A method for co-cultivation optimization based on spatial distribution feature analysis

Abstract

This invention relates to the field of agricultural data analysis, and particularly to a method for optimizing crop-livestock synergy based on spatial distribution feature analysis. The method includes: acquiring basic farmland grid data; extracting and compressing the topological convergence characteristics of closed depressions within the farmland grid to obtain a topological compression optimization factor; analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid to obtain a spatial infiltration relaxation recovery factor; performing nonlinear bounded coupling processing on the optimization factor to obtain an optimized dynamic absorption carrying capacity boundary; and substituting the optimized dynamic absorption carrying capacity boundary into a crop-livestock synergy spatial operation matching model for solution to obtain a liquid fertilizer spatial allocation scheme. This solves the problem that existing crop-livestock synergy spatial matching models, which only construct absorption carrying capacity boundaries based on two-dimensional area, struggle to identify local overload and allocation distortion of liquid fertilizer caused by the convergence of micro-topographic closed depressions and soil infiltration heterogeneity.

Description

Technical Field

[0001] This invention relates to the field of agricultural data analysis technology, and in particular to a method for optimizing crop-livestock synergy based on spatial distribution feature analysis. Background Technology

[0002] In the process of agricultural circular economy and resource utilization of livestock and poultry breeding waste, crop-livestock co-cultivation and returning waste to the field is an important method for manure disposal. This method typically involves spatially allocating liquid manure, biogas slurry, or semi-liquid organic fertilizer generated by farms according to the nutrient requirements of crops and the absorption capacity of surrounding land. This allows organic nutrients generated at the breeding end to be absorbed and utilized by the planting end, thereby reducing the input of external fertilizers and mitigating the environmental pressure caused by centralized manure discharge. To improve the efficiency of returning waste to the field, existing technologies usually abstract farms, temporary storage ponds, fertilizer outlets in pipeline networks, and farmland plots as supply and demand nodes. Spatial operation optimization methods such as graph theory matching, optimal transport, or mixed integer linear programming are then used to generate a spatial allocation scheme for liquid fertilizer with the goal of minimizing transportation distance, pipeline energy consumption, or overall scheduling cost.

[0003] In the aforementioned spatial operations optimization model, farmland grids, acting as nodes representing the demand for liquid fertilizer absorption, require pre-defined maximum absorption capacity constraints. Existing technologies typically divide the target farmland area into multiple two-dimensional grids and directly calculate the theoretical maximum absorption capacity based on the grid's planar area and the standard nutrient requirement per unit area of ​​the crop. This theoretical maximum absorption capacity is then used as the absorption capacity boundary of the farmland grid in the model solution. While this approach is computationally simple and easily integrated with optimization models such as mixed-integer linear programming, it essentially assumes the farmland grid to be a flat, homogeneous, and uniformly load-bearing ideal plane, failing to fully consider the physical characteristics of liquid fertilizer during actual surface application, including convergence, retention, and local enrichment driven by gravity. Especially in real farmland environments, the surface often exhibits micro-topographical undulations, locally closed depressions, shallow ditches, and low-lying convergence areas. Liquid fertilizer, after being applied to the surface, does not strictly distribute uniformly across a two-dimensional area but tends to concentrate in closed depressions or low-lying areas along the surface flow direction. When there is a clear closed depression within a certain farmland grid, even if the overall area of ​​the grid is large and the theoretical absorption capacity calculated based on the crop's fertilizer requirements does not exceed the standard upper limit, in actual implementation, liquid fertilizer may still concentrate at local low points, leading to problems such as excessively high local nutrient concentrations, crop burn, short-term waterlogging, and high-concentration nutrient leakage into groundwater; while relatively high areas within the grid may experience insufficient nutrient supply, resulting in inconsistencies between spatial allocation results and actual agronomic effects.

[0004] While some existing methods attempt to correct farmland carrying capacity by introducing spatial statistics such as average slope, elevation variance, or topographic relief, these global statistics primarily reflect the magnitude of topographic changes and are insufficient to identify closed topological structures within farmland grids that exhibit differences in catchment boundaries, lowest points, overflow points, and containment spaces. For example, a smooth unidirectional slope and an inwardly concave closed depression may have similar elevation variances or average slopes, but their impacts on liquid fertilizer surface flow and local convergence are entirely different. If such statistics are still used as the basis for correction, spatial management models may fail to accurately distinguish between truly high-risk convergence areas and ordinary slope areas, leading to distortion of dynamic absorption constraint boundaries. Furthermore, farmland soil media themselves exhibit spatial heterogeneity, with significant differences in saturated hydraulic conductivity at different locations. For certain low-lying areas, if the bottom soil has strong infiltration capacity, liquid fertilizer can be quickly channeled into deeper soil layers after a short period of accumulation, resulting in a relatively low risk of waterlogging and surface overload. However, if the bottom of the low-lying area is heavy or compacted soil, even a small accumulation can lead to continuous waterlogging and localized pollution risks. Existing spatial matching methods typically only use plot-average soil parameters, failing to express the spatial coupling relationship between low-lying accumulation locations and high- or low-permeability areas, easily leading to one-size-fits-all conservative constraints or over-allocation. Therefore, how to construct a dynamic absorption and carrying capacity boundary in the spatial operation matching model of integrated crop-livestock farming that can simultaneously reflect the topological accumulation effect of closed depressions and the spatial heterogeneity of soil infiltration is a pressing issue that needs to be addressed. Summary of the Invention

[0005] In view of this, the present invention aims to propose a crop-livestock synergy optimization method based on spatial distribution feature analysis, in order to solve the problem that the existing crop-livestock synergy spatial matching model only constructs the absorption and carrying capacity boundary based on two-dimensional area, which makes it difficult to identify the local overload and distribution distortion of liquid fertilizer caused by the convergence of micro-topographic closed depressions and soil infiltration heterogeneity.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0007] A method for optimizing crop-livestock co-cultivation based on spatial distribution characteristic analysis, the method comprising:

[0008] Step S1: Collect and preprocess multi-source spatial basic data of the target farmland area to obtain a basic dataset of farmland grids;

[0009] Step S2: Extract and compress the topological convergence features of closed depressions within the farmland grid to obtain the topological terrain compression optimization factor;

[0010] Step S3: By analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid, the spatial infiltration relaxation recovery factor is obtained;

[0011] Step S4: Obtain the optimized dynamic absorption bearing boundary by performing nonlinear bounded coupling processing on the topological terrain compression optimization factor and the spatial infiltration relaxation recovery factor;

[0012] Step S5: By substituting the optimized dynamic absorption carrying capacity boundary into the planting and breeding synergy spatial operation matching model for solution, the liquid fertilizer spatial allocation scheme is obtained.

[0013] Furthermore, the process of collecting and preprocessing multi-source spatial data of the target agricultural area to obtain a farmland grid basic dataset includes:

[0014] For the target farmland area to be allocated liquid fertilizer, the boundary range data of the target farmland area is obtained through a geographic information system, and the target farmland area is divided into two-dimensional spatial grids according to a preset grid scale to obtain multiple farmland grids. For any farmland grid, the corresponding planar area data is calculated based on the two-dimensional spatial boundary of the farmland grid. The surface elevation of the target farmland area is collected by UAV photogrammetry equipment or airborne lidar equipment to obtain original surface elevation point cloud data or original surface elevation image data covering the target farmland area. The original surface elevation point cloud data or original surface elevation image data is subjected to coordinate registration, noise point removal, spatial interpolation and rasterization processing to obtain high-precision digital elevation data corresponding to the spatial boundary of each farmland grid. The process involves: collecting soil spatial distribution attributes of the target farmland area through farmland IoT sensor arrays or multispectral remote sensing inversion methods to obtain spatial distribution data of soil saturated hydraulic conductivity corresponding to different spatial coordinates within the target farmland area; for any farmland grid, performing average statistical processing on the spatial distribution data of soil saturated hydraulic conductivity within the spatial boundary of the farmland grid to obtain the nominal average saturated hydraulic conductivity data corresponding to the farmland grid; obtaining the standard nutrient requirement per unit area of ​​crops within the target farmland area based on crop planting planning data and agronomic fertilization standard data for the target farmland area; and obtaining the maximum safe water accumulation depth of liquid fertilizer based on the surface application safety standard data of liquid fertilizer and the surface bearing capacity data of the target farmland area.

[0015] The basic dataset of farmland grids is obtained by associating and storing the planar area data, high-precision digital elevation model data, spatial distribution data of soil saturated hydraulic conductivity, nominal average saturated hydraulic conductivity data, standard nutrient requirements per unit area of ​​crops, and maximum safe water accumulation depth of liquid fertilizer corresponding to each farmland grid according to the farmland grid identifier.

[0016] Furthermore, the extraction and compression evaluation of the topological convergence features of closed depressions within the farmland grid to obtain a topological terrain compression optimization factor includes:

[0017] By performing closed depression hydrological topology identification processing on the high-precision digital elevation model data corresponding to the farmland grid, basic feature data of local closed depressions are obtained.

[0018] By performing topological convergence compression evaluation on the basic feature data of local closed depressions and the maximum safe water accumulation depth data of liquid fertilizer, the topological terrain compression optimization factor is obtained.

[0019] Furthermore, the step of performing closed depression hydrological topology identification processing on the high-precision digital elevation model data corresponding to the farmland grid to obtain basic feature data of local closed depressions includes:

[0020] For any target farmland grid, extract the high-precision digital elevation model data corresponding to the target farmland grid from the farmland grid basic dataset, and perform cropping processing on the high-precision digital elevation model data based on the two-dimensional spatial boundary of the target farmland grid to obtain the elevation raster data inside the target farmland grid.

[0021] For any target elevation raster in the elevation raster data within the target farmland grid, the elevation data of the target elevation raster is compared with that of the adjacent elevation raster. The adjacent elevation raster with the largest elevation drop is taken as the flow direction pointing raster of the target elevation raster. Continuous flow direction tracking is performed based on the flow direction pointing raster corresponding to all elevation rasters to obtain the surface runoff path data within the target farmland grid.

[0022] For any surface runoff path, when the surface runoff path converges in the low-elevation area inside the target farmland grid and does not directly connect to the two-dimensional spatial boundary of the target farmland grid, the low-elevation area is regarded as a local closed depression, and the number of local closed depressions is obtained based on the number of locally closed depressions identified.

[0023] For any locally closed depression, extract the elevation data corresponding to each elevation grid within the coverage area of ​​the locally closed depression from the elevation grid data inside the target farmland grid. Take the minimum elevation data within the coverage area of ​​the locally closed depression as the elevation data of the lowest point of the locally closed depression. Based on the boundary connection relationship between the locally closed depression and the surrounding elevation grid, obtain the lowest boundary elevation data of the liquid fertilizer overflowing from the locally closed depression, and take the lowest boundary elevation data as the elevation data of the overflow point of the locally closed depression.

[0024] For any locally closed depression, obtain all elevation grids that ultimately flow to the locally closed depression based on the surface runoff path data, and use the sum of the surface area of ​​all elevation grids that ultimately flow to the locally closed depression as the catchment area data of the locally closed depression.

[0025] For any local closed depression, multiple elevation levels are set between the elevation data of the lowest point of the local closed depression and the elevation data of the overflow point of the local closed depression; for any elevation level, the summation of the area of ​​continuous elevation grids within the coverage of the local closed depression whose elevation data is less than or equal to the elevation level is used as the area data enclosed by the closed contour lines corresponding to the elevation level.

[0026] By associating the data on the number of locally closed depressions corresponding to the target farmland grid, as well as the elevation data of the lowest point of each locally closed depression, the elevation data of the overflow point of each locally closed depression, the area data of the catchment area of ​​each locally closed depression, and the area data enclosed by the closed contour lines, the basic characteristic data of the locally closed depressions are obtained.

[0027] Furthermore, the process of performing topological convergence compression evaluation on the basic feature data of locally closed depressions and the maximum safe water accumulation depth data of liquid fertilizers to obtain topological terrain compression optimization factors includes:

[0028] For any target farmland grid, extract the planar area data and the maximum safe water accumulation depth of liquid fertilizer corresponding to the target farmland grid from the farmland grid basic dataset, and extract the number of local closed depressions corresponding to the target farmland grid from the local closed depression basic feature data, as well as the local closed depression lowest point elevation data, local closed depression overflow point elevation data, local closed depression catchment area data, and closed contour line enclosed area data corresponding to each local closed depression;

[0029] For any locally closed depression, between the elevation data of the lowest point of the locally closed depression and the elevation data of the overflow point of the locally closed depression, the area data enclosed by closed contour lines corresponding to each elevation level is obtained sequentially according to multiple elevation levels; the difference between the area data of the catchment area of ​​the locally closed depression and the area data enclosed by closed contour lines corresponding to each elevation level is used as the difference assessment of the convergence area corresponding to each elevation level; according to the arrangement order of the elevation levels, the difference assessment of the convergence area corresponding to each elevation level is processed by elevation direction cumulative integration to obtain the topological convergence volume assessment corresponding to the locally closed depression.

[0030] The topological convergence volume assessments of all locally closed depressions within the target farmland grid are summed to obtain the total topological convergence volume assessment of the closed depressions corresponding to the target farmland grid. The planar area data corresponding to the target farmland grid is multiplied by the maximum safe water accumulation depth data of liquid fertilizer to obtain the theoretical maximum safe carrying capacity volume corresponding to the target farmland grid. The total topological convergence volume assessment of closed depressions is used as the numerator, the theoretical maximum safe carrying capacity volume is used as the denominator, and the corresponding fraction is used as the relative penalty weight of topological convergence corresponding to the target farmland grid.

[0031] The result of adding the constant 1 to the topology convergence relative penalty weight is used as the topology convergence compression benchmark value corresponding to the target farmland grid, and the result of taking the reciprocal of the topology convergence compression benchmark value is used as the topology terrain compression optimization factor corresponding to the target farmland grid.

[0032] Furthermore, the analysis of the spatial coupling characteristics between potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid to obtain the spatial infiltration relaxation recovery factor includes:

[0033] By spatially overlaying and analyzing the basic feature data of local closed depressions with high-precision digital elevation model data, the potential water accumulation depth data of local closed depressions can be obtained.

[0034] By performing co-coordinate coupling analysis on the potential water accumulation depth data of local closed depressions and the spatial distribution data of soil saturated hydraulic conductivity, spatial synergistic characteristic data of deep infiltration are obtained.

[0035] By performing relaxation recovery assessment on deep infiltration spatial synergistic characteristic data, nominal average saturated hydraulic conductivity data, and closed depression topological convergence characteristic data, spatial infiltration relaxation recovery factors are obtained.

[0036] Furthermore, the step of obtaining potential water accumulation depth data of local closed depressions by spatially overlaying and analyzing the basic feature data of the depressions with high-precision digital elevation model data includes:

[0037] For any target farmland grid, extract the high-precision digital elevation model data corresponding to the target farmland grid from the farmland grid basic dataset, and extract the coverage data of each local closed depression corresponding to the target farmland grid and the elevation data of the overflow point of each local closed depression corresponding to the local closed depression from the local closed depression basic feature data.

[0038] For any locally closed depression, based on the coverage data of the locally closed depression, the actual ground elevation data corresponding to each spatial coordinate within the coverage area of ​​the locally closed depression is extracted from the high-precision digital elevation model data corresponding to the target farmland grid.

[0039] For any target spatial coordinate within the coverage area of ​​any locally closed depression, the difference between the elevation data of the overflow point of the locally closed depression and the actual ground elevation data corresponding to the target spatial coordinate is used as the initial potential water accumulation depth data corresponding to the target spatial coordinate.

[0040] When the initial potential water depth data corresponding to the target spatial coordinates is less than the constant 0, the potential water depth data of the local closed depression corresponding to the target spatial coordinates is set to the constant 0; when the initial potential water depth data corresponding to the target spatial coordinates is greater than or equal to the constant 0, the initial potential water depth data corresponding to the target spatial coordinates is used as the potential water depth data of the local closed depression corresponding to the target spatial coordinates.

[0041] The potential water accumulation depth data of the local closed depressions within the coverage area of ​​each local closed depression in the target farmland grid are correlated to obtain the potential water accumulation depth data of the local closed depressions corresponding to the target farmland grid.

[0042] Furthermore, the method involves performing co-coordinate coupling analysis on the potential water accumulation depth data of locally closed depressions and the spatial distribution data of soil saturated hydraulic conductivity to obtain spatially coordinated characteristic data of deep infiltration, including:

[0043] For any target farmland grid, extract the spatial distribution data of soil saturated hydraulic conductivity corresponding to the target farmland grid from the farmland grid basic dataset, and extract the potential water accumulation depth data of all spatial coordinates corresponding to the local closed depressions within the target farmland grid from the potential water accumulation depth data of local closed depressions.

[0044] For any target spatial coordinate within the coverage area of ​​any locally closed depression, extract the soil saturated hydraulic conductivity data corresponding to the target spatial coordinate from the spatial distribution data of soil saturated hydraulic conductivity.

[0045] For any target spatial coordinate within the coverage area of ​​any locally closed depression, multiply the potential water accumulation depth data of the locally closed depression corresponding to the target spatial coordinate with the soil saturated hydraulic conductivity data corresponding to the target spatial coordinate to obtain the depth permeability coupling assessment corresponding to the target spatial coordinate.

[0046] For any local closed depression, the depth permeability coupling assessment corresponding to all spatial coordinates within the coverage area of ​​the local closed depression is accumulated and integrated according to the spatial area to obtain the spatial collaborative feature data of single depression depth permeability.

[0047] The spatial collaborative feature data of depth infiltration of each single depression corresponding to all locally closed depressions within the target farmland grid are added together to obtain the spatial collaborative feature data of depth infiltration corresponding to the target farmland grid.

[0048] Furthermore, the process of obtaining a spatial infiltration relaxation recovery factor by performing relaxation recovery assessment on deep infiltration spatial synergistic characteristic data, nominal average saturated hydraulic conductivity data, and closed depression topological convergence characteristic data includes:

[0049] For any target farmland grid, extract the nominal average saturated hydraulic conductivity data corresponding to the target farmland grid from the farmland grid basic dataset, and extract the deep infiltration spatial collaborative feature data corresponding to the target farmland grid from the deep infiltration spatial collaborative feature data.

[0050] Obtain the total volume assessment of the topological convergence of the closed depression corresponding to the target farmland grid, and multiply the nominal average saturated hydraulic conductivity data corresponding to the target farmland grid with the total volume assessment of the topological convergence of the closed depression to obtain the homogeneous permeability benchmark convergence risk assessment corresponding to the target farmland grid.

[0051] The spatial coordinating characteristic data of deep infiltration corresponding to the target farmland grid is used as the numerator, the homogeneous infiltration benchmark convergence risk assessment corresponding to the target farmland grid is used as the denominator, and the corresponding fraction is used as the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid.

[0052] The result of adding constant 1 to the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid is used as the spatial infiltration relaxation recovery factor corresponding to the target farmland grid.

[0053] Furthermore, the step of obtaining the optimized dynamic absorption carrying capacity boundary by performing nonlinear bounded coupling processing on the topological terrain compression optimization factor and the spatial infiltration relaxation recovery factor includes:

[0054] For any target farmland grid, extract the planar area data and standard nutrient requirement per unit area of ​​crops corresponding to the target farmland grid from the farmland grid basic dataset, and obtain the topological compression optimization factor and spatial infiltration relaxation recovery factor corresponding to the target farmland grid.

[0055] Multiply the planar area data corresponding to the target farmland grid by the standard nutrient requirement per unit area of ​​the crop to obtain the theoretical maximum absorption carrying capacity boundary of the target farmland grid. Use the difference between the spatial infiltration relaxation recovery factor and constant 1 as the infiltration relaxation increment assessment for the target farmland grid, and perform hyperbolic tangent mapping on the infiltration relaxation increment assessment to obtain the bounded recovery coefficient of the infiltration relaxation for the target farmland grid. Use the difference between constant 1 and the topology compression optimization factor corresponding to the target farmland grid as the absorption recovery margin coefficient for the target farmland grid. Multiply the absorption recovery margin coefficient and the bounded recovery coefficient of the target farmland grid to obtain the bounded recovery compensation amount for the target farmland grid. Add the topology compression optimization factor and the bounded recovery compensation amount to obtain the dynamic carrying capacity correction coefficient for the target farmland grid. Multiply the theoretical maximum absorption carrying capacity boundary of the target farmland grid by the dynamic carrying capacity correction coefficient to obtain the optimized dynamic absorption carrying capacity boundary for the target farmland grid.

[0056] Compared with the prior art, the present invention has the following advantages:

[0057] This invention discloses a crop-livestock synergy optimization method based on spatial distribution feature analysis. By incorporating high-precision digital elevation model data of the target farmland area, closed depression topology data, and maximum safe water accumulation depth data for liquid fertilizer into the crop-livestock synergy spatial matching process, the absorption capacity boundary of farmland grids is no longer statically determined solely by two-dimensional planar area and crop fertilizer requirements. Instead, it reflects the gravity convergence pattern of liquid fertilizer under real micro-topographic conditions. For farmland grids with concave depressions, local low points, and hidden runoff paths, this scheme can identify the risk of liquid fertilizer accumulating in closed depressions and compress and correct the distributable absorption capacity of the corresponding grids. This avoids problems such as localized water accumulation, seedling burn, excessive nutrient accumulation, and high-concentration leakage pollution induced by the globally optimal solution obtained by the operational model at the data level during actual field application, thus improving the consistency between the liquid fertilizer application allocation scheme and real surface spatial conditions. Meanwhile, this technical solution further couples the spatial distribution data of soil saturated hydraulic conductivity with the local potential water accumulation depth in the same coordinate system. This allows the model to not only identify the "risk of low-lying area convergence" but also determine whether low-lying areas have strong infiltration and drainage capacity. For areas with good soil permeability at the bottom of depressions, this solution can provide bounded restoration of the rigid penalty caused by topographic convergence without exceeding the safe upper limit of the standard nutrient requirement per unit area of ​​crops. This avoids excessively excluding farmland grids with natural absorption capacity from the liquid fertilizer allocation range. As a result, the integrated crop-livestock spatial operation and matching model can achieve a more reasonable balance between environmental safety constraints, agronomic absorption capacity, and transportation scheduling efficiency, improving the precision and comprehensive utilization efficiency of livestock waste resource return to the field. Attached Figure Description

[0058] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0059] Figure 1 This is a flowchart illustrating a method for optimizing crop-livestock synergy based on spatial distribution feature analysis, as described in an embodiment of the present invention. Detailed Implementation

[0060] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] See Figure 1 This is a flowchart of a method for optimizing crop-livestock synergy based on spatial distribution feature analysis, as provided in Embodiment 1 of the present invention. Figure 1 As shown, a method for optimizing crop-livestock co-cultivation based on spatial distribution characteristic analysis may include:

[0062] Step S1: Collect and preprocess multi-source spatial basic data of the target farmland area to obtain a farmland grid basic dataset.

[0063] First, for the target farmland area to be allocated liquid fertilizer, the boundary range data of the target farmland area is obtained through a geographic information system (GIS), and the target farmland area is divided into two-dimensional spatial grids according to a preset grid scale to obtain multiple farmland grids. In this embodiment of the invention, the grid scale is set to 5 meters by 5 meters. For any farmland grid, the corresponding planar area data is calculated based on the two-dimensional spatial boundary of the farmland grid. The surface elevation of the target farmland area is collected using UAV photogrammetry equipment or airborne lidar equipment to obtain original surface elevation point cloud data or original surface elevation image data covering the target farmland area. The original surface elevation point cloud data or original surface elevation image data undergoes coordinate registration, noise point removal, spatial interpolation, and rasterization processing to obtain high-precision digital elevation model data corresponding to the spatial boundaries of each farmland grid. The spatial distribution attributes of the soil in the target farmland area are collected through a farmland IoT sensor array or multispectral remote sensing inversion method to obtain spatial distribution data of soil saturated hydraulic conductivity corresponding to different spatial coordinates within the target farmland area. For any farmland grid, the nominal average saturated hydraulic conductivity data is obtained by averaging the spatial distribution data of soil saturated hydraulic conductivity within the grid's spatial boundaries. Based on crop planting planning data and agronomic fertilization standards for the target farmland area, the standard nutrient requirements per unit area of ​​crops within that area are obtained. Based on the safety standards for surface application of liquid fertilizer and the surface bearing capacity data for the target farmland area, the maximum safe water accumulation depth for liquid fertilizer is obtained. The planar area data, high-precision digital elevation model data, spatial distribution data of soil saturated hydraulic conductivity, nominal average saturated hydraulic conductivity data, standard nutrient requirements per unit area of ​​crops, and maximum safe water accumulation depth for liquid fertilizer for each farmland grid are associated and stored according to the farmland grid identifier to obtain the basic dataset for the farmland grid.

[0064] This completes the acquisition of a farmland grid basic dataset by collecting and preprocessing multi-source spatial basic data of the target agricultural area.

[0065] Step S2 involves extracting and compressing the topological convergence features of closed depressions within the farmland grid to obtain the topological terrain compression optimization factor.

[0066] In actual farmland environments, the land surface is not an absolutely flat, ideal homogeneous plane, but inevitably contains micro-topographical undulations and localized closed depressions. Because the manure and organic fertilizers used in crop-livestock co-cultivation have liquid or semi-liquid fluid properties, when applied to grid plots, the fluids are driven by gravity to generate surface runoff, inevitably flowing and converging towards the low-lying, closed areas within the grid. Existing spatial matching algorithms, by using only the total two-dimensional area of ​​the grid as the basis for calculating the absorption capacity, mask the spatial heterogeneity caused by topographical undulations within the grid. This leads to the theoretically safe amount of liquid fertilizer allocated by the algorithm accumulating on the actual surface, far exceeding the carrying capacity of the local soil and crops at the bottom of the depressions, thus causing localized seedling burn and seepage pollution. To address this data-reality disconnect from the algorithm itself, it is necessary to break the static two-dimensional area constraint and introduce the spatial topological structure characteristics within the grid. Based on this, this method requires collecting high-precision digital elevation model data to identify all localized closed depression structures within the grid and quantifying the difference between the potential catchment area and the spatial carrying capacity of these depressions. By extracting this difference relationship, the physical trend of the grid forcing fluid to converge abnormally inward can be objectively reflected. Based on this physical trend, a preliminary optimization factor for terrain convergence characteristics is constructed, thereby dynamically correcting the ideal bearing capacity upper limit given by the traditional algorithm based on area, and directly eliminating the risk of environmental disasters caused by local enrichment from the algorithm constraint stage.

[0067] In summary, this invention first performs closed depression hydrological topology identification processing on the high-precision digital elevation model data corresponding to farmland grids to obtain basic feature data of local closed depressions. Specifically, for any target farmland grid, the high-precision digital elevation model data corresponding to the target farmland grid is extracted from the farmland grid basic dataset, and the high-precision digital elevation model data is clipped based on the two-dimensional spatial boundary of the target farmland grid to obtain the elevation raster data inside the target farmland grid. For any target elevation raster in the elevation raster data inside the target farmland grid, the elevation data of the target elevation raster is compared with that of adjacent elevation rasteres. The adjacent elevation raster with the largest elevation decrease is used as the flow direction pointing raster of the target elevation raster, and continuous flow direction tracking is performed based on the flow direction pointing raster corresponding to all elevation rasteres to obtain the surface runoff path data inside the target farmland grid. For any surface runoff path, when the surface runoff path converges in a low-elevation area within the target farmland grid but does not directly connect to the two-dimensional spatial boundary of the target farmland grid, the low-elevation area is considered a locally closed depression, and the number of locally closed depressions is obtained based on the number of identified depressions. For any locally closed depression, the elevation data corresponding to each elevation raster within the coverage area of ​​the locally closed depression is extracted from the elevation raster data within the target farmland grid, and the minimum elevation data within the coverage area of ​​the locally closed depression is taken as the elevation data of the lowest point of the locally closed depression. Based on the boundary connection relationship between the locally closed depression and the surrounding elevation raster, the lowest boundary elevation data of the runoff overflowing from the locally closed depression is obtained, and this lowest boundary elevation data is taken as the elevation data of the overflow point of the locally closed depression. For any locally closed depression, all elevation rasteres that ultimately flow to the locally closed depression are obtained based on the surface runoff path data, and the sum of the surface area of ​​all elevation rasteres that ultimately flow to the locally closed depression is taken as the catchment area data of the locally closed depression. For any locally closed depression, multiple elevation levels are established between the elevation data of the lowest point of the locally closed depression and the elevation data of the overflow point of the locally closed depression. For any elevation level, the summation of the area of ​​continuous elevation grids within the coverage area of ​​the locally closed depression whose elevation data is less than or equal to the elevation level is used as the area data enclosed by the closed contour lines corresponding to the elevation level. The number of locally closed depressions corresponding to the target farmland grid, as well as the elevation data of the lowest point of each locally closed depression, the elevation data of the overflow point of each locally closed depression, the area data of the catchment area of ​​each locally closed depression, and the area data enclosed by the closed contour lines are correlated to obtain the basic characteristic data of the locally closed depressions.

[0068] After obtaining the basic feature data of locally closed depressions, the topological convergence compression evaluation process is further performed on the basic feature data of locally closed depressions and the maximum safe water accumulation depth data of liquid fertilizer. This yields a topological compression optimization factor. Specifically, for any target farmland grid, the planar area data and the maximum safe water accumulation depth data of liquid fertilizer corresponding to the target farmland grid are extracted from the farmland grid basic dataset. Furthermore, the number of locally closed depressions corresponding to the target farmland grid, as well as the elevation data of the lowest point, overflow point, catchment area, and area enclosed by closed contour lines for each locally closed depression are extracted from the basic feature data of locally closed depressions. For any locally closed depression, the area enclosed by closed contour lines at each elevation level is sequentially obtained between the lowest point elevation data and the overflow point elevation data. The difference between the catchment area data of the locally closed depression and the area enclosed by closed contour lines at each elevation level is used as the evaluation of the convergence area difference at each elevation level. Following the elevation hierarchy, the differences in convergence area corresponding to each elevation hierarchy are evaluated by accumulating and integrating along the elevation direction to obtain the topological convergence volume assessment corresponding to local closed depressions. The topological convergence volume assessments corresponding to all local closed depressions within the target farmland grid are summed to obtain the total topological convergence volume assessment of closed depressions within the target farmland grid. The planar area data corresponding to the target farmland grid is multiplied by the maximum safe water accumulation depth data for liquid fertilizer to obtain the theoretical maximum safe carrying capacity volume for the target farmland grid. The total topological convergence volume assessment of closed depressions is used as the numerator, and the theoretical maximum safe carrying capacity volume is used as the denominator. The resulting fraction is used as the relative penalty weight for topological convergence within the target farmland grid. The result of adding the constant 1 to the relative penalty weight for topological convergence is used as the topological convergence compression benchmark value for the target farmland grid, and the reciprocal of the topological convergence compression benchmark value is used as the topological terrain compression optimization factor for the target farmland grid.

[0069] In one implementation, assume the first The planar area data corresponding to each farmland grid is: The maximum safe water depth for liquid fertilizer is... ;No. The area data of the catchment area of ​​the locally closed depression is as follows: ;No. The area enclosed by closed contour lines corresponding to each elevation level is: ; in the The total number of locally closed depressions in each farmland grid is ;No. The critical overflow point elevation of a locally closed depression is ;No. The absolute lowest point elevation of the locally closed depression is Then the first The expression for calculating the topology compression optimization factor corresponding to each farmland grid is:

[0070]

[0071] in, Indicates the first Topological terrain compression optimization factor corresponding to each farmland grid; Indicates the first Planar area data corresponding to each farmland grid; This represents the maximum safe water depth for liquid fertilizer under agronomic standards. Indicates the first The total number of locally closed depressions in each farmland grid; Indicates the first The absolute lowest point elevation of a locally closed depression; Indicates the first Elevation of the critical overflow point of a locally closed depression; Indicates the first Data on the area of ​​the catchment area of ​​a locally closed depression; Indicates the first Area data enclosed by closed contour lines corresponding to each elevation level.

[0072] It should be noted that the reason existing scalar constraints fail in meshes containing micro-topography is fundamentally because the algorithm cannot recognize the funnel-like converging effect of terrain on fluids. When the catchment area of ​​a depression is much larger than its actual bottom capacity, fluid around it will rapidly accumulate in a very small local area, causing the local liquid level to rapidly exceed the safe depth. To solve this problem, this invention delves into the interior of the landmass based on digital elevation model data, constructing an integral term for each identified closed depression. In this design, This characterizes the maximum area that can be supplied with liquid fertilizer to the depression. Characterized by elevation The integral term accurately calculates the difference in spatial volume between the projected potential inflow volume and the actual usable volume. This volume difference directly represents the convergence potential energy that forces the fluid to accumulate beyond its limit due to the terrain. A larger integral value indicates a more significant funnel effect and a higher risk of local disasters. To convert the aforementioned absolute convergence potential energy into a dimensionless optimization correction coefficient effective for the global algorithm, the formula further introduces... As the denominator. Because The total area of ​​the grid. For the safe water accumulation depth, the product of the two constitutes the theoretical maximum safe carrying capacity of the entire plot under absolutely ideal flat conditions. By dividing the sum of the accumulated topological convergence potential energy by the theoretical maximum safe carrying capacity, this method calculates the relative penalty weight of the local dangerous convergence effect in the overall ideal absorption capacity of the grid. Finally, using... The structure maps the aforementioned weights, thus constructing the optimization factor. When the mesh interior is absolutely flat and has no closed traps (i.e., the integral term is zero), Automatically degenerate into This does not affect the original water absorption allocation of the plot; however, when the catchment area inside the grid is large and the funnel is extremely steep, causing a surge in the integral term, the denominator increases nonlinearly, making... Strictly compressed to a size much smaller Within a safe range. This structure requires no manually preset adjustment parameters; it dynamically outputs the compression ratio entirely through the interaction of pure three-dimensional spatial geometric volume relationships and fluid dynamic boundaries. This forcibly reduces the total absorption quota of the dangerous grid before the allocation algorithm is executed, cutting off the possibility of the algorithm's optimal solution causing local physical space environmental disasters from the source.

[0073] Thus, the extraction and compression evaluation of the topological convergence features of closed depressions within the farmland grid were completed, and the topological terrain compression optimization factor was obtained.

[0074] Step S3: By analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid, the spatial infiltration relaxation recovery factor is obtained.

[0075] Although the optimization factor constructed in step S2 is based on the static three-dimensional spatial geometric topology within the grid, accurately quantifying the fluid convergence potential energy and forcibly compressing the absorption capacity, thus effectively avoiding the risks of local water accumulation and enrichment overload, in actual farmland geological environments, the soil medium at the bottom of depressions is not completely impermeable. As liquid manure converges towards low-lying areas, it inevitably undergoes an infiltration process into deeper soil layers. Applying the topological compression optimization factor output in step S2 directly to all plots essentially assumes that all local depressions are static stagnant water retention ponds, ignoring the spatial heterogeneity of actual soil texture distribution. In real-world scenarios, if a core depression area with extremely deep topographic convergence has soil texture with extremely high infiltration conductivity, then the liquid manure collected on the surface will quickly infiltrate the ground, preventing prolonged surface water accumulation overload and environmental pollution. In this context, topological compression optimization factors relying solely on geometric volume for unidirectional penalty can over-compress plots with excellent self-draining capacity, leading to an overly conservative absorption and allocation strategy that unnecessarily sacrifices the overall absorption efficiency of the entire crop-livestock synergy system. Therefore, it is essential to further consider the spatial misalignment or positive spatial synergy between micro-topographic structure and soil permeability characteristics during the constraint phase. This method simultaneously acquires two-dimensional soil saturated hydraulic conductivity spatial distribution data from farmland grids and spatially maps and couples it with elevation distribution data to analyze the spatial overlap between low-lying extreme areas and high-permeability areas. Based on this overlap, a flexible relaxation correction is applied to the rigid over-penalty generated in step S2, thereby maximizing the reasonable absorption space of the restoration algorithm while ensuring environmental safety.

[0076] In summary, this invention first obtains the potential water accumulation depth data of locally closed depressions by spatially overlaying and analyzing the basic feature data of the depressions with high-precision digital elevation model data. Specifically, for any target farmland grid, the high-precision digital elevation model data corresponding to the target farmland grid is extracted from the farmland grid basic dataset, and the coverage area data of each locally closed depression corresponding to the target farmland grid and the elevation data of the overflow point of each locally closed depression are extracted from the basic feature data of the locally closed depressions. For any locally closed depression, based on the coverage area data of the locally closed depression, the actual ground elevation data corresponding to each spatial coordinate within the coverage area of ​​the locally closed depression is extracted from the high-precision digital elevation model data corresponding to the target farmland grid. For any target spatial coordinate within the coverage area of ​​any locally closed depression, the difference between the elevation data of the overflow point of the locally closed depression and the actual ground elevation data corresponding to the target spatial coordinate is used as the initial potential water accumulation depth data corresponding to the target spatial coordinate. When the initial potential water accumulation depth corresponding to the target spatial coordinates is less than a constant 0, the potential water accumulation depth data of the local closed depressions corresponding to the target spatial coordinates is set to a constant 0. When the initial potential water accumulation depth corresponding to the target spatial coordinates is greater than or equal to a constant 0, the initial potential water accumulation depth data corresponding to the target spatial coordinates is used as the potential water accumulation depth data of the local closed depressions corresponding to the target spatial coordinates. The potential water accumulation depth data of the local closed depressions corresponding to all spatial coordinates within the coverage area of ​​each local closed depression within the target farmland grid are correlated to obtain the potential water accumulation depth data of the local closed depressions corresponding to the target farmland grid.

[0077] After obtaining the potential water accumulation depth data of locally closed depressions, the analysis further involves co-coordinate coupling of this data with the spatial distribution data of soil saturated hydraulic conductivity to acquire spatial synergistic feature data of deep infiltration. Specifically, for any target farmland grid, the spatial distribution data of soil saturated hydraulic conductivity corresponding to the target farmland grid is extracted from the basic farmland grid dataset, and the potential water accumulation depth data of all spatial coordinates within the coverage area of ​​each locally closed depression within the target farmland grid is extracted from the potential water accumulation depth data of locally closed depressions. For any target spatial coordinate within the coverage area of ​​any locally closed depression, the soil saturated hydraulic conductivity data corresponding to the target spatial coordinate is extracted from the spatial distribution data of soil saturated hydraulic conductivity. For any target spatial coordinate within the coverage area of ​​any locally closed depression, the potential water accumulation depth data of the locally closed depression corresponding to the target spatial coordinate is multiplied by the soil saturated hydraulic conductivity data corresponding to the target spatial coordinate to obtain the deep infiltration coupling assessment corresponding to the target spatial coordinate. For any locally closed depression, the depth permeability coupling assessment corresponding to all spatial coordinates within the coverage area of ​​the locally closed depression is accumulated and integrated according to the spatial area to obtain the spatial collaborative feature data of single depression depth permeability corresponding to the locally closed depression. The spatial collaborative feature data of single depression depth permeability corresponding to all locally closed depressions within the target farmland grid are added together to obtain the spatial collaborative feature data of depth permeability corresponding to the target farmland grid.

[0078] After obtaining the spatial coordination characteristic data of deep infiltration, a relaxation recovery assessment is performed on the deep infiltration spatial coordination characteristic data, nominal average saturated hydraulic conductivity data, and closed depression topological convergence characteristic data to obtain the spatial infiltration relaxation recovery factor. Specifically, for any target farmland grid, the nominal average saturated hydraulic conductivity data corresponding to the target farmland grid is extracted from the farmland grid basic dataset, and the deep infiltration spatial coordination characteristic data corresponding to the target farmland grid is extracted from the deep infiltration spatial coordination characteristic data. The total volume assessment of the closed depression topological convergence corresponding to the target farmland grid is obtained. The nominal average saturated hydraulic conductivity data corresponding to the target farmland grid is multiplied by the total volume assessment of the closed depression topological convergence to obtain the homogeneous infiltration benchmark convergence risk assessment corresponding to the target farmland grid. The deep infiltration spatial coordination characteristic data corresponding to the target farmland grid is used as the numerator, and the homogeneous infiltration benchmark convergence risk assessment corresponding to the target farmland grid is used as the denominator. The resulting fraction is used as the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid. The result of adding the constant 1 to the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid is used as the spatial infiltration relaxation recovery factor corresponding to the target farmland grid.

[0079] In one implementation, assume the first The integral region of each locally closed depression in two-dimensional space is: Geographic coordinates The actual ground elevation at that location is Geographic coordinates The actual soil saturated hydraulic conductivity at that location is ;No. The global average saturated hydraulic conductivity of the plots is Then the first The formula for calculating the spatial infiltration relaxation recovery factor corresponding to each farmland grid is as follows:

[0080]

[0081] in, Indicates the first Spatial infiltration relaxation recovery factor corresponding to each farmland grid; Indicates the first The integral region of a locally closed depression in two-dimensional space; Indicates the first The total number of locally closed depressions in each farmland grid; Indicates the first The absolute lowest point elevation of a locally closed depression; Indicates the first Elevation of the critical overflow point of a locally closed depression; Indicates the first Data on the area of ​​the catchment area of ​​a locally closed depression; Indicates the first Area data enclosed by closed contour lines corresponding to each elevation level; Representing geographic coordinates The actual ground elevation at the location; Representing geographic coordinates The actual saturated hydraulic conductivity of the soil at that location; Indicates the first The global average saturated hydraulic conductivity of each plot.

[0082] It should be noted that even when considering soil permeability, existing technologies often extract the average permeability parameter of the entire site as an independent term for macroscopic calculations. This conventional statistical method severs the physical spatial connection between the hydrological accumulation process and the medium's conduction process. To accurately measure the inherent drainage and storage synergy of a depression, the core of this invention lies in the numerator term... In this two-dimensional integral structure, This indicates the maximum potential water depth at that geographic coordinate point, which objectively reflects the forcing pressure at which fluid converges; while This represents the actual permeability at that point. By multiplying and integrating the two points point-to-point in the same physical coordinate system, this design quantifies the spatial covariance effect of depth-permeability. Specifically, if the deepest area of ​​a depression, i.e., the part with the most potential for water accumulation, has the highest permeability at its bottom... If the infiltration rate is both extremely high and extremely low, the integral value will show explosive growth, proving that the low-lying area and the excellent infiltration area form a perfect spatial positive synergy, possessing a strong natural ability to absorb excess accumulated fluid. Conversely, if the deepest part is compacted, impermeable clay, even if the higher areas at the edge of the depression have good permeability, the integral value will be extremely small due to the suppression of depth weight. To convert this spatial synergy into a reasonable relaxation coefficient, the average permeability of the land parcel is introduced into the denominator of the formula. Integrating the topological convergence volume in step S2 The product of the terms. The denominator represents the baseline convergence risk of the plot under the homogeneous permeability assumption. When the spatial cooperative drainage term in the numerator is much larger than the homogeneous risk penalty term in the denominator, it indicates that the actual spatial heterogeneous distribution greatly mitigates the convergence risk. In this case, the calculated value of the latter part of the formula increases, making... The value is significantly greater than 1; however, when the synergistic effect is not obvious or is negatively correlated, the latter part approaches zero. It automatically approaches 1, thus strictly maintaining the state in step S2. The original penalty intensity is eliminated. This design abandons the intervention of human experience in relaxing parameters and achieves fine-grained elastic compensation for the rigid volume penalty in the early stage based solely on the objective misalignment data of terrain undulation and medium properties within the grid. This effectively avoids the optimization algorithm from falling into unreasonable conservative allocation.

[0083] Thus, the spatial infiltration relaxation recovery factor was obtained by analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid.

[0084] Step S4: By performing nonlinear bounded coupling processing on the topological terrain compression optimization factor and the spatial infiltration relaxation recovery factor, the optimized dynamic absorption carrying capacity boundary is obtained.

[0085] Existing spatial management algorithms for integrated crop-livestock farming (in this embodiment, a single-source, multi-sink allocation model based on mixed-integer linear programming) require strict boundary constraints on the carrying capacity of demand nodes when optimizing the objective function. Existing technologies directly multiply the static two-dimensional area by the standard fertilizer requirement (i.e.... This is used as a scalar constraint. This step strictly limits the relaxation compensation effect of the spatial infiltration relaxation recovery factor to the safe difference range generated by the early penalty of the topology compression optimization factor. Its corrected carrying capacity can only infinitely approach but never exceed the maximum biological theoretical demand of the crop, thus solving the loophole of optimization failure caused by over-excitation divergence of the relaxation factor.

[0086] Specifically, for any target farmland grid, the planar area data and standard nutrient requirement per unit area of ​​the crop are extracted from the basic farmland grid dataset. The topological compression optimization factor and spatial infiltration relaxation recovery factor corresponding to the target farmland grid are then obtained. The planar area data and standard nutrient requirement per unit area of ​​the crop are multiplied to obtain the theoretical maximum absorption carrying capacity boundary of the target farmland grid. The difference between the spatial infiltration relaxation recovery factor and constant 1 is used as the infiltration relaxation increment assessment for the target farmland grid. This infiltration relaxation increment assessment is then processed using a hyperbolic tangent mapping to obtain the bounded recovery coefficient of the infiltration relaxation for the target farmland grid. The difference between constant 1 and the topological compression optimization factor corresponding to the target farmland grid is used as the absorption recovery margin coefficient for the target farmland grid. The result of multiplying the absorption recovery margin coefficient and the bounded recovery coefficient of the infiltration relaxation for the target farmland grid is used as the bounded recovery compensation amount for the target farmland grid. The result of adding the topological compression optimization factor corresponding to the target farmland grid to the bounded restoration compensation amount is used as the dynamic carrying capacity correction coefficient corresponding to the target farmland grid. The optimized dynamic carrying capacity boundary corresponding to the target farmland grid is obtained by multiplying the theoretical maximum absorption carrying capacity boundary corresponding to the target farmland grid with the dynamic carrying capacity correction coefficient.

[0087] In one implementation, it is assumed that the standard fertilizer requirement is Then the first The calculation expression for the optimized dynamic absorption carrying capacity boundary corresponding to each farmland grid is:

[0088]

[0089] in, Indicates the first The optimized dynamic absorption carrying capacity boundary corresponding to each farmland grid; Indicates the first Planar area data corresponding to each farmland grid; Indicates the standard fertilizer requirement; Indicates the first Spatial infiltration relaxation recovery factor corresponding to each farmland grid; Indicates the first Topological terrain compression optimization factor corresponding to each farmland grid; This represents the hyperbolic tangent function.

[0090] It should be noted that in this correction formula, This represents the upper limit of the maximum theoretical biological demand under ideal, flat conditions. This is the nonlinear coupling compensation operator of the present invention. Wherein, This is the starting point for compression in step S2 based on static terrain volume calculation, and This precisely represents the absorption margin space penalized by terrain. This formula uses the hyperbolic tangent function... Perform a nonlinear mapping on the relaxation factor. The range property of a function guarantees that when the range of variables is within a certain range, the function will be able to perform a function-wise operation When, its output value strictly converges to The interval. Due to the range specified in step S3. ,therefore It is always a non-negative number. Through this coupling structure, when the substrate is completely impermeable... At this point, the formula completely degenerates into Strictly maintain the terrain compression penalty; when the bottom layer of the depression has excellent permeability ( Much larger ), The limit of the operator within the square brackets gradually approaches 1, at which point the limit of the operator within the square brackets approaches 1. This makes the overall dynamic load capacity Safe and smooth recovery to The biological absolute upper limit. This nonlinear mechanism perfectly isolates the potential divergence risk of relaxation factors, ensuring that the optimized bearing boundary is compatible with three dimensions: fluid convergence penalty, media penetration relaxation, and crop physiological safety baseline.

[0091] Thus, the optimized dynamic absorption carrying capacity boundary was obtained by nonlinearly coupling the topological compression optimization factor and the spatial infiltration relaxation recovery factor.

[0092] Step S5: By substituting the optimized dynamic absorption carrying capacity boundary into the planting and breeding collaborative spatial operation matching model for solution, the liquid fertilizer spatial allocation scheme is obtained.

[0093] After obtaining the optimized dynamic absorption capacity boundary corresponding to each farmland grid, it is introduced as a dynamic absorption constraint condition for farmland demand nodes into the crop-livestock collaborative spatial operation and matching model. Specifically, for each farmland grid in the target farmland area, each farmland grid is regarded as a liquid fertilizer absorption demand node, and the livestock farm, liquid fertilizer temporary storage pond, or pipeline fertilizer outlet node is regarded as a liquid fertilizer supply node. The total amount of liquid fertilizer that can be supplied for each supply node and the transportation distance data or pipeline energy consumption cost data between each supply node and each farmland grid are obtained. Based on this, the crop-livestock collaborative spatial operation and matching model is constructed with the objective function of minimizing the total transportation distance, minimizing the total pipeline energy consumption, or minimizing the comprehensive scheduling cost in the liquid fertilizer scheduling process.

[0094] Regarding model constraints, for any farmland grid, the amount of liquid fertilizer allocated to that grid is limited to no more than the optimized dynamic absorption capacity boundary corresponding to that grid. This ensures that the liquid fertilizer allocation result does not exceed the safe absorption capacity after considering the micro-topographic convergence risk and soil infiltration recovery capacity. Simultaneously, a total supply constraint is set for each supply node, ensuring that the total amount of liquid fertilizer output by each node does not exceed its total available supply. Furthermore, constraints on total liquid fertilizer balance, transportation path accessibility, or operational equipment capacity are set according to actual crop-livestock synergy requirements. Thus, the crop-livestock synergy spatial operation and matching model completes spatial allocation optimization while satisfying the farmland environmental safety boundary.

[0095] After constructing the objective function and constraints, a mixed-integer linear programming solver is invoked to perform global optimization on the crop-livestock collaborative spatial operation matching model. This yields the liquid fertilizer allocation amount between each supply node and each farmland grid, forming a liquid fertilizer spatial allocation matrix. This liquid fertilizer spatial allocation matrix is ​​then used as the liquid fertilizer spatial allocation scheme and distributed to the farmland return pipeline control system, fertilization scheduling management platform, or autonomous driving fertilization vehicle terminal to execute liquid fertilizer return operations according to the liquid fertilizer spatial allocation scheme.

[0096] Thus, the spatial allocation scheme for liquid fertilizer was obtained by substituting the optimized dynamic absorption carrying capacity boundary into the spatial operation and matching model of planting and breeding collaboration.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing crop-livestock synergy based on spatial distribution characteristic analysis, characterized in that, The method includes: Step S1: Collect and preprocess multi-source spatial basic data of the target farmland area to obtain a basic dataset of farmland grids; Step S2: Extract and compress the topological convergence features of closed depressions within the farmland grid to obtain the topological terrain compression optimization factor; Step S3: By analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within the farmland grid, the spatial infiltration relaxation recovery factor is obtained; Step S4: Obtain the optimized dynamic absorption bearing boundary by performing nonlinear bounded coupling processing on the topological terrain compression optimization factor and the spatial infiltration relaxation recovery factor; Step S5: By substituting the optimized dynamic absorption carrying capacity boundary into the planting and breeding synergy spatial operation matching model for solution, the liquid fertilizer spatial allocation scheme is obtained; The process of extracting and compressing the topological convergence features of closed depressions within farmland grids to obtain topological topographic compression optimization factors includes: performing closed depression hydrological topological identification processing on the high-precision digital elevation model data corresponding to the farmland grids to obtain basic feature data of local closed depressions; and performing topological convergence compression evaluation processing on the basic feature data of local closed depressions and the maximum safe water accumulation depth data of liquid fertilizer to obtain topological topographic compression optimization factors. The process involves performing topological convergence compression evaluation on the basic feature data of locally closed depressions and the maximum safe water accumulation depth data of liquid fertilizer, to obtain a topological terrain compression optimization factor. This includes: for any target farmland grid, extracting the corresponding planar area data and maximum safe water accumulation depth data of liquid fertilizer from the farmland grid basic dataset; extracting the number of locally closed depressions corresponding to the target farmland grid from the locally closed depression basic feature data; and extracting the elevation data of the lowest point of each locally closed depression, the elevation data of the overflow point of each locally closed depression, the area data of the catchment area of ​​each locally closed depression, and the area data enclosed by closed contour lines. For any locally closed depression, between the elevation data of the lowest point of the locally closed depression and the elevation data of the overflow point of the locally closed depression, sequentially obtaining the area data enclosed by closed contour lines corresponding to each elevation level according to multiple elevation levels; and using the difference between the area data of the catchment area of ​​the locally closed depression and the area data enclosed by closed contour lines corresponding to each elevation level as the elevation level. The evaluation process involves assessing the convergence area differences corresponding to each elevation level; integrating the convergence area differences at each elevation level along the elevation direction to obtain the topological convergence volume assessment corresponding to local closed depressions; summing the topological convergence volume assessments corresponding to all local closed depressions within the target farmland grid to obtain the total topological convergence volume assessment of closed depressions within the target farmland grid; multiplying the planar area data of the target farmland grid with the maximum safe water accumulation depth data for liquid fertilizer to obtain the theoretical maximum safe carrying capacity of the target farmland grid; using the total topological convergence volume assessment of closed depressions as the numerator and the theoretical maximum safe carrying capacity as the denominator, and using the resulting fraction as the relative penalty weight for topological convergence of the target farmland grid; and using the result of adding the constant 1 to the relative penalty weight for topological convergence as the topological convergence compression benchmark value for the target farmland grid, and using the reciprocal of the topological convergence compression benchmark value as the topological terrain compression optimization factor for the target farmland grid. The method involves analyzing the spatial coupling characteristics of potential water accumulation depth and soil saturated hydraulic conductivity within farmland grids to obtain a spatial infiltration relaxation recovery factor. This includes: obtaining potential water accumulation depth data for local closed depressions by performing spatial overlay analysis on basic feature data and high-precision digital elevation model data; obtaining spatial synergistic feature data of deep infiltration by performing co-coordinate coupling analysis on potential water accumulation depth data of local closed depressions and spatial distribution data of soil saturated hydraulic conductivity; and obtaining a spatial infiltration relaxation recovery factor by performing relaxation recovery assessment on spatial synergistic feature data of deep infiltration, nominal average saturated hydraulic conductivity data, and topological convergence feature data of closed depressions. The method involves performing co-coordinate coupling analysis on the potential water accumulation depth data of locally closed depressions and the spatial distribution data of soil saturated hydraulic conductivity to obtain spatially coordinated feature data of deep infiltration. This includes: for any target farmland grid, extracting the spatial distribution data of soil saturated hydraulic conductivity corresponding to the target farmland grid from the basic farmland grid dataset, and extracting the potential water accumulation depth data of all spatial coordinates within the coverage area of ​​each locally closed depression within the target farmland grid from the potential water accumulation depth data of locally closed depressions; for any target spatial coordinate within the coverage area of ​​any locally closed depression, extracting the soil saturated hydraulic conductivity corresponding to the target spatial coordinate from the spatial distribution data of soil saturated hydraulic conductivity based on the target spatial coordinate. Rate data; for any target spatial coordinate within the coverage area of ​​any locally closed depression, multiply the potential water accumulation depth data of the locally closed depression corresponding to the target spatial coordinate with the soil saturated hydraulic conductivity data corresponding to the target spatial coordinate to obtain the depth permeability coupling assessment corresponding to the target spatial coordinate; for any locally closed depression, accumulate and integrate the depth permeability coupling assessments corresponding to all spatial coordinates within the coverage area of ​​the locally closed depression according to the spatial area to obtain the spatial synergistic feature data of single depression depth permeability corresponding to the locally closed depression; add the spatial synergistic feature data of single depression depth permeability corresponding to all locally closed depressions within the target farmland grid to obtain the spatial synergistic feature data of depth permeability corresponding to the target farmland grid; The process of obtaining a spatial infiltration relaxation recovery factor by performing relaxation recovery assessment on deep infiltration spatial synergistic feature data, nominal average saturated hydraulic conductivity data, and closed depression topological convergence feature data includes: for any target farmland grid, extracting the nominal average saturated hydraulic conductivity data corresponding to the target farmland grid from the farmland grid basic dataset, and extracting the deep infiltration spatial synergistic feature data corresponding to the target farmland grid from the deep infiltration spatial synergistic feature data; obtaining the total volume assessment of closed depression topological convergence corresponding to the target farmland grid, multiplying the nominal average saturated hydraulic conductivity data corresponding to the target farmland grid with the total volume assessment of closed depression topological convergence to obtain the homogeneous infiltration benchmark convergence risk assessment corresponding to the target farmland grid; using the deep infiltration spatial synergistic feature data corresponding to the target farmland grid as the numerator, and the homogeneous infiltration benchmark convergence risk assessment corresponding to the target farmland grid as the denominator, and using the resulting fraction as the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid; and adding the constant 1 to the spatial infiltration relaxation recovery assessment corresponding to the target farmland grid as the spatial infiltration relaxation recovery factor corresponding to the target farmland grid. The process of obtaining the optimized dynamic absorption carrying capacity boundary by nonlinearly and boundedly coupling the topological compression optimization factor and the spatial infiltration relaxation recovery factor includes: for any target farmland grid, extracting the planar area data and crop standard nutrient requirement per unit area data corresponding to the target farmland grid from the farmland grid basic dataset, and obtaining the topological compression optimization factor and spatial infiltration relaxation recovery factor corresponding to the target farmland grid; multiplying the planar area data and crop standard nutrient requirement per unit area data corresponding to the target farmland grid to obtain the theoretical maximum absorption carrying capacity boundary corresponding to the target farmland grid; using the difference between the spatial infiltration relaxation recovery factor corresponding to the target farmland grid and the constant 1 as the infiltration relaxation increment evaluation corresponding to the target farmland grid, and then... The infiltration relaxation increment assessment is processed using hyperbolic tangent mapping to obtain the bounded restoration coefficient of infiltration relaxation corresponding to the target farmland grid. The difference between constant 1 and the topology compression optimization factor corresponding to the target farmland grid is used as the absorption restoration margin coefficient corresponding to the target farmland grid. The result of multiplying the absorption restoration margin coefficient corresponding to the target farmland grid with the bounded restoration coefficient of infiltration relaxation is used as the bounded restoration compensation amount corresponding to the target farmland grid. The result of adding the topology compression optimization factor corresponding to the target farmland grid with the bounded restoration compensation amount is used as the dynamic carrying capacity correction coefficient corresponding to the target farmland grid. The optimized dynamic absorption carrying capacity boundary corresponding to the target farmland grid is obtained by multiplying the theoretical maximum absorption carrying capacity boundary corresponding to the target farmland grid with the dynamic carrying capacity correction coefficient. The process involves substituting the optimized dynamic absorption capacity boundary into the integrated farming and aquaculture spatial management matching model to obtain a liquid fertilizer spatial allocation scheme. This includes: for each farmland grid within the target farmland area, each farmland grid is designated as a liquid fertilizer absorption demand node, and livestock farms, liquid fertilizer storage ponds, or pipeline fertilizer outlet nodes are designated as liquid fertilizer supply nodes. The total available liquid fertilizer for each supply node and the transportation distance data or pipeline energy consumption cost data between each supply node and each farmland grid are obtained. The integrated farming and aquaculture spatial management matching model is constructed with the objective function of minimizing the total transportation distance, minimizing the total pipeline energy consumption, or minimizing the overall scheduling cost during the liquid fertilizer scheduling process. For any farmland grid... The system assigns liquid fertilizer to each farmland grid, limiting the amount allocated to that grid to no more than the optimized dynamic absorption capacity boundary of that grid. Simultaneously, it sets total supply constraints for each supply node, ensuring that the total liquid fertilizer output by each node does not exceed its total supply capacity. Furthermore, it sets liquid fertilizer balance constraints, transportation path accessibility constraints, or operational equipment capacity constraints based on actual crop-livestock synergy requirements. A mixed-integer linear programming solver is invoked to globally optimize the crop-livestock synergy spatial operation matching model, obtaining the liquid fertilizer allocation between each supply node and each farmland grid, forming a liquid fertilizer spatial allocation matrix. This liquid fertilizer spatial allocation matrix serves as the liquid fertilizer spatial allocation scheme.

2. The method for optimizing crop-livestock synergy based on spatial distribution characteristic analysis according to claim 1, characterized in that, The process involves collecting and preprocessing multi-source spatial data of the target agricultural area to obtain a basic farmland grid dataset, including: For the target farmland area to be allocated liquid fertilizer, the boundary range data of the target farmland area is obtained through a geographic information system, and the target farmland area is divided into two-dimensional spatial grids according to a preset grid scale to obtain multiple farmland grids. For any farmland grid, the corresponding planar area data is calculated based on the two-dimensional spatial boundary of the farmland grid. The surface elevation of the target farmland area is collected by UAV photogrammetry equipment or airborne lidar equipment to obtain original surface elevation point cloud data or original surface elevation image data covering the target farmland area. The original surface elevation point cloud data or original surface elevation image data is subjected to coordinate registration, noise point removal, spatial interpolation and rasterization processing to obtain high-precision digital elevation data corresponding to the spatial boundary of each farmland grid. The process involves: collecting soil spatial distribution attributes of the target farmland area through farmland IoT sensor arrays or multispectral remote sensing inversion methods to obtain spatial distribution data of soil saturated hydraulic conductivity corresponding to different spatial coordinates within the target farmland area; for any farmland grid, performing average statistical processing on the spatial distribution data of soil saturated hydraulic conductivity within the spatial boundary of the farmland grid to obtain the nominal average saturated hydraulic conductivity data corresponding to the farmland grid; obtaining the standard nutrient requirement per unit area of ​​crops within the target farmland area based on crop planting planning data and agronomic fertilization standard data for the target farmland area; and obtaining the maximum safe water accumulation depth of liquid fertilizer based on the surface application safety standard data of liquid fertilizer and the surface bearing capacity data of the target farmland area. The basic dataset of farmland grids is obtained by associating and storing the planar area data, high-precision digital elevation model data, spatial distribution data of soil saturated hydraulic conductivity, nominal average saturated hydraulic conductivity data, standard nutrient requirements per unit area of ​​crops, and maximum safe water accumulation depth of liquid fertilizer corresponding to each farmland grid according to the farmland grid identifier.

3. The method for optimizing crop-livestock synergy based on spatial distribution characteristic analysis according to claim 1, characterized in that, The process involves performing closed depression hydrological topology identification processing on the high-precision digital elevation model data corresponding to the farmland grid to obtain basic feature data of local closed depressions, including: For any target farmland grid, extract the high-precision digital elevation model data corresponding to the target farmland grid from the farmland grid basic dataset, and perform cropping processing on the high-precision digital elevation model data based on the two-dimensional spatial boundary of the target farmland grid to obtain the elevation raster data inside the target farmland grid. For any target elevation raster in the elevation raster data within the target farmland grid, the elevation data of the target elevation raster is compared with that of the adjacent elevation raster. The adjacent elevation raster with the largest elevation drop is taken as the flow direction pointing raster of the target elevation raster. Continuous flow direction tracking is performed based on the flow direction pointing raster corresponding to all elevation rasters to obtain the surface runoff path data within the target farmland grid. For any surface runoff path, when the surface runoff path converges in the low-elevation area inside the target farmland grid and does not directly connect to the two-dimensional spatial boundary of the target farmland grid, the low-elevation area is regarded as a local closed depression, and the number of local closed depressions is obtained based on the number of locally closed depressions identified. For any locally closed depression, extract the elevation data corresponding to each elevation grid within the coverage area of ​​the locally closed depression from the elevation grid data inside the target farmland grid. Take the minimum elevation data within the coverage area of ​​the locally closed depression as the elevation data of the lowest point of the locally closed depression. Based on the boundary connection relationship between the locally closed depression and the surrounding elevation grid, obtain the lowest boundary elevation data of the liquid fertilizer overflowing from the locally closed depression, and take the lowest boundary elevation data as the elevation data of the overflow point of the locally closed depression. For any locally closed depression, obtain all elevation grids that ultimately flow to the locally closed depression based on the surface runoff path data, and use the sum of the surface area of ​​all elevation grids that ultimately flow to the locally closed depression as the catchment area data of the locally closed depression. For any local closed depression, multiple elevation levels are set between the elevation data of the lowest point of the local closed depression and the elevation data of the overflow point of the local closed depression; for any elevation level, the summation of the area of ​​continuous elevation grids within the coverage of the local closed depression whose elevation data is less than or equal to the elevation level is used as the area data enclosed by the closed contour lines corresponding to the elevation level. By associating the data on the number of locally closed depressions corresponding to the target farmland grid, as well as the elevation data of the lowest point of each locally closed depression, the elevation data of the overflow point of each locally closed depression, the area data of the catchment area of ​​each locally closed depression, and the area data enclosed by the closed contour lines, the basic characteristic data of the locally closed depressions are obtained.

4. The method for optimizing crop-livestock synergy based on spatial distribution characteristic analysis according to claim 1, characterized in that, The process involves spatially overlaying and analyzing the basic characteristic data of locally closed depressions with high-precision digital elevation model data to obtain potential water accumulation depth data for these depressions, including: For any target farmland grid, extract the high-precision digital elevation model data corresponding to the target farmland grid from the farmland grid basic dataset, and extract the coverage data of each local closed depression corresponding to the target farmland grid and the elevation data of the overflow point of each local closed depression corresponding to the local closed depression from the local closed depression basic feature data. For any locally closed depression, based on the coverage data of the locally closed depression, the actual ground elevation data corresponding to each spatial coordinate within the coverage area of ​​the locally closed depression is extracted from the high-precision digital elevation model data corresponding to the target farmland grid. For any target spatial coordinate within the coverage area of ​​any locally closed depression, the difference between the elevation data of the overflow point of the locally closed depression and the actual ground elevation data corresponding to the target spatial coordinate is used as the initial potential water accumulation depth data corresponding to the target spatial coordinate. When the initial potential water depth data corresponding to the target spatial coordinates is less than the constant 0, the potential water depth data of the local closed depression corresponding to the target spatial coordinates is set to the constant 0; when the initial potential water depth data corresponding to the target spatial coordinates is greater than or equal to the constant 0, the initial potential water depth data corresponding to the target spatial coordinates is used as the potential water depth data of the local closed depression corresponding to the target spatial coordinates. The potential water accumulation depth data of the local closed depressions within the coverage area of ​​each local closed depression in the target farmland grid are correlated to obtain the potential water accumulation depth data of the local closed depressions corresponding to the target farmland grid.

Images