A genetic optimization space layout method and system for a cultivation monitoring station

Abstract

The application belongs to the technical field of agricultural information and resource monitoring, and discloses a genetic optimization space layout method and system for cultivated land monitoring stations, which comprises the following steps: acquiring geographic information and soil index data of the cultivated land monitoring stations, and performing normalization processing on the data; constructing a spatial heterogeneity comprehensive index, fusing multi-index weights based on an entropy weight method, and calculating the spatial heterogeneity comprehensive index of each cultivated land monitoring station; performing heterogeneity grade division of the cultivated land monitoring stations according to the quantile of the spatial heterogeneity comprehensive index, and generating a Thiessen polygon quantification monitoring range; and optimizing the layout of the cultivated land monitoring stations through a genetic algorithm, and iteratively selecting an optimal cultivated land monitoring station space layout scheme. The genetic optimization space layout method and system for cultivated land monitoring stations are adopted, various key indexes of soil and spatial distribution characteristics of the monitoring stations are comprehensively considered, a spatial heterogeneity comprehensive index is constructed, and the layout optimization of the monitoring stations is realized.

Description

Technical Field

[0001] This invention relates to the fields of agricultural information and agricultural resource monitoring technology, and in particular to a genetically optimized spatial layout method and system for arable land monitoring stations. Background Technology

[0002] With the continuous advancement of agricultural modernization in my country, farmland quality monitoring plays an increasingly important role in agricultural production. Soil data obtained through monitoring stations can provide a scientific basis for agricultural production and food security. However, the current layout of monitoring stations has many problems, mainly reflected in the unreasonable location of the stations. Most existing monitoring station layouts are selected based on experience or simple geographical zoning, lacking comprehensive consideration of soil characteristics and spatial heterogeneity. This layout method results in a large number of blind spots or overlapping coverage areas, insufficient monitoring accuracy, and an excessive number of monitoring stations, leading to resource waste. These problems not only reduce monitoring efficiency but also increase monitoring costs and affect the accuracy and reliability of monitoring data.

[0003] Currently, domestic and international research on the deployment of farmland quality monitoring stations mainly focuses on the construction of monitoring indicator systems, the improvement of monitoring technologies, and the preliminary layout of monitoring stations. However, most of these studies only focus on monitoring single indicators or local areas, lacking a comprehensive consideration of the integrated characteristics and spatial heterogeneity of soil multiple indicators. In practical applications, the layout of monitoring stations mostly relies on experience or simple geographical zoning, failing to fully integrate the spatial distribution characteristics and degree of heterogeneity of soil. This experience-based layout method often cannot accurately identify highly heterogeneous areas, resulting in insufficient representativeness of monitoring data and difficulty in accurately reflecting the true state of farmland quality in the region. Furthermore, existing research is mostly theoretical, with few practically implemented schemes and limited applicability, making it difficult to meet the needs of large-scale farmland monitoring.

[0004] Therefore, there is an urgent need for a method to optimize the layout of monitoring stations that can comprehensively consider the spatial heterogeneity of multiple soil indicators, so as to improve monitoring efficiency and data quality and provide a scientific basis for the accurate monitoring and management of arable land quality. Summary of the Invention

[0005] The purpose of this invention is to provide a genetically optimized spatial layout method and system for farmland monitoring stations. By comprehensively considering various key soil indicators and the spatial distribution characteristics of monitoring stations, a spatial heterogeneity comprehensive index is constructed to achieve optimized layout of monitoring stations.

[0006] To achieve the above objectives, this invention provides a genetically optimized spatial layout method for farmland monitoring stations, comprising the following steps:

[0007] Step S1, Data Acquisition and Preprocessing: Acquire geographic information and soil index data of farmland monitoring stations, and normalize the acquired data;

[0008] Step S2, Construction of Spatial Heterogeneity Comprehensive Index: Construct a spatial heterogeneity comprehensive index by integrating the weights of multiple indicators based on the entropy weight method, and calculate the spatial heterogeneity comprehensive index of each farmland monitoring station;

[0009] Step S3, Heterogeneity Classification and Spatial Coverage Modeling: Based on the quantiles of the spatial heterogeneity comprehensive index, the heterogeneity of farmland monitoring stations is classified, and the Thiessen polygon is used to quantify the monitoring range.

[0010] Step S4, Genetic Algorithm Optimization: The layout of farmland monitoring stations is optimized using a genetic algorithm, and the optimal spatial layout scheme for farmland monitoring stations is selected iteratively.

[0011] Preferably, in step S1, the geographical information and soil index data of the farmland monitoring stations are acquired, and the acquired data is normalized. The specific process is as follows:

[0012] Step S11: Obtain the geographic information of farmland monitoring stations, including latitude, longitude, and elevation, and use ArcGIS software to present the distribution of regional monitoring stations; soil index data include pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content;

[0013] Step S12: Normalize each soil index to eliminate the impact of differences in the dimensions and orders of magnitude of different soil index data, as shown below:

[0014] , =1,2,…,n; =1,2,…,m;

[0015] in, For the first The first cultivated land monitoring station The original values ​​of each soil index; For the first The set of raw values ​​of a soil index from all farmland monitoring stations; For the first The minimum value of a soil index among all farmland monitoring stations; For the first The maximum value of each soil index across all farmland monitoring stations; For the first The first cultivated land monitoring station Normalized values ​​of each soil index; This represents the total number of farmland monitoring stations. This represents the number of soil indicators.

[0016] Preferably, in step S2, a spatial heterogeneity comprehensive index is constructed. Based on the entropy weight method, the weights of multiple indicators are integrated to calculate the spatial heterogeneity comprehensive index of each farmland monitoring station. The specific process is as follows:

[0017] Step S21: Convert the soil index values ​​of each farmland monitoring station into relative weights, and calculate the first... The first soil index The proportion of individual farmland monitoring stations As shown below:

[0018] ;

[0019] Step S22: Based on the relative proportions of soil indicators at each farmland monitoring station, firstly, the dispersion of soil indicators is quantified using entropy values. Entropy values ​​of soil indicators As shown below:

[0020] ;

[0021] Next, weights are used to reflect the importance of each soil indicator in the comprehensive evaluation. Weights of each soil index As shown below:

[0022] ;

[0023] in, satisfy ;

[0024] Step S23: Based on the weights of each soil index, calculate the comprehensive index of spatial heterogeneity of soil indices at farmland monitoring stations. As shown below:

[0025] ;

[0026] in, Soil indicators for all farmland monitoring stations The mean; This is the magnification factor; This is the translation coefficient.

[0027] Preferably, in step S3, the heterogeneity level of farmland monitoring stations is classified according to the quantiles of the spatial heterogeneity composite index, and the Thiessen polygon quantification monitoring range is generated. The specific process is as follows:

[0028] Step S31: Based on the H value of the spatial heterogeneity comprehensive index, the quantile method is used to classify the farmland monitoring stations;

[0029] Step S32: Treat each farmland monitoring station as a geospatial point, and based on the latitude and longitude coordinates of that station, use the R language... sf The package generates Thiessen polygons to visualize the coverage of each farmland monitoring station.

[0030] Preferably, in step S31, the quantile method is used to classify the farmland monitoring stations, as shown below:

[0031] Farmland monitoring stations with H values ​​below the 50th percentile are marked as low heterogeneous points;

[0032] Farmland monitoring stations with H values ​​between the 50th and 75th percentiles are marked as heterogeneous points;

[0033] Farmland monitoring stations with H values ​​higher than the 75th percentile are marked as high heterogeneous sites.

[0034] Preferably, in step S4, the layout of farmland monitoring stations is optimized by a genetic algorithm, and the optimal spatial layout scheme of farmland monitoring stations is selected iteratively.

[0035] Genetic algorithm optimization includes scheme initialization, scheme encoding, fitness calculation, selection, crossover and mutation operations;

[0036] Step S41: Based on the existing farmland monitoring station layout, generate an initial scheme, with P schemes; Scheme 0 is the coordinate set of the original farmland monitoring stations:

[0037] ;

[0038] in, Indicates farmland monitoring stations latitude and longitude coordinates; Longitude coordinates; Latitude coordinates;

[0039] The remaining schemes are generated through difference perturbations: in probability Remove low heterogeneous points; New farmland monitoring stations are randomly generated within the Thiessen polygon with high heterogeneity. The newly added farmland monitoring stations are verified by GIS spatial verification to be located in farmland areas other than roads and water bodies. New farmland monitoring stations in non-farmland areas are deleted.

[0040] Step S42: Encode the layout scheme of the farmland monitoring stations, representing it as a real number vector; each station layout scheme is a sequence, as shown below:

[0041] ;

[0042] in, Used to identify the type of farmland monitoring station: 0 indicates low heterogeneity, 1 indicates medium heterogeneity, and 2 indicates high heterogeneity;

[0043] Step S43: For each layout scheme, calculate the fitness. To assess layout quality, as shown below:

[0044] ;

[0045] in, The total area of ​​the highly heterogeneous Thiessen polygon; The total area of ​​the region; The number of monitoring stations for low-heterogeneous cultivated land; The maximum distance from any point to the nearest farmland monitoring station; For reference distance; Weighting is applied to cover highly heterogeneous regions; Weighting based on the number of heterogeneous sites; Weights for spatial coverage uniformity;

[0046] Step S44: Perform parent selection and crossover variation in the farmland monitoring station layout scheme;

[0047] Step S45: Iterate the genetic algorithm and set the stopping condition for the genetic algorithm, that is, reach the maximum number of iterations or all farmland monitoring station layout schemes are the same, and complete the spatial optimization layout scheme of farmland monitoring stations.

[0048] Preferably, in step S44, parent selection and crossover mutation are performed in the farmland monitoring station layout scheme, and the specific process is as follows:

[0049] Step S441, for the first Each layout scheme is assigned a selection probability based on its fitness percentage, as shown below:

[0050] ;

[0051] in, Indicates the first The probability of selecting a layout scheme; Indicates the first The adaptability of each layout scheme; This represents the sum of fitness for all layout options; the selection probability of each layout option is sequentially mapped onto a cumulative probability axis in the interval [0,1], forming a virtual roulette wheel;

[0052] Step S442: Generate a uniform random number r in the interval [0,1], and select the corresponding scheme based on the roulette interval in which the value of r falls;

[0053] Step S443: Based on the crossover operator, randomly select two different parent schemes for crossover. For highly heterogeneous points, align the farmland monitoring stations of the same type in the two schemes according to the nearest neighbor matching principle of farmland monitoring stations by latitude and longitude. Perform simulated binary crossover on the latitude and longitude coordinates of all matched highly heterogeneous farmland monitoring stations, as shown below:

[0054] ;

[0055] in, and Indicates the offspring scheme; and Indicates the parent scheme; A random number between 0 and 1; The cross-distribution index; The cross-distribution coefficient;

[0056] Step S444: Based on the mutation operator, perform mutation operation on each child scheme:

[0057] Deletion mutation is probability Randomly delete a monitoring station for low-heterogeneous cultivated land, i.e. ;

[0058] New mutations are probabilities : Randomly add farmland monitoring stations in highly heterogeneous areas, with coordinates evenly distributed within highly heterogeneous Thiessen polygons; if the coordinates fall into non-farmland after mutation, regenerate or cancel the operation.

[0059] A genetically optimized spatial layout system for farmland monitoring stations includes the following modules:

[0060] The data acquisition and preprocessing module is used to collect the geographic information of the established farmland monitoring stations and the soil index data of each station, and to preprocess the data.

[0061] The geographic information includes latitude and longitude, and elevation; the soil index data includes pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content.

[0062] The spatial heterogeneity comprehensive index construction module determines the weights of each soil indicator at the monitoring station based on the entropy weight method, and constructs the spatial heterogeneity comprehensive index.

[0063] The heterogeneity classification and spatial coverage modeling module classifies the monitoring stations into heterogeneity levels based on the comprehensive spatial heterogeneity index of each monitoring station and its corresponding quantile, and quantifies the monitoring range using Thiessen polygons.

[0064] The genetic algorithm optimization module includes site scheme initialization, scheme encoding, fitness calculation, design of selection, crossover and mutation operators, and iterative selection of spatial optimization layout schemes for farmland monitoring sites.

[0065] Therefore, the present invention adopts the above-mentioned genetic optimization spatial layout method and system for farmland monitoring stations. By comprehensively considering multiple key soil indicators and the spatial distribution characteristics of monitoring stations, a spatial heterogeneity comprehensive index is constructed to achieve the optimization of the monitoring station layout.

[0066] This invention not only scientifically determines the location of monitoring stations based on soil data, avoiding resource waste caused by unreasonable station layout, but also achieves global optimization through genetic algorithms to ensure the optimal layout of monitoring stations.

[0067] Furthermore, this invention uses Thiessen polygon quantification to visually represent the coverage area of ​​each monitoring station, providing an important reference for the rational deployment of monitoring stations. This method improves monitoring accuracy in highly heterogeneous areas while reducing the number of inefficient stations, effectively lowering monitoring costs and increasing monitoring efficiency. This provides strong support for improving the quality of arable land and the efficient management of agricultural production in my country, contributing to the national strategy of food security.

[0068] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0069] Figure 1 This is a flowchart of a genetically optimized spatial layout method for farmland monitoring stations according to the present invention;

[0070] Figure 2 This is a distribution map of monitoring stations in the selected area of ​​this invention;

[0071] Figure 3 This is a diagram showing the heterogeneity classification results of the monitoring sites in this invention;

[0072] Figure 4 This is a map showing the monitoring range of the monitoring stations of this invention;

[0073] Figure 5 This is a flowchart illustrating the application of the genetic optimization algorithm in this invention.

[0074] Figure 6 This is a distribution map of monitoring stations optimized by a genetic algorithm according to the present invention;

[0075] Figure 7 This is a schematic diagram of a genetically optimized spatial layout system for farmland monitoring stations according to the present invention. Detailed Implementation

[0076] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0077] like Figure 1 As shown, a genetic optimization spatial layout method for farmland monitoring stations includes the following steps:

[0078] Step S1, Data Acquisition and Preprocessing: Acquire geographic information and soil index data of farmland monitoring stations, and normalize the acquired data;

[0079] Step S2, Construction of Spatial Heterogeneity Comprehensive Index: Construct a spatial heterogeneity comprehensive index by integrating the weights of multiple indicators based on the entropy weight method, and calculate the spatial heterogeneity comprehensive index of each farmland monitoring station;

[0080] Step S3, Heterogeneity Classification and Spatial Coverage Modeling: Based on the quantiles of the spatial heterogeneity comprehensive index, the heterogeneity of farmland monitoring stations is classified, and the Thiessen polygon is used to quantify the monitoring range.

[0081] Step S4, Genetic Algorithm Optimization: The layout of farmland monitoring stations is optimized using a genetic algorithm, and the optimal spatial layout scheme for farmland monitoring stations is selected iteratively.

[0082] Example 1

[0083] To address the current problems of irrational layout of farmland monitoring stations, overlapping coverage, insufficient monitoring accuracy, and resource waste due to an excessive number of stations, a genetic optimization spatial layout method for farmland monitoring stations is proposed. Balancing cost and efficiency, given a region and the number of monitoring stations within that region, the optimal layout of monitoring stations is quickly achieved based on the geographical information and soil index data of the stations. This ensures that the monitoring stations can cover all key soil characteristic areas of the region as much as possible, improving monitoring accuracy while reducing inefficiently covered areas.

[0084] This invention discloses a genetically optimized spatial layout method for farmland monitoring stations, comprising the following steps:

[0085] Step S1: Obtain the geographic information and soil index data of the farmland monitoring stations, and normalize the obtained data.

[0086] Step S11: Obtain the geographical information of farmland monitoring stations, including latitude, longitude, and elevation, and use ArcGIS software to present the distribution of regional monitoring stations; soil index data include pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content, etc.

[0087] In this embodiment, the geographic latitude and longitude information of some farmland monitoring stations in Gongzhuling is obtained, and ArcGIS software is used to present the distribution of farmland monitoring stations in the region, such as... Figure 2 As shown.

[0088] Step S12: Normalize each soil index to eliminate the impact of differences in the dimensions and orders of magnitude of different soil index data, as shown below:

[0089] , =1,2,…,n; =1,2,…,m;

[0090] in, For the first The first cultivated land monitoring station The original values ​​of each soil index; For the first The set of raw values ​​of a soil index from all farmland monitoring stations; For the first The minimum value of a soil index among all farmland monitoring stations; For the first The maximum value of each soil index across all farmland monitoring stations; For the first The first cultivated land monitoring station Normalized values ​​of each soil index; This represents the total number of farmland monitoring stations. This represents the number of soil indicators.

[0091] In this embodiment, 30 farmland monitoring stations were set up, and 5 soil indicators were established.

[0092] Step S2: Construct a spatial heterogeneity comprehensive index. Based on the entropy weight method, integrate the weights of multiple indicators to calculate the spatial heterogeneity comprehensive index of each farmland monitoring station.

[0093] Step S21: Convert the soil index values ​​of each farmland monitoring station into relative weights, and calculate the first... The first soil index The proportion of individual farmland monitoring stations As shown below:

[0094] ;

[0095] Step S22: Based on the relative proportions of soil indicators at each farmland monitoring station, firstly, the dispersion of soil indicators is quantified using entropy values. Entropy values ​​of soil indicators As shown below:

[0096] ;

[0097] Next, weights are used to reflect the importance of each soil indicator in the comprehensive evaluation. Weights of each soil index As shown below:

[0098] ;

[0099] in, satisfy .

[0100] Step S23: Based on the weights of each soil index, calculate the comprehensive index of spatial heterogeneity of soil indices at farmland monitoring stations. As shown below:

[0101] ;

[0102] in, Soil indicators for all farmland monitoring stations The mean; and This is an empirical scaling factor; This is an amplification factor to enhance the sensitivity of identifying highly heterogeneous regions; This is the translation coefficient to prevent interference from zero values.

[0103] Step S3: Based on the quantiles of the spatial heterogeneity comprehensive index, classify the heterogeneity levels of farmland monitoring stations and generate the Thiessen polygon quantitative monitoring range.

[0104] Step S31: Based on the H value of the spatial heterogeneity comprehensive index, the quantile method is used to classify the farmland monitoring stations, such as... Figure 3 As shown:

[0105] Farmland monitoring stations with H values ​​below the 50th percentile are marked as "low heterogeneous points";

[0106] Farmland monitoring stations with H values ​​between the 50th and 75th percentiles are marked as “medium heterogeneous sites”;

[0107] Farmland monitoring stations with H values ​​higher than the 75th percentile are marked as "high heterogeneous points".

[0108] Step S32: Treat each farmland monitoring station as a geospatial point, and based on the latitude and longitude coordinates of that station, use the R language... sf The package generates Thiessen polygons to visualize the coverage of each farmland monitoring station, such as... Figure 4 As shown.

[0109] Step S4: Optimize the layout of farmland monitoring stations using a genetic algorithm, iteratively selecting the optimal spatial layout scheme. The genetic algorithm optimization includes scheme initialization, scheme encoding, fitness calculation, selection, crossover, and mutation operations, such as... Figure 5 As shown.

[0110] Step S41: Scheme initialization.

[0111] Based on the existing layout of farmland monitoring stations, an initial scheme is generated, with P schemes, set to 200. Scheme 0 represents the coordinate set of the original farmland monitoring stations.

[0112] ;

[0113] in, Indicates farmland monitoring stations latitude and longitude coordinates; Longitude coordinates; These are latitude coordinates.

[0114] The remaining schemes are generated through difference perturbations: in probability Remove low heterogeneous points; New farmland monitoring stations are randomly generated within the Thiessen polygon with high heterogeneity. The newly added farmland monitoring stations are verified by GIS spatial verification to be located in farmland areas other than roads and water bodies. New farmland monitoring stations in non-farmland areas are deleted.

[0115] Step S42, Scheme coding.

[0116] The layout scheme of farmland monitoring stations is encoded and represented by a real-number vector. Each station layout scheme is a sequence, as shown below:

[0117] ;

[0118] in, Used to identify the type of farmland monitoring station: 0 indicates low heterogeneity, 1 indicates medium heterogeneity, and 2 indicates high heterogeneity.

[0119] Step S43: Fitness calculation.

[0120] For each layout scheme, calculate the fitness. To assess layout quality, as shown below:

[0121] ;

[0122] in, The total area of ​​the highly heterogeneous Thiessen polygon; The total area of ​​the region; The number of monitoring stations for low-heterogeneous cultivated land; The maximum distance from any point to the nearest farmland monitoring station; For reference distance; For high heterogeneous region coverage weight, Weighting for the number of heterogeneous sites. The weights for spatial coverage uniformity were all calibrated through pre-experiments.

[0123] Step S44: Perform parent selection and crossover variation in the farmland monitoring station layout scheme.

[0124] Step S441: First, for the first... Each layout scheme is assigned a selection probability based on its fitness percentage, as shown below:

[0125] ;

[0126] in, Indicates the first The probability of selecting a layout scheme; Indicates the first The adaptability of each layout scheme; Let P = 200, representing the sum of fitness for all layout options. The selection probability of each layout option is sequentially mapped onto a cumulative probability axis in the interval [0,1], forming a virtual "roulette".

[0127] Step S442: Next, generate a uniform random number r in the interval [0,1]. Based on the roulette interval in which the value of r falls, select the corresponding scheme. Repeat the random roulette selection 200 times, retain the best schemes to enter the next generation, and ensure the diversity of schemes.

[0128] Step S443: Then, based on the crossover operator, two different parent schemes are randomly selected for crossover. For highly heterogeneous points, farmland monitoring stations of the same type in the two schemes are aligned according to the nearest neighbor matching principle of farmland monitoring stations by latitude and longitude. Simulated binary crossover is performed on the latitude and longitude coordinates of all matched highly heterogeneous farmland monitoring stations, as shown below:

[0129] ;

[0130] in, and Indicates the offspring scheme; and Indicates the parent scheme; A random number between 0 and 1; The crossover index (usually 5) is used to control the degree of concentration of offspring individuals in the vicinity of their parents; The crossover coefficient is used to control the degree of deviation between offspring and parents.

[0131] Step S444: Finally, based on the mutation operator, perform a mutation operation on each child scheme:

[0132] Deletion mutation is probability Randomly delete one low-heterogeneity farmland monitoring station ( );

[0133] New mutations are probabilities : Randomly add farmland monitoring stations in highly heterogeneous areas, with coordinates evenly distributed within highly heterogeneous Thiessen polygons; if the coordinates fall into non-farmland after mutation, regenerate or cancel the operation.

[0134] Step S45: Iterate the genetic algorithm, setting the stopping condition for the algorithm to run, i.e., reaching the maximum number of iterations or all farmland monitoring station layout schemes being identical, to complete the optimal spatial optimization layout scheme for farmland monitoring stations, such as... Figure 6 As shown.

[0135] Example 2

[0136] like Figure 7 As shown, the present invention also provides a genetically optimized spatial layout system for farmland monitoring stations, comprising the following modules:

[0137] The data acquisition and preprocessing module is used to collect the geographic information of the established farmland monitoring stations and the soil index data of each station, and to preprocess the data.

[0138] The geographic information includes latitude and longitude, and elevation; the soil index data includes pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content.

[0139] The spatial heterogeneity comprehensive index construction module determines the weights of each soil indicator at the monitoring station based on the entropy weight method, and constructs the spatial heterogeneity comprehensive index.

[0140] The heterogeneity classification and spatial coverage modeling module classifies the monitoring stations into heterogeneity levels based on the comprehensive spatial heterogeneity index and its corresponding quantiles, and quantifies the monitoring range using Thiessen polygons.

[0141] The genetic algorithm optimization module includes site scheme initialization, scheme encoding, fitness calculation, design of selection, crossover and mutation operators, and iterative selection of the best spatial optimization layout scheme for farmland monitoring sites.

[0142] The specific implementation methods of each module and the corresponding steps are not described in this invention.

[0143] In summary, the present invention adopts the above-mentioned genetic optimization spatial layout method and system for farmland monitoring stations. By comprehensively considering multiple key soil indicators and the spatial distribution characteristics of monitoring stations, a spatial heterogeneity comprehensive index is constructed to achieve the optimization of the monitoring station layout.

[0144] This invention not only scientifically determines the location of monitoring stations based on soil data, avoiding resource waste caused by unreasonable station layout, but also achieves global optimization through genetic algorithms to ensure the optimal layout of monitoring stations.

[0145] Furthermore, this invention uses the Thiessen polygon visualization monitoring range to intuitively display the coverage area of ​​each monitoring station, providing an important reference for the rational layout of monitoring stations. This method improves the monitoring accuracy in highly heterogeneous areas while reducing the number of inefficient stations, effectively lowering monitoring costs and increasing monitoring efficiency, thus providing strong support for agricultural production and food security monitoring and management.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. 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 still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A genetically optimized spatial layout method for farmland monitoring stations, characterized in that, Includes the following steps: Step S1, Data Acquisition and Preprocessing: Acquire geographic information and soil index data of farmland monitoring stations, and normalize the acquired data; Step S2, Construction of Spatial Heterogeneity Composite Index: A spatial heterogeneity composite index is constructed by integrating multiple indicator weights using the entropy weight method. The specific process is as follows: Step S21: Convert the soil index values ​​of each farmland monitoring station into relative weights, and calculate the first... The first soil index The proportion of individual farmland monitoring stations As shown below: ； in, For the first The first cultivated land monitoring station Normalized values ​​of each soil index; Step S22: Based on the relative proportions of soil indicators at each farmland monitoring station, firstly, the dispersion of soil indicators is quantified using entropy values. Entropy values ​​of soil indicators As shown below: ； in, This represents the total number of farmland monitoring stations. Next, weights are used to reflect the importance of each soil indicator in the comprehensive evaluation. Weights of each soil index As shown below: ； in, satisfy ; The number of soil indicators; Step S23: Based on the weights of each soil index, calculate the comprehensive index of spatial heterogeneity of soil indices at farmland monitoring stations. As shown below: ； in, Soil indicators for all farmland monitoring stations The mean; This is the magnification factor; The translation coefficient; Step S3, Heterogeneity Classification and Spatial Coverage Modeling: Based on the quantiles of the spatial heterogeneity comprehensive index, the heterogeneity levels of farmland monitoring stations are classified, and the Thiessen polygon is used to quantify the monitoring range. The specific process is as follows: Step S31: Based on the H value of the spatial heterogeneity comprehensive index, the quantile method is used to classify the farmland monitoring stations; Step S32: Treat each farmland monitoring station as a geospatial point, and based on the latitude and longitude coordinates of that station, use the R language... sf The package generates Thiessen polygons to visualize the coverage of each farmland monitoring station; Step S4, Genetic Algorithm Optimization: The layout of farmland monitoring stations is optimized using a genetic algorithm, and the optimal spatial layout scheme of farmland monitoring stations is selected iteratively. Genetic algorithm optimization includes scheme initialization, scheme encoding, fitness calculation, selection, crossover and mutation operations; Step S41: Based on the existing farmland monitoring station layout, generate an initial scheme, with P schemes; Scheme 0 is the coordinate set of the original farmland monitoring stations: ； in, Indicates farmland monitoring stations latitude and longitude coordinates; Longitude coordinates; Latitude coordinates; The remaining schemes are generated through difference perturbations: in probability Remove low heterogeneous points; New farmland monitoring stations are randomly generated within the Thiessen polygon with high heterogeneity. The newly added farmland monitoring stations are verified by GIS spatial verification to be located in farmland areas that are not roads or water bodies. New farmland monitoring stations in non-farmland areas are deleted. Step S42: Encode the layout scheme of the farmland monitoring stations, representing it as a real number vector; each station layout scheme is a sequence, as shown below: ； in, Used to identify the type of farmland monitoring station: 0 indicates low heterogeneity, 1 indicates medium heterogeneity, and 2 indicates high heterogeneity; Step S43: For each layout scheme, calculate the fitness. To assess layout quality, as shown below: ； in, The total area of ​​the highly heterogeneous Thiessen polygon; The total area of ​​the region; The number of monitoring stations for low-heterogeneous cultivated land; The maximum distance from any point to the nearest farmland monitoring station; For reference distance; Weighting is applied to cover highly heterogeneous regions; Weighting based on the number of heterogeneous sites; Weights for spatial coverage uniformity; Step S44: Perform parent selection and crossover variation in the farmland monitoring station layout scheme; Step S45: Iterate the genetic algorithm and set the stopping condition for the genetic algorithm, that is, reach the maximum number of iterations or all farmland monitoring station layout schemes are the same, and complete the spatial optimization layout scheme of farmland monitoring stations.

2. The genetically optimized spatial layout method for farmland monitoring stations according to claim 1, characterized in that, In step S1, the geographical information and soil index data of the farmland monitoring stations are acquired, and the acquired data are normalized. The specific process is as follows: Step S11: Obtain the geographic information of farmland monitoring stations, including latitude, longitude, and elevation, and use ArcGIS software to present the distribution of regional monitoring stations; soil index data include pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content; Step S12: Normalize each soil index to eliminate the impact of differences in the dimensions and orders of magnitude of different soil index data, as shown below: , =1,2,…,n; =1,2,…,m; in, For the first The first cultivated land monitoring station The original values ​​of each soil index; For the first The set of raw values ​​of a soil index from all farmland monitoring stations; For the first The minimum value of a soil index among all farmland monitoring stations; For the first The maximum value of each soil index across all farmland monitoring stations; For the first The first cultivated land monitoring station Normalized values ​​of each soil index; This represents the total number of farmland monitoring stations. This represents the number of soil indicators.

3. The genetically optimized spatial layout method for farmland monitoring stations according to claim 1, characterized in that, In step S31, the quantile method is used to classify the farmland monitoring stations, as shown below: Farmland monitoring stations with H values ​​below the 50th percentile are marked as low heterogeneous points; Farmland monitoring stations with H values ​​between the 50th and 75th percentiles are marked as heterogeneous points; Farmland monitoring stations with H values ​​higher than the 75th percentile are marked as high heterogeneous sites.

4. The genetic optimization spatial layout method for farmland monitoring stations according to claim 1, characterized in that, In step S44, parent selection and crossover mutation are performed in the farmland monitoring station layout scheme. The specific process is as follows: Step S441, for the first Each layout scheme is assigned a selection probability based on its fitness percentage, as shown below: ； in, Indicates the first The probability of selecting a layout scheme; Indicates the first The adaptability of each layout scheme; This represents the sum of fitness for all layout options; the selection probability of each layout option is sequentially mapped onto a cumulative probability axis in the interval [0,1], forming a virtual roulette wheel; Step S442: Generate a uniform random number r in the interval [0,1], and select the corresponding scheme based on the roulette interval in which the value of r falls; Step S443: Based on the crossover operator, randomly select two different parent schemes for crossover. For highly heterogeneous points, align the farmland monitoring stations of the same type in the two schemes according to the nearest neighbor matching principle of farmland monitoring stations by latitude and longitude. Perform simulated binary crossover on the latitude and longitude coordinates of all matched highly heterogeneous farmland monitoring stations, as shown below: ； in, and Indicates the offspring scheme; and Indicates the parent scheme; A random number between 0 and 1; It is the cross-distribution index; The cross-distribution coefficient; Step S444: Based on the mutation operator, perform mutation operation on each child scheme: Deletion mutation is probability Randomly delete a monitoring station for low-heterogeneous cultivated land, i.e. ; New mutations are probabilities : Randomly add farmland monitoring stations in highly heterogeneous areas, with coordinates evenly distributed within highly heterogeneous Thiessen polygons; if the coordinates fall into non-farmland after mutation, regenerate or cancel the operation.

5. A method for genetically optimized spatial layout of cultivated land monitoring stations according to any one of claims 1-4, characterized in that, A genetically optimized spatial layout system for a farmland monitoring station, the system comprising the following modules: The data acquisition and preprocessing module is used to collect the geographic information of the established farmland monitoring stations and the soil index data of each station, and to preprocess the data. The geographic information includes latitude and longitude, and elevation; the soil index data includes pH value, organic matter content, total nitrogen content, available phosphorus content, and available potassium content. The spatial heterogeneity comprehensive index construction module determines the weights of each soil indicator at the monitoring station based on the entropy weight method, and constructs the spatial heterogeneity comprehensive index. The heterogeneity classification and spatial coverage modeling module classifies the monitoring stations into heterogeneity levels based on the comprehensive spatial heterogeneity index of each monitoring station and its corresponding quantile, and quantifies the monitoring range using Thiessen polygons. The genetic algorithm optimization module includes site scheme initialization, scheme encoding, fitness calculation, design of selection, crossover and mutation operators, and iterative selection of spatial optimization layout schemes for farmland monitoring sites.

Images