitrogenase-catalyzed reactions (typically, reducing one molecule of nitrogen requires 16 ATP). Combined with the oxidative phosphorylation efficiency of heterotrophic bacterial communities (P / O ratio, e.g., 2.5), calculate the amount of glucose oxidation required to produce 16 ATP, convert it into carbon mass, and set it as the baseline threshold for carbon source consumption per unit gene abundance (e.g., 0.5 μg carbon per million copies). Construct a negative correlation response function (e.g., a linear decay function) between soil mechanical resistance and effective porosity, and substitute the vertical bioimpedance depth into the calculation of the current effective soil porosity. Based on the mass transfer principle of porous media, calculate the pore tortuosity coefficient, which is inversely proportional to porosity (e.g., tortuosity = 1 / porosity). 0.5 A metabolic energy consumption compensation model is constructed to calculate the additional energy consumption multiplier (environmental stress gain factor) required by microorganisms to actively transport along tortuous paths. For example, a doubling of tortuosity increases energy consumption by 1.5 times. This factor is used to perform a multiplicative correction on the basic threshold of carbon source consumption (e.g., 0.5 × 1.5 = 0.75) to generate the final microbial metabolic activation benchmark. Subsequently, the energy threshold difference between the carbon bacteria metabolic energy supply ratio and the microbial metabolic activation benchmark is calculated. A nonlinear negative exponential decay model based on the energy threshold difference is constructed, and a decay coefficient k (e.g., 0.8) is set. This model is used to perform dynamic weighting operations on a preset standard nitrogen fixation conversion coefficient (e.g., 0.1 μg of nitrogen fixed per million copies per day). If the energy supply ratio is lower than the benchmark, the conversion coefficient will decrease sharply according to an exponential law. For example, when the threshold difference = -0.5 and the decay coefficient k = 0.8, the correction coefficient = standard coefficient × e (0.8 × -0.5) =Standard coefficient × e -0.4 Finally, the actual biological nitrogen fixation rate of the voxel under the current energy and impedance constraints is obtained, and nitrogen fixation correction data is generated.

[0031] S303: Analyze the spatial distribution characteristics of nitrogen fixation correction data, establish a mapping table between correction data and original carbon source constraints, serialize and encode the nitrogen fixation correction data of each grid point according to spatial topological order, and encapsulate the encoded data with added metadata tags to generate nitrogen fixation correction records. The spatial distribution characteristics of nitrogen fixation correction data are analyzed, and a mapping table containing grid ID, spatial coordinates, and correction values ​​is constructed. Using spatial filling curve algorithms such as Hilbert curves or Z-order curves, the nitrogen fixation correction data of the 3D grid points is serialized and encoded according to spatial topological order, transforming the multidimensional array into a one-dimensional data stream to improve storage and retrieval efficiency. Metadata tags are added to each data node, including information such as computation time, the version of the impedance correction parameter used, and the carbon source threshold standard, and structured encapsulation is performed. For example, the record format of a certain grid point is "ID:1024,Pos:(10,20,5), Rate:0.08, Tag:v1.0_20240520". Finally, a nitrogen fixation correction record containing the nitrogen fixation capacity of all voxels in the entire domain is generated. This record not only stores the numerical values ​​but also retains computational traceability information, providing traceable biological nitrogen source data support for subsequent fertilization decisions.

[0032] Please see Figure 5 The specific steps for obtaining the bio-fertilizer dataset are as follows: S401: Call the impedance environment dataset and microbial distribution matrix, calculate the phosphorus solubilization efficiency damping coefficient that characterizes the environmental inhibition effect based on the soil acid-base buffer capacity, combine the absolute abundance of phosphorus solubilization bacteria genes, and use the phosphorus solubilization efficiency damping coefficient to perform a correction calculation on the standard phosphorus solubilization rate to generate the corrected phosphorus solubilization rate. Using the impedance environment dataset and microbial distribution matrix, soil pHBC data for each voxel was extracted. Based on the principles of chemical kinetics, the damping coefficient of phosphorus solubilization efficiency, characterizing the environmental inhibitory effect, was calculated. A damping function was constructed: damping coefficient = 1 / (1 + sensitivity coefficient × pHBC), where the sensitivity coefficient was set according to the sensitivity of phosphorus solubilizing enzymes to pH changes (e.g., 0.1). If pHBC = 50 mmol / kg, then the damping coefficient = 1 / (1 + 0.1 × 50) = 1 / 6 ≈ 0.167. This coefficient reflects the adsorption, fixation, or chemical precipitation of enzymatic reaction products by soils with high buffer capacity. Simultaneously, a standard phosphorus solubilization rate was obtained. The process involved: parsing metagenomic information, identifying the inorganic phosphorus solubilization gene cluster encoding glucose dehydrogenase (GCD) and the organic phosphorus mineralization gene cluster encoding phosphatase (phoD), and extracting their gene copy number density (e.g., GCD = 2000 units, phoD = ...). 3000 units); obtain the standard catalytic kinetic constants (kcat) of various enzymes, set the theoretical phosphorus solubilization constants of monomers under substrate saturation conditions (e.g., GCD = 0.05 μg P / unit / day, phoD = 0.03 μg P / unit / day); calculate the theoretical flux using multiplication (GCD: 2000 × 0.05 = 100; phoD: 3000 × 0.03 = 90), and sum them to obtain the standard phosphorus solubilization rate (190 μg P / day); finally, use the aforementioned phosphorus solubilization efficiency damping coefficient (0.167) to perform a multiplication correction calculation on the standard phosphorus solubilization rate (190), i.e., 190 × 0.167 = 31.73 μg P / day, to generate the corrected phosphorus solubilization rate.

[0033] S402: Call the nitrogen fixation correction data in the nitrogen fixation correction record, combine it with the corrected phosphorus solubilization rate, determine the time span parameter of the crop growth cycle, perform time-domain integration operation on both along the time span parameter, accumulate and calculate the total effective nitrogen and total effective phosphorus released by microbial metabolic transformation during the crop growth cycle, and generate effective total settlement data. The nitrogen fixation correction data (e.g., a certain voxel nitrogen fixation rate = 0.08 μg N / day) and the corrected phosphorus solubilization rate generated in step S401 (e.g., 31.73 μg P / day) are retrieved from the nitrogen fixation correction record. Combining this with the target crop's growth cycle parameters (e.g., maize's entire growth period = 120 days), the growth cycle is divided into several time steps (e.g., one step per day). Time-domain integration is performed on both along the time span parameter, i.e., the metabolic output of each day is accumulated. Considering the fluctuation of microbial activity with seasonal temperature changes, a temperature influence factor is introduced to fine-tune the daily rate (e.g., a summer coefficient of 1.2, a spring / autumn coefficient of 0.8). Here, a simplified example of constant rate integration is used: Total effective biological nitrogen fixation = 0.08 × 120 = 9.6 μg N (per unit microvolume), Total effective biological phosphorus solubilization = 31.73 × 120 = 3807.6 micrograms of phosphorus; Extending the calculation results of microscopic volume to a unit area (e.g., hectares), the total effective nitrogen and phosphorus released by microbial metabolic transformation during the crop growth cycle are generated, i.e., the total effective amount settlement data.

[0034] S403: Extract the effective total settlement data, perform data alignment and structured assembly based on spatial grid indexing on the effective biological nitrogen fixation total and the effective biological phosphorus solubilization total to generate a biological fertilizer supply dataset; Effective total amount settlement data was extracted, and the effective biological nitrogen fixation and effective biological phosphorus solubilization were aligned using a unified spatial grid indexing system. A data structure for the bio-fertilizer supply dataset was constructed, containing the coordinate information of each grid cell, the total biological nitrogen fixation value, the total biological phosphorus solubilization value, and key correction factors used in the calculation process (such as damping coefficients). Structured assembly was performed to integrate the scattered calculation results into a single database file or data stream; for example, the record for grid A is {N_total: 15 kg / ha, P_total: 5 kg / ha}, and the record for grid B is {N_total: 18 kg / ha, P_total: 6 kg / ha}. This dataset comprehensively quantifies the effective nutrient baseline that the soil microbial ecosystem can provide during the crop growing season, providing a scientific basis for accurately deducting fertilizer application.

[0035] Please see Figure 6 The specific steps for obtaining the fertilizer discharge control signal are as follows: S501: Call the bio-fertilizer dataset, extract the total effective biological nitrogen fixation and total effective biological phosphorus solubilization, retrieve the target nutrient requirements for the entire growth period according to crop type, calculate the difference between the target nutrient requirements standard and the total effective biological nitrogen fixation and total effective biological phosphorus solubilization, and generate the net fertilizer replenishment requirement. The system retrieves the total effective biological nitrogen fixation and effective biological phosphorus solubility from the bio-fertilizer dataset for each grid in the field. Based on the preset crop type (e.g., maize) and its target yield (e.g., 10 tons / hectare), it searches the agricultural knowledge base to obtain the target nutrient requirements for the entire growth period (e.g., nitrogen requirement 250 kg / hectare, phosphorus requirement 90 kg / hectare). For each grid cell, it performs subtraction operations: Net fertilizer supply requirement (nitrogen) = Target nitrogen requirement - Total effective biological nitrogen fixation; Net fertilizer supply requirement (phosphorus) = Target phosphorus requirement - Total effective biological phosphorus solubility. For example, if a grid provides 50 kg of nitrogen through biological fixation and 20 kg of phosphorus through biological solubility, then the net nitrogen supply for that grid = 250 - 50 = 200 kg / hectare, and the net phosphorus supply = 90 - 20 = 200 kg / hectare. 70 kg / ha; if the calculation result is negative, set it to zero, indicating that the biological fertilizer supply has met the demand; this step generates a distribution map of the net fertilizer supply demand for the entire field.

[0036] S502: Call the net fertilizer supply demand, read the operating width parameter of the fertilizer application machinery, use the operating width parameter of the fertilizer application machinery to define the effective coverage cross section of a single operation of the machinery, perform discretization grid reconstruction along the operation path direction on the spatial distribution field of the net fertilizer supply demand, divide it into independent fertilizer control blocks aligned with the operating width of the machinery, and generate a width discretization fertilizer mapping table. The system retrieves the net fertilizer requirement distribution map and reads the operating width parameters of the fertilization machinery (e.g., width = 12 meters). Using this width parameter, the entire field is divided into several parallel mechanical operating strips. Within each operating strip, the continuous spatial grid is reorganized along the machinery's travel direction into a series of independent fertilization control blocks. The width of each block equals the machinery's width, and its length depends on the control system's response frequency and travel speed (e.g., set to 5 meters). For each fertilization control block, the arithmetic mean of the net fertilizer requirement of all original small grids within its coverage area is calculated as the unified target fertilization amount for that block. For example, a block covering 10 small grids has nitrogen fertilizer requirements ranging from 190 to 210 kg / ha, with an average of 200 kg / ha. A width-discrete fertilization mapping table is generated, which clarifies the specific fertilization standards that the machinery should implement during each segment of its journey in the field, achieving a match between agronomical requirements and the machinery's operational capabilities.

[0037] S503: Call the width discretized fertilization mapping table, monitor the machine's travel speed parameters in real time, calculate the dynamic fertilizer discharge rate at the current grid position, construct a time-domain coupled control model of fertilizer application amount, working width and travel speed, and generate fertilizer discharge control signals; The discrete fertilization mapping table is invoked, and the machine's travel speed parameters (e.g., 2 m / s) are monitored in real time by wheel speed sensors or GPS receivers installed on the tractor. Based on the physical flow formula, the dynamic fertilizer discharge rate at the current grid position is calculated. Fertilizer discharge rate (kg / s) = target fertilizer application rate (kg / m²) × operating width (m) × travel speed (m / s). Assuming the target fertilizer application rate is 200 kg / hectare (i.e., 0.02 kg / m²), the operating width is 12 m, and the speed is 2 m / s, then the fertilizer discharge rate = 0.02 × 12 × 2 = 0.48 kg / s. A time-domain coupled control model of fertilizer application rate, operating width, and travel speed is constructed, mapping the calculated fertilizer discharge rate to the rotational speed command of the fertilizer discharge motor or the opening command of the hydraulic valve. Table 2 shows the mapping relationship of the fertilizer discharge control signal at different speeds. Finally, a fertilizer discharge control signal containing pulse width modulation (PWM) duty cycle or voltage value is generated to drive the actuator to accurately discharge fertilizer.

[0038] Table 2 Example of Fertilizer Control Signal Mapping

[0039] As shown in Table 2, the system dynamically adjusts the action of the actuator according to the speed to ensure a constant amount of fertilizer applied per unit area.

[0040] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A scientific fertilization method based on detecting soil conditions and microbial activity, characterized in that, Includes the following steps: S1: Using a hydraulic cone penetrometer and stratified soil sampling equipment, vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes are obtained. Line integrals and exponential weighting of the vertical cone index records are used to generate an impedance environment dataset. S2: Call the impedance environment dataset, calculate the cumulative impedance path based on the vertical bioimpedance depth, construct the anti-impedance weighting factor, spatially interpolate the absolute abundance of nitrogen-fixing bacteria genes and the absolute abundance of phosphate-solubilizing bacteria genes, and generate a microbial distribution matrix. S3: Call the microbial distribution matrix and the impedance environment dataset, calculate the ratio of soil active organic carbon content to absolute abundance of nitrogen-fixing bacteria genes, compare it with the microbial metabolic activation benchmark, and attenuate the standard nitrogen fixation conversion coefficient to obtain nitrogen fixation correction data and generate nitrogen fixation correction records. S4: Call the microbial distribution matrix and the impedance environment dataset, correct the standard phosphorus solubilization rate and integrate to obtain the total effective biological phosphorus solubilization, combine the nitrogen fixation correction record to obtain the total effective biological nitrogen fixation, and compile them to generate a biological fertilizer supply dataset. S5: Call the biological fertilizer supply dataset, calculate the difference between crop nutrient requirements and the total effective biological nitrogen fixation and total effective biological phosphorus solubilization, obtain the net fertilizer supply requirement, and map it into a fertilizer discharge control signal based on the operating width parameters and driving speed parameters of the fertilizer application machinery.

2. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 1, characterized in that, The impedance environment dataset includes vertical bioimpedance depth, active organic carbon content, soil pH buffer capacity, and absolute abundance of microbial genes. The microbial distribution matrix includes spatial distribution voxels of nitrogen-fixing bacteria, spatial distribution voxels of phosphate-solubilizing bacteria, and a spatial anti-impedance weight field. The nitrogen fixation correction record includes calculated carbon-bacterial ratio, nitrogen fixation conversion attenuation coefficient, and nitrogen fixation amount correction value. The bio-fertilizer supply dataset includes total effective biological nitrogen fixation, total effective biological phosphorus solubilization, and phosphorus solubilization efficiency damping coefficient. The fertilizer discharge control signal includes net fertilizer supply demand, fertilizer discharge valve opening command, and variable fertilizer application execution parameters.

3. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 1, characterized in that, The specific steps for obtaining the impedance environment dataset are as follows: S101: In-situ testing and sampling analysis are performed using a hydraulic cone penetrometer and stratified soil sampling equipment to obtain vertical cone index records, active organic carbon content, soil acid-base buffer capacity, and absolute abundance of microbial genes. Combined with sampling coordinates, full data alignment and standardization in the spatiotemporal dimensions are performed to generate a multidimensional set of basic soil detection elements. S102: Call the multidimensional soil foundation detection element set, perform continuous numerical line integration operation along the vertical depth direction on the vertical conic index record, identify the stress mutation interval in the integration path where the value exceeds the soil compaction resistance benchmark, perform nonlinear exponential penalty weight calculation for the stress mutation interval, map the linear physical geometric depth to the bioeffective impedance depth, and generate the vertical bioimpedance depth. S103: Call the multidimensional soil basic detection element set, associate and encapsulate the vertical bioimpedance depth with active organic carbon content, soil acid-base buffer capacity and absolute abundance of microbial genes, and generate an impedance environment dataset.

4. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 3, characterized in that, The specific process for obtaining the soil compaction resistance benchmark is as follows: A negative correlation response model of the root elongation rate of the target crop to soil mechanical resistance is constructed. The critical resistance threshold when the root elongation rate decays to zero in the negative correlation response model is identified. The soil particle size distribution parameters of the test area are obtained. The internal friction angle correction coefficient based on the soil particle size distribution parameters is calculated. The mechanical strength correction is performed on the critical resistance threshold using the internal friction angle correction coefficient. The corrected value is set as the soil compaction resistance benchmark. The process of performing nonlinear exponential penalty weight calculation for stress mutation intervals is as follows: The difference between the average strength value recorded by the vertical conic index within the stress abrupt change interval and the soil compaction resistance benchmark is calculated to generate the resistance over-limit amplitude. An impedance enhancement index function with the natural constant as the base is constructed. The resistance over-limit amplitude is input into the impedance enhancement index function to solve the bioenergy consumption ratio of root penetration per unit soil layer. The product of the physical thickness of the stress abrupt change interval and the bioenergy consumption ratio is calculated to obtain the equivalent bioimpedance distance. The equivalent bioimpedance distances at all depth levels are accumulated to generate the vertical bioimpedance depth.

5. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 3, characterized in that, The specific steps for obtaining the microbial distribution matrix are as follows: S201: Call the impedance environment dataset, construct a three-dimensional voxelized spatial grid covering the target fertilization area, retrieve all voxel center points to be estimated and known sampling points within the grid, establish a straight vector path connecting each pair of points, perform discrete line integral accumulation operation on the spatial distribution field perpendicular to the bioimpedance depth along the straight vector path, and generate a cumulative impedance path tensor. S202: Call the accumulated impedance path tensor, perform a nonlinear reciprocal transformation on the accumulated path values, construct an inverse proportional mapping model between physical impedance and spatial correlation strength, calculate the weight contribution value of the sampling point to be estimated grid point, and perform normalization processing on the weight contribution value to generate an anti-impedance weight factor field. S203: Extract the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes based on the absolute abundance of microbial genes. Use the aforementioned anti-impedance weighting factor field to perform a weighted linear combination estimation of the absolute abundance of nitrogen-fixing bacteria genes and phosphate-solubilizing bacteria genes based on impedance distance, reconstruct the global continuous field, and generate a microbial distribution matrix.

6. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 5, characterized in that, The specific steps for obtaining the nitrogen fixation correction record are as follows: S301: Call the microbial distribution matrix and the impedance environment dataset, extract the absolute abundance of nitrogen-fixing bacteria genes and the content of active organic carbon in the soil according to the global spatial grid index alignment, calculate the ratio of active organic carbon in the soil to the absolute abundance of nitrogen-fixing bacteria genes, and generate the carbon bacteria metabolic energy supply ratio. S302: Obtain the microbial metabolic activation benchmark, calculate the energy threshold difference between the carbon bacteria metabolic energy supply ratio and the microbial metabolic activation benchmark, construct a nonlinear negative exponential decay model based on the energy threshold difference, use the nonlinear negative exponential decay model to perform dynamic weighting operation on the preset standard nitrogen fixation conversion coefficient, obtain the actual biological nitrogen fixation rate, and generate nitrogen fixation correction data. S303: Analyze the spatial distribution characteristics of the nitrogen fixation correction data, establish a mapping table between the correction data and the original carbon source constraints, perform serialization encoding on the nitrogen fixation correction data of each grid point according to the spatial topological order, perform structured encapsulation with added metadata tags on the encoded data, and generate nitrogen fixation correction records.

7. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 6, characterized in that, The process of obtaining the microbial metabolic activation benchmark is as follows: The standard ATP stoichiometry of nitrogenase-catalyzed reactions was obtained. Combined with the oxidative phosphorylation efficiency of heterotrophic bacteria, the theoretical carbon source mass required to maintain the minimum effective nitrogen fixation activity was calculated and set as the basic threshold for carbon source consumption per unit gene abundance. A negative correlation response function between soil mechanical resistance and effective porosity is constructed. The vertical bioimpedance depth is substituted into the negative correlation response function as the independent variable to solve the current effective soil porosity. The porosity tortuosity coefficient, which is inversely proportional to the effective soil porosity, is calculated to generate an effective diffusion resistance coefficient characterizing the substrate diffusion path extension effect. A metabolic energy consumption compensation model based on the effective diffusion resistance coefficient is constructed to quantify the additional active transport energy consumption required by microorganisms to maintain intracellular substrate flux and overcome the porosity coefficient. An environmental stress gain factor is generated, and the environmental stress gain factor is used to perform nonlinear multiplication correction on the carbon source consumption baseline threshold to generate the microbial metabolic activation benchmark.

8. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 6, characterized in that, The specific steps for obtaining the bio-fertilizer dataset are as follows: S401: Call the impedance environment dataset and the microbial distribution matrix, calculate the phosphorus solubilization efficiency damping coefficient that characterizes the environmental inhibition effect based on the soil acid-base buffer capacity, combine the absolute abundance of phosphorus solubilization bacteria genes, and use the phosphorus solubilization efficiency damping coefficient to perform a correction calculation on the standard phosphorus solubilization rate to generate a corrected phosphorus solubilization rate. S402: Call the nitrogen fixation correction data in the nitrogen fixation correction record, combine it with the corrected phosphorus solubilization rate, determine the time span parameter of the crop growth cycle, perform time-domain integration operation on both along the time span parameter, accumulate and calculate the total effective nitrogen and total effective phosphorus released by microbial metabolism and transformation during the crop growth cycle, and generate effective total amount settlement data. S403: Extract the effective total settlement data, perform data alignment and structured assembly based on spatial grid indexing on the effective biological nitrogen fixation total and the effective biological phosphorus solubilization total to generate a biological fertilizer supply dataset.

9. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 8, characterized in that, The process for obtaining the standard phosphorus desorption rate is as follows: The metagenomic information contained in the microbial distribution matrix was analyzed to identify the inorganic phosphorus dissolution gene cluster encoding glucose dehydrogenase and the organic phosphorus mineralization gene cluster encoding phosphatase, and the gene copy number density of each gene cluster per unit soil volume was extracted. The standard catalytic kinetic constants of various phosphorus-solubilizing enzymes were obtained, ideal reaction boundary conditions of substrate saturation and no inhibitory factors were set, the maximum phosphorus conversion efficiency of the monomer gene product was calculated, and set as the theoretical phosphorus-solubilizing constant of the monomer. The theoretical phosphorus solubility constant of the monomer is used to perform a product operation on the gene copy number density to obtain the theoretical phosphorus production flux of different functional bacterial groups. The fluxes corresponding to the inorganic phosphorus dissolution gene cluster and the organic phosphorus mineralization gene cluster are accumulated to generate the standard phosphorus solubility rate.

10. The scientific fertilization method based on detecting soil conditions and microbial activity according to claim 8, characterized in that, The specific steps for acquiring the fertilizer discharge control signal are as follows: S501: Call the biological fertilizer dataset, extract the total effective biological nitrogen fixation and the total effective biological phosphorus solubilization, and retrieve the target nutrient requirements for the entire growth period according to the crop type. Calculate the difference between the target nutrient requirements standard and the total effective biological nitrogen fixation and the total effective biological phosphorus solubilization, and generate the net fertilizer replenishment requirement. S502: Call the net fertilizer supply demand, read the operating width parameter of the fertilization machinery, use the operating width parameter of the fertilization machinery to define the effective coverage cross section of a single operation of the machinery, perform discretization grid reconstruction along the operating path direction on the spatial distribution field of the net fertilizer supply demand, divide it into independent fertilization control blocks aligned with the operating width of the machinery, and generate a width discretization fertilization mapping table. S503: Call the width discretized fertilization mapping table, monitor the machine's travel speed parameters in real time, calculate the dynamic fertilizer discharge rate at the current grid position, construct a time-domain coupled control model of fertilizer application amount, working width and travel speed, and generate fertilizer discharge control signals.

Images