Combination explosion problem-oriented agricultural management measure space continuous optimization configuration method

Abstract

The invention discloses an agricultural management measure space continuous optimization configuration method for a combinatorial explosion problem, and relates to the technical field of agricultural resource management, and the method comprises the steps: obtaining the feature data of a plurality of to-be-configured space units in a target region, and dividing the plurality of to-be-configured space units into a plurality of homogeneous clusters; for each homogeneous cluster, a continuous proportion decision variable is defined, the value range of the continuous proportion decision variable is [0, 1], and a target function model comprising at least two optimization targets which conflict with each other is constructed; performing optimization solution on the objective function model by adopting a multi-objective evolutionary algorithm to obtain a Pareto optimal solution set; and selecting an optimal solution from the Pareto optimal solution set, and generating a spatial decision scheme based on the continuous proportion decision variable value of each homogeneous cluster in the optimal solution. According to the method, the problem of large-scale space optimization is feasible in calculation, the requirements of decision makers of different levels are met, and the practicability and operability of the result are enhanced.

Description

Technical Field

[0001] This invention relates to the fields of agricultural resource management, spatial optimization and decision-making technology, and in particular to a method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems. Background Technology

[0002] Faced with the dual challenges of global food security and climate change, promoting water-saving irrigation and optimized fertilization are crucial for achieving a green transformation in agriculture. However, determining "where to implement and how much to implement" on a national or regional scale is a complex problem of spatial resource optimization. Traditional decision-making methods often employ uniform promotion or simple zoning based on expert experience, making it difficult to quantitatively weigh multiple objectives such as increased production, emission reduction, and cost savings, thus failing to achieve systematic optimization of overall benefits.

[0003] A more scientific approach is to construct a spatial explicit optimization model. Current technologies typically treat each independent spatial unit (such as a grid or field) as a binary decision variable (implement or not implement), thus forming a combinatorial optimization problem. When the number of spatial units reaches tens or even hundreds of thousands (e.g., a 1 km grid scale across the entire country), the independent "yes / no" decision for each unit will lead to an exponential increase in the number of possible combinations of solutions (reaching 2^35). N This leads to a "combinatorial explosion," where problems of immense scale occur. For such large-scale combinatorial optimization problems, even with modern computers, exhaustively enumerating all possible solutions to find the optimal solution is completely infeasible within an acceptable timeframe. While existing technologies employ multi-objective evolutionary algorithms (such as NSGA-II) to solve complex optimization problems have advantages, their direct application to binary decision-making problems involving massive spatial units is still limited by the computational bottleneck caused by the combinatorial explosion. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a spatially continuous optimization configuration method for agricultural management measures for combinatorial explosion problems, which solves the computational bottleneck caused by combinatorial explosion when existing methods are directly applied to binary decision problems with massive spatial units.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a spatially continuous optimization configuration method for agricultural management measures for combinatorial explosion problems, which includes the following steps:

[0008] S1. Obtain feature data of multiple spatial units to be configured within the target area, and use a clustering algorithm to divide the multiple spatial units to be configured into several homogeneous clusters;

[0009] S2. For each homogeneous cluster, define a continuous proportional decision variable, the value range of which is [0,1], and construct an objective function model containing at least two conflicting optimization objectives;

[0010] S3. The objective function model is optimized and solved using a multi-objective evolutionary algorithm to obtain a Pareto optimal solution set;

[0011] S4. Select an optimal solution from the Pareto optimal solution set, and generate a spatial decision scheme based on the continuous proportional decision variable values ​​of each homogeneous cluster in the optimal solution.

[0012] As a preferred embodiment of the spatial continuous optimization configuration method for agricultural management measures for the combinatorial explosion problem described in this invention, the feature data includes environmental factors, management measure factors, and expected benefit potential factors.

[0013] As a preferred embodiment of the spatial continuous optimization configuration method for agricultural management measures for the combinatorial explosion problem described in this invention, the optimization objective has three aspects: maximizing the total agronomic benefits, maximizing the total environmental benefits, and minimizing the total implementation cost.

[0014] As a preferred embodiment of the spatial continuous optimization allocation method for agricultural management measures addressing the combinatorial explosion problem described in this invention, the specific steps for the production space decision-making scheme are as follows:

[0015] S401. Assign the optimal continuous proportional decision variable value of each homogeneous cluster to all spatial units within that cluster to generate a first raster layer. The value of each pixel in this layer represents the priority of implementing the agricultural management measures at its location.

[0016] S402. For each homogeneous cluster, determine the number of spatial units within the cluster that require the implementation of the agricultural management measures based on the optimal continuous ratio decision variable value, and randomly select the corresponding number of units from all spatial units within the cluster, mark them as "recommended conversion", and generate the second raster layer.

[0017] As a preferred embodiment of the spatial continuous optimization configuration method for agricultural management measures for the combinatorial explosion problem described in this invention, step S2 specifically involves:

[0018] S201. Define the decision variable as an n-dimensional vector X, and calculate it as follows.

[0019] ;

[0020] Where, x i∈ [0,1] represents the proportion of grid cells in the i-th cluster recommended for conversion to water-saving irrigation, 1≤i≤n;

[0021] S202. Construct an objective function that maximizes the gain in total national output.

[0022] ,

[0023] Among them, Y i Let N be the yield gain (t) of the representative grid of the i-th cluster. i The total number of graticles contained in the i-th cluster;

[0024] Construct an objective function to maximize the total national GHG emission reduction.

[0025] ,

[0026] ,

[0027] Among them, G i ΔCH4 represents the GHG emission reduction potential (tCO2e), and ΔCH4 represents the CH4 emission reduction (tCH4 / km). 2 ΔN2O is the change in N2O (t N2O / km) 2 (a) and (b) represent the global warming potential of CH4 and N2O, respectively.

[0028] Construct an objective function that minimizes the total cost of nationwide promotion.

[0029] ,

[0030] Where A = 100 hectares / grid, which is the area of ​​each grid; C = 1000 yuan / hectare, which is the cost of promoting water-saving irrigation per unit area;

[0031] S203. Set constraints.

[0032] ,

[0033] in, This represents the floor operation, where c and d represent the promotion ratio at the current baseline level and the theoretical upper limit of the promotion ratio, respectively.

[0034] As a preferred embodiment of the spatial continuous optimization configuration method for agricultural management measures oriented towards the combinatorial explosion problem described in this invention, step S3 specifically involves:

[0035] S301. Using the NSGA-II algorithm on an n-dimensional continuous hypercube [0,1] n Search within the space;

[0036] S302. Set the algorithm parameters, including population size m, number of generations n, crossover probability, mutation probability, crossover distribution index, mutation distribution index, and penalty coefficient;

[0037] S303. Obtain the Pareto front containing m non-dominated solutions, and select the solution with the highest comprehensive score as the optimal solution using the standardized scoring method. .

[0038] As a preferred embodiment of the spatial continuous optimization allocation method for agricultural management measures for the combinatorial explosion problem described in this invention, the standardized scoring method is specifically as follows:

[0039] First, the range standardization process is performed on the three objective values ​​in the Pareto front.

[0040] = , k = 1, 2;

[0041] = ;

[0042] in, and These represent the standardized production gain and GHG emission reduction, respectively. Indicate the cost-effectiveness of standardization;

[0043] Then calculate the overall score for each Pareto solution.

[0044] Score = ( + + ) ÷ 3;

[0045] The solution with the highest overall score is selected as the final recommended solution, and its corresponding clustering transformation ratio vector is denoted as . each It is the i-th clustering transformation ratio under the optimal solution.

[0046] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in the first aspect of the present invention.

[0047] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in the first aspect of the present invention.

[0048] The beneficial effects of this invention are as follows: By combining clustering and continuous proportional decision-making, this invention expands the search space from a discrete, combinatorially exploding 2-1 search space. N The space is transformed into a continuous, efficiently searchable [0,1] region. n The spatial model (where n is the number of clusters, n << N) makes large-scale spatial optimization problems computationally feasible. Utilizing a multi-objective evolutionary algorithm, it explores the Pareto front in a continuous space, clearly revealing the trade-offs between agronomical, environmental, and economic benefits, supporting scientific decision-making. Simultaneously, it produces a "continuous priority map" to guide macro-level planning and an "irrigation type decision classification map" to guide micro-level implementation, meeting the needs of decision-makers at different levels and enhancing the practicality and operability of the results. It can be applied to any agricultural and environmental management field involving large-scale spatial unit resource allocation and multi-objective trade-offs, such as promoting water-saving irrigation, optimizing fertilization, variety layout, and delineating ecological protection red lines. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart of the spatial continuous optimization configuration method for agricultural management measures for the combinatorial explosion problem in Example 1.

[0051] Figure 2 In Example 2, the three-dimensional Pareto front plot and its two-dimensional projection obtained by the NSGA-II algorithm illustrate the trade-off between production gain, greenhouse gas emission reduction, and implementation cost; A shows the three-dimensional Pareto front plot illustrating the trade-off between the three objectives; B shows the trade-off between production gain and greenhouse gas emission reduction; C shows the trade-off between production gain and implementation cost; D shows the trade-off between greenhouse gas emission reduction and implementation cost; E shows the parallel coordinate plot of the normalized target values.

[0052] Figure 3 The following is a diagram showing the spatial decision-making scheme results generated based on the optimal Pareto solution in Example 2: A. Continuous priority spatial distribution diagram (macro-strategic diagram); B. Irrigation type decision classification diagram (micro-implementation diagram). Detailed Implementation

[0053] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0054] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0055] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0056] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a spatially continuous optimization configuration method for agricultural management measures for combinatorial explosion problems, including the following steps:

[0057] S1. Obtain feature data of multiple spatial units to be configured within the target area. The feature data includes environmental factors, management measure factors, and expected benefit potential factors. Use a clustering algorithm to divide the multiple spatial units to be configured into several homogeneous clusters.

[0058] S2. For each homogeneous cluster, define a continuous proportional decision variable to represent the proportion of spatial units within that cluster that are recommended to implement the target agricultural management measures; construct an objective function model containing at least two conflicting optimization objectives; specifically,

[0059] S201. Define the decision variable as an n-dimensional vector X, and calculate it as follows.

[0060] ;

[0061] Where, x i ∈ [0,1] represents the proportion of grid cells in the i-th cluster recommended for conversion to water-saving irrigation, 1≤i≤n;

[0062] S202. Construct an objective function that maximizes the gain in total national output.

[0063] ,

[0064] Among them, Y i Let N be the yield gain (t) of the representative grid of the i-th cluster. i The total number of graticles contained in the i-th cluster;

[0065] Construct an objective function to maximize the total national GHG emission reduction.

[0066] ,

[0067] ,

[0068] Among them, G i ΔCH4 represents the GHG emission reduction potential (tCO2e), and ΔCH4 represents the CH4 emission reduction (tCH4 / km). 2 ΔN2O is the change in N2O (t N2O / km) 2 (a) and (b) represent the global warming potential of CH4 and N2O, respectively.

[0069] Construct an objective function that minimizes the total cost of nationwide promotion.

[0070] ,

[0071] Where A = 100 hectares / grid, which is the area of ​​each grid; C = 1000 yuan / hectare, which is the cost of promoting water-saving irrigation per unit area;

[0072] S203. Set constraints.

[0073] ,

[0074] in, This represents the floor operation, where c and d represent the promotion ratio at the current baseline level and the theoretical upper limit of the promotion ratio, respectively.

[0075] S3. The objective function model is optimized and solved using a multi-objective evolutionary algorithm to obtain a Pareto optimal solution set, specifically:

[0076] S301. Using the NSGA-II algorithm on an n-dimensional continuous hypercube [0,1] n Search within the space;

[0077] S302. Set the algorithm parameters, including population size m, number of generations n, crossover probability, mutation probability, crossover distribution index, mutation distribution index, and penalty coefficient;

[0078] S303. Obtain the Pareto front containing m non-dominated solutions, and select the solution with the highest comprehensive score as the optimal solution using the standardized scoring method. .

[0079] The standardized scoring method is as follows:

[0080] First, the range standardization process is performed on the three objective values ​​in the Pareto front.

[0081] = , k = 1, 2;

[0082] = ;

[0083] in, and These represent the standardized production gain and GHG emission reduction, respectively. Indicate the cost-effectiveness of standardization;

[0084] Then calculate the overall score for each Pareto solution.

[0085] Score = ( + + ) ÷ 3;

[0086] The solution with the highest overall score is selected as the final recommended solution, and its corresponding clustering transformation ratio vector is denoted as . each It is the i-th clustering transformation ratio under the optimal solution.

[0087] S4. Select an optimal solution from the Pareto optimal solution set, and generate a spatial decision scheme based on the continuous proportional decision variable values ​​of each homogeneous cluster in the optimal solution, specifically including:

[0088] S401. Assign the optimal continuous proportional decision variable value of each homogeneous cluster to all spatial units within that cluster to generate a first raster layer. The value of each pixel in this layer represents the priority of implementing the agricultural management measures at its location.

[0089] S402. For each homogeneous cluster, determine the number of spatial units within the cluster that require the implementation of the agricultural management measures based on the optimal continuous ratio decision variable value, and randomly select the corresponding number of units from all spatial units within the cluster, mark them as "recommended conversion", and generate the second raster layer.

[0090] This invention creatively uses clustering and continuous proportional decision-making to transform the discrete combinatorial explosion space into a continuous and efficient searchable space, making large-scale spatial optimization problems computationally feasible. Then, it uses a multi-objective evolutionary algorithm to explore the Pareto front in the continuous space, clearly revealing the trade-offs between multiple optimization objectives, generating a continuous priority map and a decision classification map to meet the needs of different decision-makers.

[0091] Example 2: This example is based on Example 1. The difference between this example and Example 1 is that this example uses a specific example (promotion of water-saving irrigation in paddy fields in China) to further illustrate this application.

[0092] A spatially continuous optimization method for promoting water-saving irrigation measures in Chinese paddy fields to address the combinatorial explosion problem includes the following steps:

[0093] S1. Data Preparation: Collect environmental and production characteristic data from all rice planting grids nationwide (a total of 284,789 valid grids, of which 157,417 are currently flooded irrigation grids and are to be optimized and transformed), including: 1) Environmental factors: annual mean temperature, annual precipitation, soil organic carbon, total nitrogen content, pH value, soil bulk density, and clay content; 2) Management measures: nitrogen application rate; 3) Conversion potential: yield gain (ΔYield), CH4 emission reduction potential (ΔCH4), and total GHG emission reduction potential (ΔGHG). Although the GHG emission reduction potential already includes the contribution of CH4, CH4 is the most important greenhouse gas in paddy fields, and its individual changes have important guiding significance for management decisions. Therefore, the CH4 emission reduction potential was retained separately in this step. In addition, the N2O change was excluded, mainly to avoid collinearity: the N2O change has been integrated into the GHG emission reduction potential through weighted calculation.

[0094] Cluster analysis: For 157,417 flood irrigation grids to be modified, the Mini-batch K-means algorithm was used to divide them into 500 homogeneous clusters based on the 11 features mentioned above. The algorithm parameters were set to 500 clusters, batch size of 1000, 5 initializations, and a maximum of 100 iterations. The explained variance of the clustering results reached 0.999, indicating good clustering performance. The cluster ID of each grid was recorded.

[0095] S2 includes:

[0096] S201. Define the decision variable: The decision variable is a 500-dimensional vector X, calculated as follows.

[0097] ;

[0098] Where, x i ∈ [0,1] represents the proportion of rasters in the i-th cluster that are recommended to be converted to water-saving irrigation. For example, x3=0.7 means that 70% of the rasters in the 3rd cluster are recommended to be converted from flood irrigation to water-saving irrigation, while the remaining 30% remain as they are.

[0099] S202. Constructing the objective function:

[0100] Objective 1 (Maximize the gain in total national output) ),

[0101]

[0102] Among them, Y i Let N be the yield gain (t) of the representative grid of the i-th cluster. i Let be the total number of grid cells contained in the i-th cluster, and the spatial resolution of each grid cell is preferably 1 km;

[0103] Objective 2 (Maximize total national GHG emission reductions) ),

[0104] ,

[0105] ,

[0106] Among them, G i ΔCH4 represents the GHG emission reduction potential (tCO2e), and ΔCH4 represents the CH4 emission reduction (tCH4 / km). 2 ΔN2O is the change in N2O (t N2O / km) 2 The coefficients 27.9 and 273 represent the 100-year global warming potential of CH4 and N2O, respectively (IPCC, 2021).

[0107] Objective 3 (Minimize total national promotion cost) ),

[0108] ,

[0109] Where A = 100 hectares / grid, is the area of ​​each grid; C = 1000 yuan / hectare, is the cost of promoting water-saving irrigation per unit area; therefore, the cost per grid = 100 hectares × 1000 yuan / hectare = 100,000 yuan;

[0110] S203, Set constraints

[0111]

[0112] in, This represents the floor function. This constraint limits the feasible range of the national water-saving irrigation promotion rate: the final promotion rate (the ratio of the sum of baseline water-saving area and newly converted area to the total rice area) must be between the current baseline level of 44.73% and the theoretical upper limit of 90% (the upper and lower limits in this constraint are given with reference to the paper: Xu, Q., Ao, D., Abdo, AI, Chen, J., Chen, X., Liang, H., Jan, K., Groenigen, V., & Jiang, Y. (2025). Scaling up water-saving irrigation reduces yield-scaled greenhouse gas emissions by over one-quarter in China's ricepaddies. Agricultural Water Management, 321, 109937.).

[0113] S3. Continuous space optimization solution:

[0114] S301. Using the NSGA-II algorithm on a 500-dimensional continuous hypercube [0,1] 500 Search within the space;

[0115] S302. Set the algorithm parameters: population size 100, number of generations 500, crossover probability 0.8, mutation probability 0.1, crossover distribution index 20, mutation distribution index 20, penalty coefficient 10. 6 ;

[0116] S303, Execution Results: Obtained a Pareto front containing 100 non-dominated solutions (e.g., Figure 2 As shown in the figure, the solution with the highest comprehensive score is selected as the optimal solution using a standardized scoring method. .

[0117] The standardized scoring method is as follows:

[0118] First, the three objectives in the Pareto front are normalized using range standardization. For the maximization objectives (yield gain and GHG emission reduction):

[0119] = , k = 1, 2;

[0120] = ;

[0121] Where f̂1 and f̂2 represent the standardized production gain and GHG emission reduction, respectively, and f̂3 represents the standardized cost-effectiveness. This transformation ensures that larger standardized objective values ​​indicate better performance, and that the values ​​are uniformly within the range of [0,1]. Then, the overall score for each Pareto solution is calculated:

[0122] .

[0123] The solution with the highest overall score is selected as the final recommended solution, and its corresponding clustering transformation ratio vector is denoted as . . It represents the "optimal vector" or "optimal solution". This represents the 500 components of the vector, each It represents the i-th clustering transformation ratio under the optimal solution. This standardization method ensures the comparability of three objectives with different dimensions and optimization directions, and that they contribute fairly to the overall score.

[0124] S4. Spatial Decision-Making Scheme Generation:

[0125] S401, Generate a continuous priority map ( Figure 3 A):

[0126] Create a blank raster layer R priority ;

[0127] For each flooded irrigation grid cell, based on its cluster ID (denoted as k), from the optimal solution vector Extract the optimal conversion rate corresponding to this cluster. And assign this value (between 0 and 1) to R. priority The corresponding grid position;

[0128] The graph shows that high values ​​indicate high-priority promotion areas.

[0129] S402, Generate an irrigation type decision classification map ( Figure 3 B):

[0130] For each cluster i, calculate the actual number of raster cells n that need to be converted. i ;

[0131] Randomly select n from all the grids of cluster i. i One is marked as "Recommended Conversion" (value=1), and the rest are marked as "Maintain Status Quo" (value=0);

[0132] Based on the baseline water-saving areas, a final irrigation type decision classification map is generated;

[0133] S403, The national weighted average promotion ratio (P) corresponding to the optimal solution nationalThe calculation is as follows:

[0134] ,

[0135] Where, N total The total number of flood irrigation grids nationwide is represented by this indicator, which comprehensively reflects the overall coverage level achievable through water-saving irrigation under spatially differentiated strategies.

[0136] S5. Summary of Macroeconomic Benefits: Based on Figure 3 The irrigation type decision classification diagram for B shows that, under the water-saving irrigation promotion plan with a coverage rate of 80.3%, the total national rice production can be increased by about 6.47 million tons, greenhouse gas emissions can be reduced by about 63.5 million tons of CO2e, and the total promotion cost is about 101.4 billion yuan.

[0137] This invention uses the promotion of water-saving irrigation in Chinese paddy fields as an example to specifically demonstrate the algorithm's effectiveness. For example... Figure 2 As shown, NSGA-II operates in a continuous decision space ([0,1]). 500 The algorithm was successfully solved in the context of production, emissions reduction, and cost three-dimensional Pareto fronts, demonstrating its effectiveness in multi-objective trade-offs. Figure 3 This demonstrates the algorithm's two-level spatial output: Figure 3 A (continuous priority map) visually reveals that the main rice-producing areas in East and Central China are high-priority promotion areas (high values), while Figure 3 B (Implementation Classification Diagram) further provides specific, implementable grid-level recommendation schemes. Results show that the algorithm-recommended schemes can achieve a synergistic benefit of increasing production by 6.47 million tons and reducing CO2e emissions by 63.5 million tons, with a nationwide adoption rate of 80.3%, verifying the feasibility and superiority of this method in solving ultra-large-scale spatial configuration problems.

[0138] This embodiment also provides a computer device applicable to the case of a method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as proposed in the above embodiment.

[0139] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0140] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the spatially continuous optimization configuration method for agricultural management measures oriented towards the combinatorial explosion problem as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0141] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A spatially continuous optimization method for agricultural management measures oriented towards the combinatorial explosion problem, characterized in that: include, S1. Obtain feature data of multiple spatial units to be configured within the target area, and use a clustering algorithm to divide the multiple spatial units to be configured into several homogeneous clusters; S2. For each homogeneous cluster, define a continuous proportional decision variable, the value range of which is [0,1], and construct an objective function model containing at least two conflicting optimization objectives; S3. The objective function model is optimized and solved using a multi-objective evolutionary algorithm to obtain a Pareto optimal solution set; S4. Select an optimal solution from the Pareto optimal solution set, and generate a spatial decision scheme based on the continuous proportional decision variable values ​​of each homogeneous cluster in the optimal solution.

2. The method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in claim 1, characterized in that: The characteristic data includes environmental factors, management measures factors, and expected benefit potential factors.

3. The method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in claim 1, characterized in that: The optimization objectives are threefold: maximizing the total agronomic benefits from implementing the agricultural management measures, maximizing the total environmental benefits, and minimizing the total implementation costs.

4. The spatial continuous optimization configuration method for agricultural management measures for the combinatorial explosion problem as described in claim 3, characterized in that: The specific steps for production space decision-making are as follows: S401. Assign the optimal continuous proportional decision variable value of each homogeneous cluster to all spatial units within that cluster to generate a first raster layer. The value of each pixel in this layer represents the priority of implementing the agricultural management measures at its location. S402. For each homogeneous cluster, determine the number of spatial units within the cluster that require the implementation of the agricultural management measures based on the optimal continuous ratio decision variable value, and randomly select the corresponding number of units from all spatial units within the cluster, mark them as "recommended conversion", and generate the second raster layer.

5. The method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in claim 4, characterized in that: Step S2 specifically involves: S201. Define the decision variable as an n-dimensional vector X, and calculate it as follows. ； Where, x i ∈ [0,1] represents the proportion of grid cells in the i-th cluster recommended for conversion to water-saving irrigation, 1≤i≤n; S202. Construct an objective function that maximizes the gain in total national output. ， Among them, Y i Let N be the yield gain (t) of the representative grid of the i-th cluster. i The total number of rasters contained in the i-th cluster; Construct an objective function to maximize the total national GHG emission reduction. ， ， Among them, G i ΔCH4 represents the GHG emission reduction potential (tCO2e), and ΔCH4 represents the CH4 emission reduction (tCH4 / km). 2 ΔN2O is the change in N2O (tN2O / km) 2 (a) and (b) represent the global warming potential of CH4 and N2O, respectively. Construct an objective function that minimizes the total cost of nationwide promotion. ， Where A = 100 hectares / grid, which is the area of ​​each grid; C = 1000 yuan / hectare, which is the cost of promoting water-saving irrigation per unit area; S203. Set constraints. ， in, This represents the floor operation, where c and d represent the promotion ratio at the current baseline level and the theoretical upper limit of the promotion ratio, respectively.

6. The method for spatially continuous optimization of agricultural management measures for combinatorial explosion problems as described in claim 5, characterized in that: Step S3 specifically involves: S301. Using the NSGA-II algorithm on an n-dimensional continuous hypercube [0,1] n Search within the space; S302. Set the algorithm parameters, including population size m, number of generations n, crossover probability, mutation probability, crossover distribution index, mutation distribution index, and penalty coefficient; S303. Obtain the Pareto front containing m non-dominated solutions, and select the solution with the highest comprehensive score as the optimal solution using the standardized scoring method. .

7. The method for spatially continuous optimization of agricultural management measures for combinatorial explosion problems as described in claim 6, characterized in that: The standardized scoring method is as follows: First, the range standardization process is performed on the three objective values ​​in the Pareto front. = , k = 1, 2； = ； in, and These represent the standardized production gain and GHG emission reduction, respectively. Indicate the cost-effectiveness of standardization; Then calculate the overall score for each Pareto solution. Score = ( + + ) ÷ 3; The solution with the highest overall score is selected as the final recommended solution, and its corresponding clustering transformation ratio vector is denoted as . Each It is the i-th clustering transformation ratio under the optimal solution.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for continuous spatial optimization of agricultural management measures for combinatorial explosion problems as described in any one of claims 1 to 7.

Images