Self-adaptive precise fertilization amount calculation method and system based on crop growth model and Gaussian distribution

Abstract

The invention provides a self-adaptive precise fertilization amount calculation method based on a crop growth model and Gaussian distribution. The method comprises the following steps: constructing a dynamic fertilization core parameter library and initializing Gaussian distribution parameters; collecting multi-dimensional basic input data of the target area; based on the dynamic fertilization core parameter library and the multi-dimensional basic input data, performing growth sensitive period coupling on a target crop through Gaussian distribution to obtain a fertilization time window; calculating the single fertilization amount and the accumulated fertilization amount of the fertilization time window through the three-factor correction coefficient; and performing layered grid and soil fertility partition coupling calculation according to the single fertilization amount and the accumulated fertilization amount to obtain the total fertilization amount of the target area. According to the method, through Gaussian distribution coupling crop growth sensitive period, three-factor correction, layered grid and soil fertility partition coupling calculation, crop characteristics, soil conditions and meteorological changes are adapted, and the accuracy and actual fitness of regional accurate fertilization amount calculation are improved.

Description

Technical Field

[0001] This invention relates to the field of agricultural environmental simulation and emission inventory accounting technology, and in particular to an adaptive and precise fertilizer application calculation method and system based on crop growth models and Gaussian distribution. Background Technology

[0002] Accurate quantification of fertilizer application is the cornerstone of assessing the environmental impact of agricultural activities. In regional and even national-scale environmental models, such as simulating haze caused by ammonia (NH3) emissions in the atmosphere or estimating the emissions of nitrous oxide (N2O), a potent greenhouse gas from farmland, the reliability of the results highly depends on whether the input fertilizer application data closely reflects reality. However, current fertilizer application data used in macro-environmental studies often exhibit significant uncertainty. This stems primarily from two aspects: firstly, crop planting time is uncertain due to interannual fluctuations in meteorological conditions; secondly, different regions and farmers show significant "local customs" in their choices of fertilization frequency, timing, and fertilizer types (e.g., preference for compound fertilizers and farmyard manure as base fertilizer, and urea as top dressing). Existing models typically use fixed fertilization schedules and uniform fertilizer types for estimation. This simplistic "one-size-fits-all" approach leads to a severe disconnect between the estimated application rates and the complex local realities, resulting in insufficient credibility of subsequent environmental impact assessments.

[0003] To improve this situation, existing techniques have attempted to introduce randomness into models to characterize uncertainty. For example, existing techniques propose using a Gaussian (normal) distribution to describe the random fluctuations in fertilization timing around the theoretical date. However, this method has a fundamental flaw: it forcibly simplifies the fertilization process for all crops into a uniform three-stage process: "before planting, at planting, and before harvest." This extremely simplified model completely fails to capture the specificities of different crops, such as rice and fruit trees, in their growth cycles and nutrient requirements, nor does it reflect the diversity of fertilization strategies resulting from different regional agronomic practices. Therefore, its estimation results still fail to accurately reflect the actual fertilization situation at the regional scale. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive precision fertilization calculation method and system based on crop growth models and Gaussian distribution. By coupling crop growth sensitive periods, three-factor correction, and hierarchical grid with soil fertility zoning calculations through Gaussian distribution, the method adapts to crop characteristics, soil conditions, and meteorological changes, and improves the accuracy and consistency of regional precision fertilization calculations.

[0005] To achieve the above objectives, the present invention provides the following solution: An adaptive and precise fertilization calculation method based on crop growth models and Gaussian distribution includes the following steps: A dynamic fertilization core parameter library was constructed, and Gaussian distribution parameters were initialized based on the dynamic fertilization core parameter library. The dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library, and crop growth sensitive period sub-library. Collect multi-dimensional basic input data for the target area; the multi-dimensional basic input data includes: basic crop data, dynamic soil fertility data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data; Based on the core parameter library of dynamic fertilization and multi-dimensional basic input data, the growth sensitive period of the target crop is coupled by Gaussian distribution to obtain the fertilization time window. The single application amount and cumulative application amount within the fertilization time window are calculated using three-factor correction coefficients, which include: soil fertility correction coefficient, meteorological correction coefficient, and crop growth status correction coefficient. The total application amount for the target area is obtained by coupling calculation with the layered grid and soil fertility zoning based on the single application amount and the cumulative application amount.

[0006] Optionally, a dynamic fertilization core parameter library is constructed, and Gaussian distribution parameters are initialized based on the dynamic fertilization core parameter library, including: By determining the theoretical number of fertilizations, theoretical relative fertilization time, main fertilizer types, and theoretical application rate benchmarks for target crops based on standard agronomic practices, local planting experience, and crop nutrient requirements, a crop sub-bank is obtained. Based on the soil type, soil organic matter content level, soil nitrogen, phosphorus and potassium available nutrient content range and soil pH range of the target area, an adaptation rule between the basic parameters for fertilizer application correction and fertilizer type is constructed to obtain a soil sub-library. Based on the climate zones of the target area, the range of variation coefficients of precipitation, temperature and sunshine hours during the growing season, the correlation between the fertilization time fluctuation adjustment parameters and the meteorological influence weights is constructed, and a meteorological sub-database is obtained. Based on the correlation between the theoretical number of fertilizations and the proportion of chemical fertilizer nutrients during the nutrient requirement sensitive period of the target crop, a sub-library of crop growth sensitive periods was constructed. The initial mean and initial standard deviation of the Gaussian distribution are determined based on the theoretical relative fertilization time and the fertilization sensitivity period.

[0007] Optionally, the initial standard deviation is the product of the fertilizer requirement sensitivity period coefficient and the basic fluctuation days; the fertilizer requirement sensitivity period coefficient is determined based on the fertilizer requirement sensitivity period, and the basic fluctuation days are the default value determined by the survey data on the fertilization time fluctuation of similar crops in the same region.

[0008] Optionally, the adaptation rules include: When soil nitrogen levels are below 0.8 times the standard value, increase the proportion of nitrogen fertilizer among major fertilizer types and decrease the proportion of compound fertilizer. When soil phosphorus levels are below 0.8 times the standard value, increase the proportion of diammonium phosphate and decrease the proportion of potassium chloride in the main fertilizer types. When soil potassium levels are below 0.8 times the standard value, increase the proportion of potassium chloride or potassium sulfate in the main fertilizer types and reduce the proportion of compound fertilizer. When the soil pH is below 5.5, add alkaline fertilizers to the main types of fertilizers and reduce the proportion of physiologically acidic fertilizers. When the soil pH value is higher than 7.5, add acidic fertilizers to the main types of fertilizers and reduce the proportion of physiologically alkaline fertilizers. When the target crop is in a period of high fertilizer demand, increase the proportion of fast-acting fertilizers and decrease the proportion of slow-release fertilizers among the main types of fertilizers.

[0009] Optionally, the basic crop data includes: crop planting date, crop planting area, crop planting density, and crop variety type; dynamic soil fertility data includes: soil type, soil organic matter content, soil available nitrogen content, soil available phosphorus content, soil available potassium content, soil pH value, and soil moisture content; real-time meteorological data includes: cumulative precipitation, average temperature, accumulated temperature, number of extreme temperature days, sunshine hours, and average wind speed; spatial heterogeneity data includes: topography type, altitude, slope, and aspect; and fertilizer recommendation data is a fertilizer application rate recommendation table per unit area.

[0010] Optionally, based on the dynamic fertilization core parameter library and multi-dimensional basic input data, the target crop's growth sensitivity period is coupled using a Gaussian distribution to obtain the fertilization time window, including: Through formula The meteorological sensitivity coefficient was calculated, where This represents the real-time precipitation variation coefficient. As a weighting factor for precipitation, This is the real-time temperature variation coefficient. Temperature weighting, This represents the real-time solar radiation variation coefficient. Weighting based on solar radiation; Through formula The corrected Gaussian standard deviation was calculated, where The latest date sequence, The earliest date, This represents the fertilizer-sensitive period coefficient. Through formula The corrected Gaussian mean was calculated, where This refers to the theoretical relative fertilization time. This is the deviation coefficient for the reproductive period; A Gaussian probability distribution function is constructed based on the meteorological sensitivity coefficient, the meteorological sensitivity coefficient, and the corrected Gaussian mean. The fertilization time window is determined based on the Gaussian probability distribution function.

[0011] Optionally, the single application rate and cumulative application rate within the fertilization time window are calculated using a three-factor correction coefficient, including: Through formula The soil fertility correction coefficient was calculated, where This is the coefficient of variation in mechanical quality. Organic matter weight, This is the nitrogen deviation coefficient. For nitrogen weight, This is the phosphorus deviation coefficient. For phosphorus weight, This is the potassium deviation coefficient. For potassium weight, This is the pH deviation coefficient. pH weighting; Through formula The meteorological correction coefficients were calculated, where This is the precipitation deviation coefficient. Weighting based on precipitation impact, This is the extreme temperature deviation coefficient. Weighting for the impact of extreme temperatures, This is the solar radiation deviation coefficient. Weighting for the impact of sunlight; Through formula The crop growth status correction coefficient was calculated, where This represents the real-time plant height deviation coefficient. Plant height is the weight factor. This represents the real-time biomass deviation coefficient. Biomass weight; The amount of fertilizer applied per application is calculated based on the soil fertility correction coefficient, the meteorological correction coefficient, and the crop growth status correction coefficient; the formula for calculating the amount of fertilizer applied per application is as follows: ,in, For planting area, For the first The benchmark value for the average application rate of chemical fertilizer per mu; The daily fertilizer application amount is calculated based on the amount of fertilizer applied in a single application, and the cumulative fertilizer application amount is calculated based on the daily fertilizer application amount.

[0012] Optionally, based on the single application amount and cumulative application amount, a layered grid and soil fertility zoning coupled calculation is performed to obtain the total application amount for the target area, including: The target area is divided into multiple basic grid units by a preset resolution, and the basic grid units are further divided into multiple fertility sub-regions according to the soil fertility level. Attribute assignment, Gaussian distribution adaptation, and three-factor correction coefficient adaptation are performed on the fertility sub-region; Calculate the cumulative fertilizer application amount for each fertility sub-region, and then summarize the cumulative fertilizer application amount for each region to obtain the total application amount for the target region.

[0013] An adaptive precision fertilization calculation system based on crop growth models and Gaussian distribution, comprising: The parameter library construction module is used to build a dynamic fertilization core parameter library and initialize Gaussian distribution parameters based on the dynamic fertilization core parameter library; the dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library and crop growth sensitive period sub-library; The data acquisition module is used to collect multi-dimensional basic input data of the target area. The multi-dimensional basic input data includes: basic crop data, dynamic soil fertility data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data. The window partitioning module is used to couple the growth-sensitive period of the target crop with Gaussian distribution based on the dynamic fertilization core parameter library and multi-dimensional basic input data to obtain the fertilization time window. The fertilizer application rate correction module is used to calculate the single application rate and cumulative fertilizer application rate within the fertilization time window using three-factor correction coefficients. The three-factor correction coefficients include: soil fertility correction coefficient, weather correction coefficient, and crop growth status correction coefficient. The total fertilizer application calculation module is used to perform layered grid and soil fertility zoning calculations based on single fertilizer application and cumulative fertilizer application to obtain the total application amount for the target area.

[0014] According to specific embodiments provided by the present invention, the following technical effects are disclosed: The adaptive precision fertilization calculation method and system based on crop growth models and Gaussian distribution provided by the present invention includes: constructing a dynamic fertilization core parameter library and initializing Gaussian distribution parameters according to the dynamic fertilization core parameter library; the dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library, and crop growth sensitive period sub-library; collecting multi-dimensional basic input data of the target area; the multi-dimensional basic input data includes: crop basic data, soil fertility dynamic data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data; based on the dynamic fertilization core parameter library and multi-dimensional basic input data, coupling the growth sensitive period of the target crop with Gaussian distribution to obtain the fertilization time window; calculating the single fertilization amount and cumulative fertilization amount of the fertilization time window through three-factor correction coefficients; the three-factor correction coefficients include: soil fertility correction coefficient, meteorological correction coefficient, and crop growth status correction coefficient; and performing hierarchical grid and soil fertility zoning coupling calculation based on the single fertilization amount and cumulative fertilization amount to obtain the total application amount of the target area. This method constructs a dynamic fertilization core parameter library, collects multi-dimensional basic data, and combines Gaussian distribution with crop growth sensitive period and three-factor correction, as well as hierarchical grid and soil fertility zoning coupling calculation. It adapts to crop characteristics, soil conditions and meteorological changes, and improves the accuracy and actual fit of regional precision fertilization calculation. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. 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.

[0016] Figure 1 This is a flowchart of the adaptive precision fertilization calculation method of the present invention; Figure 2 This is a schematic diagram of the adaptive precision fertilization calculation system of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] like Figure 1 As shown, this invention provides an adaptive and precise fertilization calculation method based on a crop growth model and Gaussian distribution, comprising the following steps: Step 100: Construct a dynamic fertilization core parameter library and initialize Gaussian distribution parameters based on the dynamic fertilization core parameter library; the dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library and crop growth sensitive period sub-library; Step 200: Collect multi-dimensional basic input data for the target area; the multi-dimensional basic input data includes: crop basic data, soil fertility dynamic data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data; Step 300: Based on the dynamic fertilization core parameter library and multi-dimensional basic input data, the growth sensitive period of the target crop is coupled through Gaussian distribution to obtain the fertilization time window; Step 400: Calculate the single application amount and cumulative application amount of fertilizer within the fertilization time window using the three-factor correction coefficients; the three-factor correction coefficients include: soil fertility correction coefficient, meteorological correction coefficient, and crop growth status correction coefficient; Step 500: Perform layered grid and soil fertility zoning coupled calculations based on the single application amount and cumulative application amount to obtain the total application amount for the target area.

[0020] Preferably, a dynamic fertilization core parameter library is constructed, and Gaussian distribution parameters are initialized based on the dynamic fertilization core parameter library, including: By determining the theoretical number of fertilizations, theoretical relative fertilization time, main fertilizer types, and theoretical application rate benchmarks for target crops based on standard agronomic practices, local planting experience, and crop nutrient requirements, a crop sub-bank is obtained. Based on the soil type, soil organic matter content level, soil nitrogen, phosphorus and potassium available nutrient content range and soil pH range of the target area, an adaptation rule between the basic parameters for fertilizer application correction and fertilizer type is constructed to obtain a soil sub-library. Based on the climate zones of the target area, the range of variation coefficients of precipitation, temperature and sunshine hours during the growing season, and the range of variation coefficients of sunshine hours during the growing season, a meteorological sub-database is constructed to adjust the fertilization time fluctuation parameters and the meteorological influence weights. Based on the correlation between the theoretical number of fertilizations and the proportion of chemical fertilizer nutrients during the nutrient requirement sensitive period of the target crop, a sub-library of crop growth sensitive periods was constructed. The initial mean and initial standard deviation of the Gaussian distribution are determined based on the theoretical relative fertilization time and the fertilization sensitivity period.

[0021] In the specific implementation process, step 100 first constructs a crop sub-database. By consulting the National Agricultural Industry Standard Agronomic Practice Manual, the nutrient requirement stage characteristics of the target crop are clarified. More than 50 growers are visited to record actual fertilization behavior. Combining the obtained farmer planting survey data and the nutrient requirement characteristics of crop varieties, the theoretical number of fertilizations is determined. The theoretical relative fertilization time is calculated based on the crop planting day. At the same time, the main fertilizer types are determined based on the survey results of the mainstream fertilizer varieties in the region. Among them, the base fertilizer for wheat in the north prefers N-P2O5-K2O compound fertilizer, while in some areas in the south, urea and compound fertilizer are mixed at a ratio of 1:1. The theoretical application rate benchmark value is obtained by referring to the fertilization guidelines issued by the local agricultural technology extension station and combining the average application rate of the same crop in the main producing areas. Next, a soil sub-bank was constructed. Soil samples were first taken from the target area using a grid method to detect the actual distribution ranges of soil type, organic matter content level, available nitrogen, phosphorus, and potassium nutrient content, and pH value. The nutrient content was compared with the corresponding standard values ​​in the "Soil Nutrient Classification Standard," dividing the soil into three ranges: ≥1.0 times the standard value, 0.8-1.0 times the standard value, and <0.8 times the standard value. Similarly, pH values ​​were divided into three ranges: <5.5, 5.5-7.5, and >7.5. Then, based on the principle of soil nutrient supply and demand balance and the fertilizer acid-base regulation mechanism, adaptation rules were constructed. Specifically, when soil nitrogen is below 0.8 times the nitrogen standard value, the proportion of nitrogen fertilizer in the main fertilizer types was increased by 20%-30%, and the proportion of compound fertilizer was reduced by 15%-25%; when soil phosphorus is below 0.8 times the phosphorus standard value, the proportion of nitrogen fertilizer in the main fertilizer types was increased by 20%-30%, and the proportion of compound fertilizer was reduced by 15%-25%. The proportion of diammonium phosphate in the main fertilizer types should be 25%-35%, while the proportion of potassium chloride should be reduced by 20%-30%. When the soil potassium level is below 0.8 times the standard value, the proportion of potassium chloride or potassium sulfate in the main fertilizer types should be increased by 20%-30%, while the proportion of compound fertilizer should be reduced by 15%-25%. When the soil pH is below 5.5, alkaline fertilizers such as quicklime should be added to the main fertilizer types, with an addition amount calculated at 50-80 kg / mu for every 1 unit decrease in pH value, and the proportion of physiologically acidic fertilizers should be reduced by 20%-30%. When the soil pH is above 7.5, acidic fertilizers such as ammonium sulfate should be added to the main fertilizer types, while the proportion of physiologically alkaline fertilizers should be reduced by 20%-30%. When the target crop is in a nutrient-sensitive period, the proportion of fast-acting fertilizers in the main fertilizer types should be increased by 30%-40%, while the proportion of slow-release fertilizers should be reduced by 25%-35%.

[0022] Subsequently, a meteorological sub-database was constructed, collecting meteorological data from the target area over the past 10 years during the growing season. The ratio of the standard deviation of precipitation to the average precipitation was used as the coefficient of variation for precipitation during the growing season; the ratio of the standard deviation of temperature to the average temperature was used as the coefficient of variation for temperature; and the ratio of the standard deviation of sunshine duration to the average sunshine duration was used as the coefficient of variation for sunshine duration. Based on the magnitude of the coefficient of variation, three intervals were defined: low (<0.2), medium (0.2-0.5), and high (>0.5). Pearson correlation analysis was used to determine the correlation between each meteorological factor and the fluctuation of fertilization time. The weights of precipitation, temperature, and sunshine duration were summed to 1. In some embodiments, the weights of precipitation, temperature, and sunshine duration were 0.4, 0.35, and 0.25, respectively. This established the correlation between different coefficient of variation intervals and the adjustment parameter for fluctuation of fertilization time. When the coefficient of variation of precipitation was in the high interval, the fluctuation adjustment parameter was 1.2; in the medium interval, it was 1.0; and in the low interval, it was 0.8.

[0023] A crop growth sensitivity period sub-database was then constructed. Crop physiological experiments were used to identify nutrient-sensitive periods such as the jointing stage of wheat, the tasseling stage of corn, and the fruit enlargement stage of fruit trees. The theoretical number of fertilizations corresponding to each sensitivity period was statistically analyzed. A correlation was established between the theoretical number of fertilizations and the proportion of fertilizer nutrients based on the proportion of nutrient demand during the sensitivity period to the total nutrient demand during the entire growth period, ensuring that the proportion of key nutrients during the sensitivity period is 15%-25% higher than that during non-sensitivity periods. Finally, Gaussian distribution parameters were initialized. The initial mean was directly taken from the corresponding theoretical relative fertilization time in the crop sub-database, and the initial standard deviation was the product of the nutrient-sensitive period coefficient and the basic fluctuation days. The nutrient-sensitive period coefficient was determined based on the nutrient-sensitive period: 1.2 for key sensitivity periods, 0.8 for general sensitivity periods, and 0.5 for non-sensitivity periods. The basic fluctuation days were obtained by collecting fertilization time survey data of similar crops in the same region for more than 5 years and calculating the standard deviation of fertilization time fluctuation as the default value. In some embodiments, the basic fluctuation days were 30 days for grain crops and 45 days for fruit trees.

[0024] Preferably, the basic crop data includes: crop planting date, crop planting area, crop planting density, and crop variety type; dynamic soil fertility data includes: soil type, soil organic matter content, soil available nitrogen content, soil available phosphorus content, soil available potassium content, soil pH value, and soil moisture content; real-time meteorological data includes: cumulative precipitation, average temperature, accumulated temperature, number of extreme temperature days, sunshine hours, and average wind speed; spatial heterogeneity data includes: topography type, altitude, slope, and aspect; and fertilizer recommendation data is a fertilizer application rate recommendation table per unit area.

[0025] Specifically, step 300, based on the crop growth sensitive period sub-database and meteorological sub-database in the dynamic fertilization core parameter library, and the real-time meteorological data and crop basic data in the multi-dimensional basic input data, first calculates the meteorological sensitivity coefficient. The calculation formula is as follows: ; in The real-time precipitation variation coefficient is calculated by dividing the standard deviation of the difference between the current cumulative precipitation during the growing season in the target area and the average precipitation during the same period in the past 10 years by the average precipitation. Precipitation weights are preset in the meteorological sub-database (determined by the analytic hierarchy process, with values ​​ranging from 0.3 to 0.5). The real-time temperature variation coefficient is calculated by dividing the standard deviation of the difference between the current average temperature during the reproductive period and the average temperature of the same period in the past 10 years by the average temperature. The temperature weights are preset in the temperature weight meteorological sub-database (with values ​​ranging from 0.25 to 0.4). The real-time sunshine variation coefficient is calculated by dividing the standard deviation of the difference between the current sunshine duration during the reproductive period and the average sunshine duration during the same period in the past 10 years by the average sunshine duration. The value is the sunshine weight (ranging from 0.15 to 0.3), and the sum of the three weights is 1.

[0026] Next, calculate the corrected Gaussian standard deviation using the following formula: ; in The latest date (the number of days between the latest fertilization date and the planting date for similar crops, determined based on local planting experience surveys). The earliest date (based on the number of days between the earliest fertilization date and the planting date for similar crops determined by the survey). The coefficient for fertilizer demand sensitivity period is 1.3 for critical sensitive period, 0.9 for general sensitive period, and 0.6 for non-sensitive period.

[0027] Next, calculate the corrected Gaussian mean using the following formula: ; in This represents the theoretical relative fertilization time for this fertilization within the crop sub-bank. The growth period progress deviation coefficient is calculated by the ratio of real-time crop height and biomass to the corresponding indicators of the theoretical growth period. If the real-time indicators are higher than the theoretical values, a positive deviation of 0.05-0.1 is taken; if they are lower than the theoretical values, a negative deviation of -0.1 to -0.05 is taken.

[0028] Subsequently, a Gaussian probability distribution function is constructed based on the meteorological sensitivity coefficient, the corrected Gaussian mean, and its expression is: ; in Let be the probability density of fertilization for the i-th fertilization on day t (t being the day sequence relative to the planting day). Finally, the fertilization time window is determined. First, a preliminary time range is defined by combining the nutrient-demand sensitive period corresponding to this fertilization in the crop growth sensitive period sub-database. Then, based on the Gaussian probability distribution function, continuous day sequence intervals with a probability density ≥ 0.05 are selected, while ensuring that this interval covers the core period of the nutrient-demand sensitive period. Ultimately, a fertilization time window that combines probabilistic rationality with the crop's nutrient-demand characteristics is obtained. In some embodiments, the fertilization time window for wheat greening fertilizer is 172-188 days after planting.

[0029] Specifically, step 400 first calculates the soil fertility correction coefficient, using the following formula: ; in It is the organic matter deviation coefficient (i.e., the ratio of the actual organic matter content of the soil in the target area to the corresponding grade standard value in the soil sub-bank minus 1). The organic matter weight (determined by expert scoring, and in some embodiments, it is 0.2). It is the nitrogen deviation coefficient (i.e., the ratio of the actual content of soil available nitrogen to the standard value minus 1). The nitrogen weight (in some embodiments, the value is 0.3). It is the phosphorus deviation coefficient (i.e., the ratio of the actual content of available phosphorus in the soil to the standard value minus 1). The phosphorus weight (in some embodiments, the value is 0.25). It is the potassium deviation coefficient (i.e., the ratio of the actual content of available potassium in the soil to the standard value minus 1). The potassium weight (in some embodiments, the value is 0.2). It is the pH deviation coefficient (i.e., the difference between the actual pH value of the soil and the median value of the suitable pH range divided by half the width of the suitable range). The pH weights are 0.05 in some embodiments, and the sum of all weights is 1.

[0030] Next, the meteorological correction factor is calculated using the following formula: ; in It is the precipitation deviation coefficient (i.e., the ratio of the cumulative precipitation during the current growing season in the target area to the average precipitation of the corresponding climate zone in the meteorological sub-database minus 1). The weighting for the impact of precipitation (in some embodiments, the value is 0.4). It is the extreme temperature deviation coefficient (i.e., the ratio of the number of extreme temperature days in the current reproductive period to the average number of extreme temperature days minus 1). The weighting for the impact of extreme temperatures (in some embodiments, the value is 0.3). It is the sunshine deviation coefficient (i.e., the ratio of the current sunshine hours to the average sunshine hours during the reproductive period minus 1). The weights for the impact of sunlight (in some embodiments, the value is 0.3) are used, and the sum of all weights is 1.

[0031] Next, calculate the crop growth status correction coefficient. The calculation formula is as follows: ; in The real-time plant height deviation coefficient (i.e., the ratio of the current actual plant height of the target crop to the theoretical plant height of the corresponding growth stage in the crop sub-bank minus 1) The plant height is weighted (in some embodiments, the value is 0.5). This is the real-time biomass deviation coefficient (i.e., the ratio of the current actual biomass to the theoretical biomass minus 1). The biomass weight is 0.5 in some embodiments, and the sum of all weights is 1.

[0032] Then, the total amount of fertilizer applied for the i-th fertilization is calculated based on the soil fertility correction coefficient, the meteorological correction coefficient, and the crop growth status correction coefficient. The calculation formula is as follows: ; in, The planting area is from the multi-dimensional basic input data, and 666.67 is the conversion factor between square meters and mu. The first in the crop sub-library The baseline value for the average application rate of fertilizer per mu (unit of land area) is determined. Next, the daily fertilizer application rate is calculated based on the Gaussian probability distribution function obtained in step 300, using the formula... Calculations show that Let be the amount of fertilizer applied on day t for the i-th fertilization. Finally, sum the amounts of fertilizer applied on all days within the fertilization time window to obtain the cumulative amount of fertilizer applied for that fertilization time window.

[0033] Preferably, the total application amount for the target area is obtained by coupling calculation with a layered grid and soil fertility zoning based on the single application amount and cumulative application amount, including: The target area is divided into multiple basic grid units by a preset resolution, and the basic grid units are further divided into multiple fertility sub-regions according to the soil fertility level. Attribute assignment, Gaussian distribution adaptation, and three-factor correction coefficient adaptation are performed on the fertility sub-region; Calculate the cumulative fertilizer application amount for each fertility sub-region, and then summarize the cumulative fertilizer application amount for each region to obtain the total application amount for the target region.

[0034] In the specific implementation process, step 500 determines the preset resolution based on the area size and spatial heterogeneity characteristics of the target area. When the area is less than 1000 square kilometers, a resolution of 500m×500m is used, and when the area is ≥1000 square kilometers, a resolution of 1km×1km is used. GIS technology is used to overlay the administrative division map and topographic map of the target area. The target area is divided into multiple non-overlapping, fully covered basic grid units through grid partitioning tools. Each unit is assigned a unique grid number to ensure that the grid boundary does not cross the key agricultural production zone. Next, based on the dynamic soil fertility data such as soil organic matter content, nitrogen, phosphorus and potassium available nutrient content, and pH value in the multi-dimensional basic input data, the fuzzy C-means clustering (FCM) algorithm is used to perform cluster analysis on each basic grid unit. Referring to the nutrient content range standard in the soil sub-database (i.e., organic matter content ≥2% is high fertility, 1%-2% is medium fertility, and <1% is low fertility), the clustering results are divided into three fertility levels: high, medium, and low. Each level is a fertility sub-region. At the same time, it is ensured that the coefficient of variation of soil fertility indicators within each fertility sub-region is ≤0.2 to ensure the homogeneity of the partition. Subsequently, attribute values ​​were assigned. For each fertility sub-region, missing soil fertility data in the grid cells were supplemented using GIS spatial kriging interpolation. The average values ​​of soil organic matter, available nitrogen, phosphorus, and potassium nutrients, and pH were calculated as soil fertility attribute values ​​for the sub-region. Spatial heterogeneity data were overlaid to obtain the sub-region's topographic type, average altitude, average slope, and dominant slope aspect. The summaries of crop planting area, variety type, and planting density within the sub-region were also compiled. Real-time meteorological data was then integrated to obtain the average values ​​of meteorological attributes such as cumulative precipitation and average temperature for the sub-region, thus ensuring the completeness and accuracy of soil, crop, meteorological, and spatial attributes for each fertility sub-region. For the crop type of each fertility sub-region, the theoretical fertilization frequency and theoretical relative fertilization time from the dynamic fertilization core parameter library were called. Combined with the average meteorological attributes of the sub-region, the meteorological sensitivity coefficient, corrected Gaussian mean, and standard deviation of the sub-region were recalculated using the calculation formula from step 300, and a Gaussian probability distribution function adapted to the actual conditions of the sub-region was constructed.

[0035] Next, a three-factor correction coefficient adaptation is performed. Based on the average soil fertility attribute of the sub-region, the soil fertility correction coefficient is calculated using the formula in step 400. The meteorological correction coefficient is calculated based on the average meteorological attribute of the sub-region. Then, the crop growth status correction coefficient is calculated based on the average real-time crop height and biomass of the sub-region, thus ensuring that the correction coefficients are highly matched with the actual conditions of the sub-region. Finally, the cumulative fertilizer application amount of each fertility sub-region is calculated and summarized. Based on the adapted Gaussian distribution function and the three-factor correction coefficients, the single application amount of each fertilizer in the sub-region and the daily application amount within the application time window are calculated using the single application amount formula in step 400. The summation is used to obtain the sub-region cumulative fertilizer application amount. Finally, the cumulative fertilizer application amounts of all fertility sub-regions are added together to obtain the total application amount of the target area. During the summary, GIS spatial overlay analysis is used to remove duplicate calculations at grid boundaries to ensure the accuracy of the total application amount statistics.

[0036] like Figure 2 As shown, this invention also provides an adaptive precision fertilization calculation system based on a crop growth model and Gaussian distribution, including: The parameter library construction module is used to build a dynamic fertilization core parameter library and initialize Gaussian distribution parameters based on the dynamic fertilization core parameter library; the dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library and crop growth sensitive period sub-library; The data acquisition module is used to collect multi-dimensional basic input data of the target area. The multi-dimensional basic input data includes: basic crop data, dynamic soil fertility data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data. The window partitioning module is used to couple the growth-sensitive period of the target crop with Gaussian distribution based on the dynamic fertilization core parameter library and multi-dimensional basic input data to obtain the fertilization time window. The fertilizer application rate correction module is used to calculate the single application rate and cumulative fertilizer application rate within the fertilization time window using three-factor correction coefficients. The three-factor correction coefficients include: soil fertility correction coefficient, weather correction coefficient, and crop growth status correction coefficient. The total fertilizer application calculation module is used to perform layered grid and soil fertility zoning calculations based on single fertilizer application and cumulative fertilizer application to obtain the total application amount for the target area.

[0037] The beneficial effects of this invention are as follows: 1) By constructing a dynamic fertilization core parameter library that includes crop sub-libraries, soil sub-libraries, meteorological sub-libraries, and crop growth sensitive period sub-libraries, and combining it with multi-dimensional basic input data, it breaks through the simplified "one-size-fits-all" mode of existing technologies. It can dynamically adjust the number of fertilizations, fertilizer types, and nutrient ratios according to the fertilizer requirements of different crops, soil fertility status, meteorological fluctuations, and local planting habits, and adaptably cover a variety of crops such as grain crops, fruit trees, and vegetables, as well as regions with different climates and soil types. 2) A Gaussian distribution coupled with crop growth sensitivity period was introduced. The distribution parameters were corrected by the meteorological sensitivity coefficient and the growth period progress deviation coefficient, which accurately characterized the random fluctuation of fertilization time. At the same time, the influence of various environmental and crop growth factors on fertilizer application was quantified by the three-factor correction coefficients of soil fertility, meteorology and crop growth status. This solved the estimation bias problem caused by fixed time and uniform parameters, making the calculation results closer to the actual application situation. This improved the accuracy of fertilizer application calculation while reducing uncertainty. 3) The calculation is coupled with layered grid and soil fertility zoning. The basic grid is divided according to the preset resolution and the fertility sub-regions are subdivided according to the soil fertility level. The attribute values ​​are assigned and the parameters are adapted in combination with spatial heterogeneity data. It can output the total application amount of the region and generate spatialized fertilization data at the grid scale, which meets the dual needs of macro statistics and refined environmental simulation, and accurately captures spatial heterogeneity. 4) The parameters of the crop sub-bank are determined through standard agronomic practices and local planting experience surveys. The soil sub-bank is pre-set with fertilizer matching rules based on soil nutrient content and pH value. The meteorological sub-bank is combined with regional climate zones to construct fluctuation adjustment correlations. It fully considers the impact of "local habits" on the number of fertilizations and fertilizer types, avoids the problem of forcibly simplifying the fertilization process, and truly restores the actual fertilization behavior patterns in different regions, making the method more in line with actual agricultural production.

[0038] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0039] Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. Furthermore, those skilled in the art will recognize that, based on the ideas of this invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. An adaptive and precise fertilization calculation method based on crop growth models and Gaussian distribution, characterized in that, Includes the following steps: A dynamic fertilization core parameter library is constructed, and Gaussian distribution parameters are initialized based on the dynamic fertilization core parameter library; the dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library and crop growth sensitive period sub-library; Collect multi-dimensional basic input data of the target area; the multi-dimensional basic input data includes: basic crop data, dynamic soil fertility data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data; Based on the dynamic fertilization core parameter library and the multi-dimensional basic input data, the fertilization time window is obtained by coupling the growth sensitive period of the target crop with Gaussian distribution. The single application amount and cumulative application amount within the fertilization time window are calculated using three-factor correction coefficients, which include: soil fertility correction coefficient, meteorological correction coefficient, and crop growth status correction coefficient. The total application amount for the target area is obtained by performing a layered grid coupled with soil fertility zoning calculation based on the single application amount and the cumulative application amount.

2. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 1, characterized in that, Construct a dynamic fertilization core parameter library, and initialize Gaussian distribution parameters based on the dynamic fertilization core parameter library, including: The theoretical number of fertilizations, theoretical relative fertilization time, main fertilizer types, and theoretical application rate benchmarks for the target crop are determined by standard agronomic practices, local planting experience, and crop fertilizer requirements, thus obtaining the crop sub-database. Based on the soil type, soil organic matter content level, soil nitrogen, phosphorus and potassium available nutrient content range and soil pH range of the target area, an adaptation rule between the basic parameters for fertilizer application correction and fertilizer type is constructed to obtain the soil sub-library. Based on the climate zones of the target area, the range of variation coefficients of precipitation, temperature, and sunshine hours during the growing season, the correlation between the fertilization time fluctuation adjustment parameter and the meteorological influence weight is constructed, and the meteorological sub-database is obtained. Based on the nutrient requirement sensitivity period of the target crop, a correlation relationship between the theoretical number of fertilizations and the proportion of chemical fertilizer nutrients is constructed to obtain the crop growth sensitivity period sub-library; The initial mean and initial standard deviation of the Gaussian distribution are determined based on the theoretical relative fertilization time and the fertilizer demand sensitive period.

3. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 2, characterized in that, The initial standard deviation is the product of the fertilizer demand sensitivity period coefficient and the basic fluctuation days; the fertilizer demand sensitivity period coefficient is determined based on the fertilizer demand sensitivity period, and the basic fluctuation days is a default value determined by the survey data on the fertilization time fluctuation of similar crops in the same region.

4. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 3, characterized in that, The adaptation rules include: When soil nitrogen levels are below 0.8 times the standard nitrogen value, increase the proportion of nitrogen fertilizer among the main types of fertilizers and decrease the proportion of compound fertilizer. When soil phosphorus levels are below 0.8 times the standard phosphorus value, increase the proportion of diammonium phosphate and decrease the proportion of potassium chloride in the main fertilizer types. When soil potassium levels are below 0.8 times the standard potassium value, increase the proportion of potassium chloride or potassium sulfate in the main fertilizer types and decrease the proportion of compound fertilizer. When the soil pH is below 5.5, add alkaline fertilizers to the main types of fertilizers and reduce the proportion of physiologically acidic fertilizers. When the soil pH value is higher than 7.5, add acidic fertilizers to the main types of fertilizers and reduce the proportion of physiologically alkaline fertilizers. When the target crop is in the fertilizer-sensitive period, the proportion of fast-acting fertilizers among the main fertilizer types should be increased and the proportion of slow-release fertilizers should be decreased.

5. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 1, characterized in that, The basic crop data includes: crop planting date, crop planting area, crop planting density, and crop variety type; the dynamic soil fertility data includes: soil type, soil organic matter content, soil available nitrogen content, soil available phosphorus content, soil available potassium content, soil pH value, and soil moisture content; the real-time meteorological data includes: cumulative precipitation, average temperature, accumulated temperature, number of days with extreme temperatures, sunshine hours, and average wind speed; the spatial heterogeneity data includes: topography type, altitude, slope, and aspect; and the fertilizer recommendation data is a fertilizer application rate recommendation table per unit area.

6. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 1, characterized in that, Based on the aforementioned dynamic fertilization core parameter library and the aforementioned multi-dimensional basic input data, a fertilization time window is obtained by coupling the growth sensitivity period of the target crop through a Gaussian distribution, including: Through formula The meteorological sensitivity coefficient was calculated, where This represents the coefficient of variation of real-time precipitation. As a weighting factor for precipitation, This is the real-time temperature variation coefficient. Temperature weighting, This represents the real-time solar radiation variation coefficient. Weighting based on solar radiation; Through formula The corrected Gaussian standard deviation was calculated, where The latest date sequence, The earliest date, This represents the fertilizer-sensitive period coefficient. Through formula The corrected Gaussian mean was calculated, where This refers to the theoretical relative fertilization time. This is the deviation coefficient for the reproductive period. A Gaussian probability distribution function is constructed based on the meteorological sensitivity coefficient, the meteorological sensitivity coefficient, and the corrected Gaussian mean. The fertilization time window is determined based on the Gaussian probability distribution function.

7. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 1, characterized in that, The single application amount and cumulative application amount within the fertilization time window are calculated using a three-factor correction coefficient, including: Through formula The soil fertility correction coefficient was calculated, where This is the coefficient of variation of the mechanical quality. Organic matter weight, This is the nitrogen deviation coefficient. For nitrogen weight, This is the phosphorus deviation coefficient. For phosphorus weight, This is the potassium deviation coefficient. For potassium weight, This is the pH deviation coefficient. pH weighting; Through formula The meteorological correction coefficient is calculated, where This is the precipitation deviation coefficient. Weighting based on precipitation impact, This is the extreme temperature deviation coefficient. Weighting for the impact of extreme temperatures, This is the solar radiation deviation coefficient. Weighting for the impact of sunlight; Through formula The crop growth state correction coefficient was calculated, where This represents the real-time plant height deviation coefficient. Plant height is the weight factor. This represents the real-time biomass deviation coefficient. Biomass weight; The single application rate of fertilizer is calculated based on the soil fertility correction coefficient, the meteorological correction coefficient, and the crop growth status correction coefficient; the formula for calculating the single application rate of fertilizer is: ,in, For planting area, For the first The benchmark value for the average application rate of chemical fertilizer per mu; The daily fertilizer application amount is calculated based on the single application amount, and the cumulative fertilizer application amount is calculated based on the daily fertilizer application amount.

8. The adaptive precision fertilization calculation method based on crop growth model and Gaussian distribution according to claim 1, characterized in that, Based on the single application amount and the cumulative application amount, a layered grid and soil fertility zoning coupled calculation is performed to obtain the total application amount for the target area, including: The target area is divided into multiple basic grid units by a preset resolution, and the basic grid units are further divided into multiple fertility sub-regions according to the soil fertility level. The fertility sub-region is assigned attribute values, adapted to Gaussian distribution, and fitted with the three-factor correction coefficients. The cumulative fertilizer application amount of each fertility sub-region is calculated, and the cumulative fertilizer application amount is summarized by region to obtain the total application amount of the target region.

9. An adaptive precision fertilization calculation system based on crop growth models and Gaussian distribution, characterized in that, include: The parameter library construction module is used to construct a dynamic fertilization core parameter library and initialize Gaussian distribution parameters based on the dynamic fertilization core parameter library. The dynamic fertilization core parameter library includes: crop sub-library, soil sub-library, meteorological sub-library, and crop growth sensitive period sub-library; The data acquisition module is used to collect multi-dimensional basic input data of the target area; the multi-dimensional basic input data includes: basic crop data, dynamic soil fertility data, real-time meteorological data, spatial heterogeneity data, and fertilization recommendation data; The window partitioning module is used to obtain the fertilization time window by coupling the growth sensitive period of the target crop with the dynamic fertilization core parameter library and the multi-dimensional basic input data through Gaussian distribution. The fertilizer application rate correction module is used to calculate the single application rate and cumulative application rate of the fertilization time window through three-factor correction coefficients; the three-factor correction coefficients include: soil fertility correction coefficient, meteorological correction coefficient and crop growth status correction coefficient; The total fertilizer application calculation module is used to perform layered grid and soil fertility zoning coupled calculations based on the single fertilizer application amount and the cumulative fertilizer application amount to obtain the total application amount for the target area.

Images