Intelligent water and fertilizer analysis method and system based on regionalized crop yield improvement in farmland

Abstract

The present invention relates to the technical field of intelligent water and fertilizer analysis, and in particular to an intelligent water and fertilizer analysis method and system based on regionalized crop yield improvement in farmland. The method comprises the following steps: acquiring farmland regional structure data and extracting irrigation water flow direction, deploying sensors to monitor farmland status, and generating irrigation basin farmland status data; collecting crop morphological images through electronic monitoring equipment and water flow direction data, evaluating the degree of leaf yellowing and curling, and extracting soil volume water saturation. According to soil moisture data, changes in nutrient element content are extracted, and the nutrient leaching loss gradient is quantified, further simulating soil nitrogen cycle anomalies; finally, based on the nitrogen cycle anomaly data, intelligent water and fertilizer dosage controlled-release analysis is performed on the degree of leaf yellowing and curling, and the results are sent to a terminal to improve crop yield. The present invention makes the intelligent water and fertilizer analysis technology more perfect by optimizing the processing of the intelligent water and fertilizer analysis technology.

Description

Technical Field

[0001] The present invention relates to the technical field of water and fertilizer intelligent analysis, and in particular to a water and fertilizer intelligent analysis method and system based on regionalized farmland crop yield improvement. Background Art

[0002] In recent years, with the rapid development of technologies such as the Internet of Things, big data, cloud computing, and artificial intelligence, intelligent agricultural management has made significant progress. Sensor-based farmland monitoring, crop growth data collection, and precise water and fertilizer regulation are gradually being applied to agricultural production. By accurately adjusting water and fertilizer usage through real-time monitoring of soil, climate, and crop growth data, this approach not only increases crop yields but also effectively reduces resource waste, lowers environmental pollution, and achieves sustainable agricultural production. In particular, in the area of ​​intelligent water and fertilizer analysis, previous water and fertilizer management methods have often relied on standardized fertilization standards, ignoring the differences between regions, crops, and soils. This approach often leads to wasteful use of water and fertilizer resources and even negatively impacts the environment. Intelligent water and fertilizer analysis methods for crop yield improvement based on regionalized farmland can differentiate water and fertilizer management based on factors such as regional characteristics, crop needs, and soil conditions. By comprehensively analyzing data such as irrigation water flow, soil moisture, nutrient content, and crop growth status, combined with intelligent control technology, precise regulation of water and fertilizer usage can be achieved. This not only increases crop yields but also reduces water and fertilizer input costs, minimizes unnecessary resource consumption, and avoids environmental pollution caused by overfertilization, promoting sustainable and green agriculture. However, traditional intelligent water and fertilizer analysis methods based on regionalized crop yield improvement in farmland cannot accurately analyze nutrient loss in the soil, resulting in large errors in water and fertilizer management. Summary of the Invention

[0003] Based on this, it is necessary to provide an intelligent water and fertilizer analysis method and system based on regionalized crop yield improvement in farmland to solve at least one of the above technical problems.

[0004] To achieve the above objectives, a water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland is provided, the method comprising the following steps:

[0005] Step S1: Acquire farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and perform real-time monitoring of the farmland status in the irrigation basin, thereby generating irrigation basin farmland status monitoring data;

[0006] Step S2: Using electronic monitoring equipment and based on farmland irrigation water flow data, crop morphological images are collected to obtain crop morphological images in the water flow area; the degree of leaf yellowing and curling in the crop morphological images in the water flow area is assessed to obtain leaf yellowing and curling degree data; and based on the leaf yellowing and curling degree data, the soil volume water saturation of crops with yellowing and curling leaves in the irrigation basin is extracted from the farmland status monitoring data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0007] Step S3: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantifying the nutrient leaching loss gradient of the soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulating and quantifying nitrogen cycle anomalies based on the nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data;

[0008] Step S4: Based on the soil nitrogen cycle abnormality data, the leaf yellowing and curling degree data is analyzed for soil water and fertilizer intelligent controlled release, and the soil water and fertilizer intelligent controlled release data is obtained. The soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield.

[0009] Preferably, step S2 includes the following steps:

[0010] Step S21: using electronic monitoring equipment and based on the farmland irrigation water flow data, crop morphology images are collected to obtain crop morphology images in the water flow area;

[0011] Step S22: performing growth inhibition distribution recognition on the crop morphological image in the water flow area to obtain crop growth inhibition distribution data;

[0012] Step S23: evaluating the degree of leaf yellowing and curling of the crop morphological images in the water flow area based on the crop growth inhibition distribution data to obtain leaf yellowing and curling degree data;

[0013] Step S24: extracting the soil volume water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0014] Preferably, step S3 includes the following steps:

[0015] Step S31: Calculating the soil particle dispersion density based on the soil volume water saturation to obtain soil particle dispersion density data;

[0016] Step S32: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data;

[0017] Step S33: quantifying the nutrient leaching loss gradient of the soil nutrient element content change data based on the soil particle dispersion density data and the farmland irrigation water flow direction data to obtain nutrient leaching loss gradient data;

[0018] Step S34: quantifying the nitrogen cycle anomaly simulation based on the nutrient leaching loss gradient data and the soil volumetric water saturation to obtain soil nitrogen cycle anomaly data.

[0019] Preferably, step S33 includes the following steps:

[0020] Step S331: performing soil particle size interval evaluation on the soil particle dispersion density data to obtain the soil particle size interval;

[0021] Step S332: performing hydrodynamic soil porosity fluctuation calculation based on the soil particle size range and the farmland irrigation water flow direction data to obtain hydrodynamic soil porosity fluctuation data;

[0022] Step S333: performing soil profile nutrient loss numerical iteration on the soil nutrient element content change data based on the hydrodynamic soil porosity fluctuation data to obtain soil profile nutrient loss iteration data;

[0023] Step S334: evaluating the expansion of the nutrient sinking range on the iterative data of nutrient loss in the soil profile to obtain the expansion of the nutrient sinking range;

[0024] Step S335: quantify the nutrient leaching loss gradient based on the soil profile nutrient loss iteration data and the nutrient sinking range expansion amount to obtain nutrient leaching loss gradient data.

[0025] Preferably, step S333 includes the following steps:

[0026] The hydrodynamic nonlinear pore expansion relationship is analyzed for the hydrodynamic soil porosity fluctuation data to obtain the hydrodynamic nonlinear pore expansion relationship;

[0027] The nonlinear pore expansion relationship of hydrodynamics is used to analyze the shear stress of water flow in soil profiles and obtain the pore water flow stress data of soil profiles.

[0028] The velocity difference of different directions of the pore water flow stress data of the soil profile is deduced to obtain the velocity difference of different directions of the pore water flow in the profile.

[0029] Based on the hydrodynamic nonlinear pore expansion relationship and the velocity difference of the cross-sectional pores at different seepage directions, the seepage convection pressure data of the cross-sectional pores were obtained.

[0030] According to the Richards equation, the velocity difference of different cross-sectional pore infiltration directions and the cross-sectional pore infiltration convection pressure data, the soil nutrient element content change data were numerically iterated to obtain the soil profile nutrient loss iterative data.

[0031] Preferably, step S34 includes the following steps:

[0032] Step S341: performing soil oxygen content decrease gradient analysis based on soil volume water saturation to obtain soil oxygen content decrease gradient data;

[0033] Step S342: performing an anaerobic microbial activity enhancement function relationship analysis based on the nutrient leaching loss gradient data and the soil oxygen content decrease gradient data to obtain an anaerobic microbial activity enhancement relationship function;

[0034] Step S343: performing denitrification rate enhancement coupling according to the anaerobic microbial activity enhancement relationship function to obtain denitrification rate enhancement coupling data;

[0035] Step S344: performing asymptotic fitting estimation of nitrogen release on the denitrification rate enhancement coupling data to obtain nitrogen release fitting estimation data;

[0036] Step S345: Perform nitrogen cycle anomaly simulation and quantification based on the nitrogen release fitting estimation data to obtain soil nitrogen cycle anomaly data.

[0037] Preferably, step S344 includes the following steps:

[0038] Based on the denitrification rate enhanced coupling data, nitrogen release is mapped proportionally to obtain rate-matched nitrogen release mapping data;

[0039] Based on the denitrification rate enhancement coupling data, the rate-matched nitrogen release mapping data were analyzed for multiple extreme values ​​of ammonia release, and the rate-matched ammonia release multiple extreme values ​​were obtained.

[0040] Perform time series delay calculation on multiple extreme values ​​of rate-matched ammonia release to obtain multiple extreme value time series delay data;

[0041] Based on the multiple extreme value time series delay data and the rate matching of multiple extreme values ​​of ammonia release, the mean difference of cyclic repeatability release is calculated to obtain the extreme value mean difference of ammonia release cycle;

[0042] The nitrogen release asymptotic fitting estimation was performed on the extreme mean difference of the ammonia release cycle according to the maximum likelihood estimation method to obtain the nitrogen release fitting estimation data.

[0043] Preferably, step S4 includes the following steps:

[0044] Step S41: performing convolution calculation on the soil nitrogen cycle abnormal data to obtain soil nitrogen cycle abnormal convolution data;

[0045] Step S42: performing soil water saturation control on the soil volume water saturation according to the soil nitrogen cycle abnormal convolution data and the nutrient leaching loss gradient data to obtain soil water saturation control data;

[0046] Step S43: Based on the soil moisture saturation control data and the soil nitrogen cycle abnormality convolution data, the leaf yellowing and curling degree data is subjected to soil water and fertilizer intelligent controlled release analysis to obtain soil water and fertilizer intelligent controlled release data, and the soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield.

[0047] Preferably, step S43 includes the following steps:

[0048] Step S431: performing a nutrient deficiency long-term cumulative effect analysis on the soil nitrogen cycle abnormality convolution data to obtain nutrient deficiency long-term cumulative effect data;

[0049] Step S432: performing a nutrient absorption capacity degradation assessment on the leaf yellowing and curling degree data to obtain crop nutrient absorption capacity degradation data;

[0050] Step S433: analyzing the soil aeration enhancement ratio based on the soil water saturation control data to obtain the soil aeration enhancement ratio;

[0051] Step S434: performing a phased multi-objective optimization of the water-fertilizer ratio based on the long-term cumulative effect data of nutrient deficiency and the soil aeration enhancement ratio on the crop nutrient absorption capacity degradation data to obtain the multi-objective optimization data of the water-fertilizer ratio;

[0052] Step S435: Perform intelligent controlled-release analysis of soil water and fertilizer usage based on the multi-objective optimization data of water and fertilizer ratio, obtain soil water and fertilizer usage intelligent controlled-release data, and send the soil water and fertilizer usage intelligent controlled-release data to the terminal to perform intelligent water and fertilizer analysis for improving crop yield.

[0053] Preferably, the present invention further provides a water and fertilizer intelligent analysis system based on regionalized crop yield improvement in farmland, which is used to execute the water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland as described above. The water and fertilizer intelligent analysis system based on regionalized crop yield improvement in farmland comprises:

[0054] The irrigation basin farmland status monitoring module is used to obtain farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and conduct real-time monitoring of the irrigation basin farmland status, thereby generating irrigation basin farmland status monitoring data;

[0055] The soil volumetric water saturation extraction module is used to collect crop morphological images based on farmland irrigation water flow data using electronic monitoring equipment to obtain crop morphological images in the water flow area; evaluate the degree of leaf yellowing and curling in the crop morphological images in the water flow area to obtain leaf yellowing and curling degree data; and extract the soil volumetric water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volumetric water saturation of crops with yellowing and curling leaves.

[0056] The soil nitrogen cycle anomaly analysis module is used to extract nutrient element content changes from irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantify the nutrient leaching loss gradient of soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulate and quantify nitrogen cycle anomalies based on nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data;

[0057] The intelligent analysis and control module for water and fertilizer usage is used to perform intelligent controlled release analysis of soil water and fertilizer usage based on the data on leaf yellowing and curling abnormalities in the soil nitrogen cycle, obtain intelligent controlled release data of soil water and fertilizer usage, and send the intelligent controlled release data of soil water and fertilizer usage to the terminal to perform intelligent water and fertilizer analysis to improve crop yield.

[0058] The beneficial effect of the present invention is that by acquiring farmland regional structural data and extracting irrigation water flow direction, the operation of the irrigation system and the distribution of water flow within the farmland can be accurately understood. This step facilitates the precise deployment of multi-directional sensors, ensuring coverage of different areas of the farmland for comprehensive monitoring. The acquisition of real-time monitoring data provides a foundation for subsequent precision irrigation and farmland management, ensuring efficient collection of irrigation basin status data and providing a reliable basis for further analysis. By collecting crop morphological images and assessing the degree of leaf yellowing and curling through electronic monitoring equipment, crop health can be promptly detected, particularly early signs of uneven water and fertilizer distribution or pests and diseases. This step provides a visual assessment of crop growth status by analyzing crop morphological images of the water flow area. Leaf yellowing and curling data can help identify potential water or nutrient deficiencies, ensuring that farmland managers can take timely intervention measures to prevent crop growth from being affected, thereby optimizing agricultural production. Extracting soil volumetric water saturation, combined with analysis of changes in nutrient content, can accurately assess soil moisture and nutrient status. This process quantifies changes in nutrients in the soil, helping to reveal the soil's nutrient supply capacity and water status. In particular, quantifying nutrient leaching loss gradients can help understand the severity of nutrient loss. Combined with simulation analysis of nitrogen cycle anomalies, this allows for precise identification of abnormalities in soil nitrogen supply. This analysis provides a scientific basis for developing appropriate fertilization and irrigation plans, avoiding resource waste and environmental pollution. Intelligent controlled-release analysis of water and fertilizer application based on soil nitrogen cycle anomaly data enables precise water and fertilizer management, avoiding over-fertilization and over-irrigation. This process scientifically regulates soil water and fertilizer application, enabling intelligent release of water and fertilizer based on actual crop needs. This significantly improves water and fertilizer utilization efficiency, reduces resource waste, and increases crop yields. By transmitting intelligent analysis data to terminals, agricultural managers receive timely decision support, further optimizing farmland management and production, and promoting sustainable agricultural production. Therefore, the intelligent water and fertilizer analysis method based on regionalized crop yield improvement of farmland provided by the present invention is an optimization processing of the traditional intelligent water and fertilizer analysis method based on regionalized crop yield improvement of farmland, which solves the problem that the traditional intelligent water and fertilizer analysis method based on regionalized crop yield improvement of farmland cannot accurately analyze the nutrient loss in the soil, thereby causing large errors in the management and use of water and fertilizer, improves the accuracy of the analysis of nutrient loss in the soil, and reduces the errors in the management and use of water and fertilizer. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic diagram of the steps of an intelligent water and fertilizer analysis method based on regionalized crop yield improvement in farmland;

[0060] Figure 2 for Figure 1Detailed implementation steps of step S2 in FIG.

[0061] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG. DETAILED DESCRIPTION

[0062] See also Figures 1 to 3 , a water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland, the method comprising the following steps:

[0063] Step S1: Acquire farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and perform real-time monitoring of the farmland status in the irrigation basin, thereby generating irrigation basin farmland status monitoring data;

[0064] Step S2: Using electronic monitoring equipment and based on farmland irrigation water flow data, crop morphological images are collected to obtain crop morphological images in the water flow area; the degree of leaf yellowing and curling in the crop morphological images in the water flow area is assessed to obtain leaf yellowing and curling degree data; and based on the leaf yellowing and curling degree data, the soil volume water saturation of crops with yellowing and curling leaves in the irrigation basin is extracted from the farmland status monitoring data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0065] Step S3: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantifying the nutrient leaching loss gradient of the soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulating and quantifying nitrogen cycle anomalies based on the nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data;

[0066] Step S4: Based on the soil nitrogen cycle abnormality data, the leaf yellowing and curling degree data is analyzed for soil water and fertilizer intelligent controlled release, and the soil water and fertilizer intelligent controlled release data is obtained. The soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield.

[0067] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for intelligent water and fertilizer analysis based on regionalized crop yield improvement in farmland according to the present invention. In this example, the method for intelligent water and fertilizer analysis based on regionalized crop yield improvement in farmland includes the following steps:

[0068] Step S1: Acquire farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and perform real-time monitoring of the farmland status in the irrigation basin, thereby generating irrigation basin farmland status monitoring data;

[0069] In this embodiment of the present invention, farmland regional structure data is acquired based on high-precision remote sensing mapping technology, employing multispectral remote sensing imagery, LiDAR point cloud data, and Geographic Information System (GIS) technology for data collection and processing. First, remote sensing images of the target farmland are acquired using high-resolution multispectral imaging satellites. Indices such as the Normalized Difference Vegetation Index (NDVI) and the Ratio Vegetation Index (RVI) are used to enhance the distribution of farmland vegetation. Second, a LiDAR scanning system is used on an unmanned aerial vehicle (UAV) to calculate the farmland surface elevation model using laser ranging and reflection intensity. This model captures the internal topography of the farmland, including ridges, drainage channels, and slope distribution. The multispectral remote sensing imagery and LiDAR point cloud data are georeferenced and combined with GIS spatial analysis methods to generate farmland regional structure data. To extract irrigation water flow direction from the farmland regional structure data, a digital elevation model (DEM) is used to calculate the water flow direction matrix. This is then combined with a watershed segmentation algorithm (such as the D8 algorithm) to determine the irrigation path, resulting in farmland irrigation water flow direction data. Multi-directional sensors are deployed based on the farmland irrigation water flow data. Soil moisture sensors, conductivity sensors, and temperature and humidity sensors are deployed using wireless sensor networks (WSNs). LoRa communication technology is used to establish a remote data acquisition system to achieve real-time monitoring of the status of farmland in the irrigation basin, and ultimately generate irrigation basin farmland status monitoring data.

[0070] Step S2: Using electronic monitoring equipment and based on farmland irrigation water flow data, crop morphological images are collected to obtain crop morphological images in the water flow area; the degree of leaf yellowing and curling in the crop morphological images in the water flow area is assessed to obtain leaf yellowing and curling degree data; and based on the leaf yellowing and curling degree data, the soil volume water saturation of crops with yellowing and curling leaves in the irrigation basin is extracted from the farmland status monitoring data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0071] In this embodiment of the present invention, crop morphology images are acquired based on farmland irrigation water flow data using a hyperspectral imager and an RGB camera. It is important to note that crops are planted simultaneously and of the same type, and crops of the same type are planted in the same area. First, a hyperspectral imager is set up in key irrigation water flow areas, and a fixed-angle automatic camera system is used to capture images in different time periods. Second, a multi-scale segmentation algorithm is used to pre-process the collected crop morphology images of the water flow areas, removing shadows and soil background interference while retaining the crop canopy area. When evaluating the degree of leaf yellowing and curling in the extracted crop canopy area, color histogram statistical analysis and edge detection methods are used to calculate the leaf yellowing rate and curling degree. Leaf texture information is extracted using the local binary pattern (LBP) feature description method, and leaf health status is determined using a support vector machine (SVM) classifier to obtain leaf yellowing and curling degree data. Based on the yellowing and curling degree data of leaves, the soil volumetric water saturation of crops with yellowing and curling leaves was extracted from the farmland status monitoring data in the irrigation basin. The soil moisture content of soil layers at different depths was calculated using the geostatistical interpolation method, and the soil volumetric water saturation of the root zone of crops with yellowing and curling leaves was estimated in combination with the time series analysis method. Finally, the soil volumetric water saturation data of crops with yellowing and curling leaves were generated.

[0072] Step S3: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantifying the nutrient leaching loss gradient of the soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulating and quantifying nitrogen cycle anomalies based on the nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data;

[0073] In this embodiment of the present invention, nutrient element content changes are extracted from farmland status monitoring data in irrigation basins based on the soil volumetric water saturation of crops with yellowing and curling leaves. Near-infrared spectroscopy (NIR) analysis and ion chromatography (IC) are used to detect changes in the content of major nutrients such as nitrogen (N), phosphorus (P), and potassium (K) in the soil solution. Partial least squares regression (PLSR) is used to reduce the dimensionality of the NIR spectral data. The concentrations of each nutrient in the soil are calculated using a standard curve, and kriging interpolation is used to construct spatial distribution maps of soil nutrient element content changes in different regions. To quantify the nutrient leaching loss gradient of soil nutrient element content change data, a one-dimensional vertical water flow-solute transport model (such as the HYDRUS-1D model) is used to simulate the dynamic process of soil nutrient migration with water under rainfall and irrigation conditions. Leaching loss gradients of nitrogen, phosphorus, and potassium are calculated in soil layers at different depths to obtain nutrient leaching loss gradient data. Based on the nutrient leaching loss gradient data, nitrogen cycle anomaly simulation and quantification were carried out. The migration dynamics of nitrogen in the soil-plant system were determined using stable isotope analysis (15N tracer technology). Combined with the improved nitrogen mass balance equation, nitrogen cycle anomaly parameters were calculated to obtain soil nitrogen cycle anomaly data.

[0074] Step S4: Based on the soil nitrogen cycle abnormality data, the leaf yellowing and curling degree data is analyzed for soil water and fertilizer intelligent controlled release, and the soil water and fertilizer intelligent controlled release data is obtained. The soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield.

[0075] In an embodiment of the present invention, based on the abnormal data of soil nitrogen cycle, the data on the degree of leaf yellowing and curling are used to perform intelligent controlled release analysis of soil water and fertilizer usage, and a spatiotemporal variability analysis method is used to construct a dynamic model of soil water and fertilizer supply and demand, and analyze areas where soil water and fertilizer supply and demand do not match. During the intelligent controlled release analysis process, a fuzzy control algorithm is used to set the critical threshold of soil water and fertilizer, and the optimal amount of nitrogen fertilizer, phosphorus fertilizer, potassium fertilizer and water in different areas is calculated in combination with the abnormal data of soil nitrogen cycle and the data on the degree of leaf yellowing and curling. A drip irrigation system is combined with an intelligent fertilization system to achieve precise water and fertilizer regulation, and variable-speed drip irrigation technology is used to adjust the irrigation flow according to the real-time soil moisture, and a high-throughput liquid fertilization system is used to accurately place fertilizers. The intelligent controlled release data of soil water and fertilizer usage is sent to a remote terminal control system, and 5G Internet of Things technology is used to realize data transmission, and the field water and fertilizer integration equipment is automatically controlled to perform intelligent water and fertilizer analysis to improve crop yield.

[0076] Step S2 includes the following steps:

[0077] Step S21: using electronic monitoring equipment and based on the farmland irrigation water flow data, crop morphology images are collected to obtain crop morphology images in the water flow area;

[0078] Step S22: performing growth inhibition distribution recognition on the crop morphological image in the water flow area to obtain crop growth inhibition distribution data;

[0079] Step S23: evaluating the degree of leaf yellowing and curling of the crop morphological images in the water flow area based on the crop growth inhibition distribution data to obtain leaf yellowing and curling degree data;

[0080] Step S24: extracting the soil volume water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0081] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0082] Step S21: using electronic monitoring equipment and based on the farmland irrigation water flow data, crop morphology images are collected to obtain crop morphology images in the water flow area;

[0083] In this embodiment of the present invention, crop morphology imagery is acquired based on farmland irrigation water flow data. High-resolution optical imaging equipment, multispectral imagers, and unmanned aerial vehicle (UAV) remote sensing systems are used to collaboratively collect data. First, fixed electronic monitoring equipment, including visible light cameras and near-infrared imagers, is deployed in key irrigation water flow areas. A multi-axis pan-tilt system automatically adjusts the camera angle to ensure that the shooting angle aligns with the crop growth direction. Second, a UAV equipped with a hyperspectral imaging system is flown at low altitude along a specific trajectory, and orthophoto stitching technology is used to generate a complete crop morphology image of the water flow area. During the acquisition process, adaptive exposure control technology is used to adjust imaging parameters to eliminate the impact of varying sunlight angles on image quality. Radiometric correction is performed on the acquired multispectral and RGB images, and linear normalization is used to unify the pixel grayscale range. Background segmentation algorithms (such as the GrabCut algorithm) are used to remove non-crop areas, and morphological filtering methods are used to remove small noise points to ensure the accuracy of crop morphology images in the water flow area.

[0084] Step S22: performing growth inhibition distribution recognition on the crop morphological image in the water flow area to obtain crop growth inhibition distribution data;

[0085] In an embodiment of the present invention, growth inhibition distribution identification is based on crop morphological images of water flow areas, using multiple vegetation indices such as the Normalized Difference Vegetation Index (NDVI), Greenness Index (GI), and Leaf Area Index (LAI) to jointly calculate crop growth characteristics. First, band calculation is performed on the multispectral image, and the green plant coverage of the crop canopy is calculated using the NDVI formula. The contrast of the green leaf areas is enhanced in combination with the GI index. Second, crop leaf density information is extracted based on the LAI index, and the image contrast is adjusted using an adaptive histogram equalization method to enhance the distinguishability of crop growth areas. The K-means clustering algorithm is used to classify crop growth areas, dividing them into areas of vigorous growth and areas of growth inhibition. Spatial autocorrelation analysis (Moran's I) is used to calculate the degree of clustering of the spatial distribution of crop growth, and a spatial variation model of growth changes is established in combination with the semi-variance analysis method, ultimately generating crop growth inhibition distribution data. The Laplace operator is used to sharpen the boundaries of the growth inhibition area to ensure data accuracy, and the discrete noise points are eliminated based on the morphological closing operation to make the crop growth inhibition distribution data more coherent.

[0086] Step S23: evaluating the degree of leaf yellowing and curling of the crop morphological images in the water flow area based on the crop growth inhibition distribution data to obtain leaf yellowing and curling degree data;

[0087] In the embodiment of the present invention, the evaluation of leaf yellowing and curling degree is based on the crop growth inhibition distribution data, and the yellowing degree and curling morphology of the leaves are quantified by color feature analysis, edge detection and texture feature extraction methods. First, the RGB image is converted to Lab color space using color space conversion technology, and the color space is extracted. Channel values ​​are used to calculate the yellowing index (YI) to characterize the degree of leaf yellowing. Secondly, the Canny edge detection algorithm is used to obtain the leaf contour, and the curvature change of the leaf edge is calculated based on the morphological gradient analysis method to quantify the degree of leaf curling. The local binary pattern (LBP) method is used to extract the texture features of the leaf surface, and the principal component analysis (PCA) is combined to perform feature dimensionality reduction to reduce redundant information. The extracted leaf yellowing index, curling curvature and texture feature data are normalized, and the support vector machine (SVM) classifier is used to classify the leaf health status. The area ratio of the leaf yellowing area is calculated based on the region growing algorithm, and a spatial correlation model is established in combination with the growth inhibition distribution data to evaluate the severity of leaf yellowing and curling in different areas, and finally generate leaf yellowing and curling degree data.

[0088] Step S24: extracting the soil volume water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

[0089] In this embodiment of the present invention, soil volumetric water saturation is extracted based on leaf yellowing and curling data, calculated through multi-source data fusion and time series analysis. First, soil moisture sensors are deployed within the farmland monitoring area. Soil moisture content is measured using the time domain reflectometry (TDR) method, supplemented with the frequency domain reflectometry (FDR) method to determine soil moisture variations at different depths. Second, based on the leaf yellowing and curling data, a random forest regression algorithm is used to construct a model for the relationship between leaf morphology and soil moisture. Least Squares Support Vector Regression (LSSVR) is used to optimize prediction accuracy. Soil moisture trends are analyzed using a time series decomposition method, and kriging interpolation is used to generate soil volumetric water saturation distribution maps for different regions. Based on the water transport characteristics of different soil layers, the Richards equation is used to calculate root zone soil water potential. This is combined with leaf stomatal conductance data to estimate crop transpiration, thereby inverting soil water deficit status. Finally, a spatial regression analysis method is used to construct a mathematical model for the relationship between soil moisture and leaf status, ensuring the scientific and accurate nature of the soil volumetric water saturation data.

[0090] Step S3 includes the following steps:

[0091] Step S31: Calculating the soil particle dispersion density based on the soil volume water saturation to obtain soil particle dispersion density data;

[0092] Step S32: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data;

[0093] Step S33: quantifying the nutrient leaching loss gradient of the soil nutrient element content change data based on the soil particle dispersion density data and the farmland irrigation water flow direction data to obtain nutrient leaching loss gradient data;

[0094] Step S34: quantifying the nitrogen cycle anomaly simulation based on the nutrient leaching loss gradient data and the soil volumetric water saturation to obtain soil nitrogen cycle anomaly data.

[0095] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0096] Step S31: Calculating the soil particle dispersion density based on the soil volume water saturation to obtain soil particle dispersion density data;

[0097] In this embodiment of the present invention, soil particle dispersion density is calculated based on soil volumetric water saturation through soil particle classification analysis and sedimentation velocity modeling. First, soil particle size at different sampling depths is measured using a laser particle size analyzer. The dispersion coefficient of soil particles of different sizes is calculated based on Stokes' sedimentation law. Second, combined with soil volumetric water saturation data, fractal theory is used to calculate the degree of soil particle aggregation, and the particle dispersion index (PDI) is used to characterize the discrete state of soil particles. A soil particle dispersion potential energy model is constructed, and the discrete element method (DEM) is used to simulate the dynamic process of particle dispersion under different soil moisture conditions. The adsorption, repulsion, and gravitational forces between particles are calculated. The particle dispersion density is then calculated by combining soil moisture-void ratio analysis. The particle dispersion density data is fitted using a Gaussian mixture model (GMM), and a spatial distribution map is generated using the Kriging interpolation method. Ultimately, the soil particle dispersion density data is obtained.

[0098] Step S32: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data;

[0099] In this embodiment of the present invention, soil nutrient content changes are extracted based on soil volumetric water saturation. Data is extracted using multi-element spectral analysis and dynamic equilibrium calculations, combined with farmland status monitoring data from irrigation basins. First, atomic absorption spectroscopy (AAS) and inductively coupled plasma optical emission spectroscopy (ICP-OES) are used to determine the initial concentrations of nutrients such as nitrogen, phosphorus, and potassium in soil samples. High-throughput ion chromatography is then used to obtain trace element content data (such as iron, zinc, and copper). Second, based on soil volumetric water saturation, a one-dimensional solute transport model is used to calculate the migration rates of nutrients under different water states. The release, absorption, and loss of soil nutrients are then estimated using the soil-plant system nutrient cycle equation. Coefficient of variation analysis is used to calculate the spatial nutrient variation intensity in different regions, and wavelet analysis is used to denoise the time series nutrient data to obtain true nutrient variation trends. Principal component analysis (PCA) is used to identify the factors that most influence soil nutrient dynamics, and a random forest regression algorithm is used to predict future nutrient variation trends. Ultimately, soil nutrient content variation data is obtained.

[0100] Step S33: quantifying the nutrient leaching loss gradient of the soil nutrient element content change data based on the soil particle dispersion density data and the farmland irrigation water flow direction data to obtain nutrient leaching loss gradient data;

[0101] In this embodiment of the present invention, nutrient leaching loss gradients are quantified based on soil particle density data and irrigation water flow data. This is done by establishing a soil solute transport model and a percolation coefficient calculation method. First, soil column experiments are used to determine solute penetration curves under different soil particle density conditions. The water infiltration rate is calculated using the Brutsaert nonlinear seepage equation, and the two-dimensional Darcy's law is used to estimate the hydraulic gradients of different soil layers. Second, based on the irrigation water flow data, the finite difference method (FDM) is used to simulate the flow path, and the solute balance equation is used to calculate the nutrient transport in each watershed segment. The influence of soil particle density on water flow rate is analyzed using Fourier transforms, and geostatistical methods are used to calculate nutrient leaching loss gradients in different soil layers. The nutrient leaching gradient distribution map is then plotted using the contour line method, and the spatial uniformity of nutrient leaching is calculated using the fractal dimension. Ultimately, nutrient leaching loss gradient data is obtained.

[0102] Step S34: quantifying the nitrogen cycle anomaly simulation based on the nutrient leaching loss gradient data and the soil volumetric water saturation to obtain soil nitrogen cycle anomaly data.

[0103] In an embodiment of the present invention, the simulation and quantification of nitrogen cycle anomalies are based on nutrient leaching loss gradient data and soil volume water saturation, and are quantitatively simulated by a biogeochemical cycle model and a nitrogen flux calculation method. First, the soil ammonia volatilization amount is determined using a static chamber method and gas chromatography analysis technology, and the release of N2O, NO and N2 during the soil nitrification-denitrification process is determined in combination with solid phase microextraction-mass spectrometry technology. Secondly, based on the nutrient leaching loss gradient data, the step-by-step inversion method is used to infer the nitrate nitrogen leaching loss amount, and the soil nitrogen conversion rate is calculated in combination with the first-order kinetic equation. The mass balance principle is used to establish a mathematical model of nitrogen input, consumption and loss, and the risk of nitrogen cycle anomalies under different water and fertilizer management modes is evaluated by the Monte Carlo random simulation method. The spatial regression analysis method is used to calculate the nitrogen accumulation in different regions, and the Kriging interpolation method is combined to generate a nitrogen anomaly distribution map, and finally the soil nitrogen cycle anomaly data is obtained.

[0104] Step S33 includes the following steps:

[0105] Step S331: performing soil particle size interval evaluation on the soil particle dispersion density data to obtain the soil particle size interval;

[0106] Step S332: performing hydrodynamic soil porosity fluctuation calculation based on the soil particle size range and the farmland irrigation water flow direction data to obtain hydrodynamic soil porosity fluctuation data;

[0107] Step S333: performing soil profile nutrient loss numerical iteration on the soil nutrient element content change data based on the hydrodynamic soil porosity fluctuation data to obtain soil profile nutrient loss iteration data;

[0108] Step S334: evaluating the expansion of the nutrient sinking range on the iterative data of nutrient loss in the soil profile to obtain the expansion of the nutrient sinking range;

[0109] Step S335: quantify the nutrient leaching loss gradient based on the soil profile nutrient loss iteration data and the nutrient sinking range expansion amount to obtain nutrient leaching loss gradient data.

[0110] In this embodiment of the present invention, soil particle size ranges are assessed based on soil particle dispersion density data and calculated using laser particle size analysis combined with probabilistic statistical methods. First, soil samples at different depths are measured using a laser particle size analyzer to obtain particle size distribution curves. The relative content of each particle size segment is then calculated based on Mie scattering theory. Next, the degree of soil particle aggregation is calculated using fractal dimension analysis, and the particle size distribution data is fitted using a Gaussian mixture model (GMM) to separate different particle size components. The probability density functions of particles of different sizes are calculated using a Poisson distribution model, and the normality of the particle size distribution is assessed using the Kolmogorov-Smirnov test. Subsequently, the particle size data is classified using the K-means clustering algorithm, the boundaries of each particle size component are calculated, and the volume fraction of each particle size range is calculated based on soil particle density and porosity. Finally, the resulting soil particle size range data is cross-compared with soil texture classification data from different farmland areas, and outliers are corrected to determine the soil particle size ranges. The calculation of hydrodynamic soil porosity fluctuations is based on soil particle size ranges and irrigation water flow data, using fluid dynamics methods combined with soil permeability analysis. First, the permeability coefficients of soils in different particle size ranges are calculated based on Darcy's law, and the effective porosity of each soil layer is estimated using the Kozeny-Carman equation. Second, a water transport model is constructed using the finite element method (FEM) in conjunction with irrigation water flow data to simulate hydrodynamic variations in different soil regions. The flow pattern in each watershed segment is calculated using the Reynolds number, and the Navier-Stokes equation is used to simulate variations in soil pore water velocity. Subsequently, the porosity-discharge relationship curve is combined with a nonlinear regression method to calculate hydrodynamic variations under different porosity conditions, and time series analysis is used to extract porosity fluctuation characteristics. Finally, a spatial interpolation method is used to generate a hydrodynamic soil porosity fluctuation distribution map, and error correction is performed to obtain hydrodynamic soil porosity fluctuation data. Nutrient loss from soil profiles is numerically simulated based on the hydrodynamic soil porosity fluctuation data and soil nutrient content variation data. First, based on hydrodynamic porosity data, the Richards equation was used to calculate water infiltration rates at different soil depths, and the solute transport equation (ADE) was combined to calculate the migration rates of different nutrients within the soil profile. Second, using a soil profile stratification model, the farmland soil was divided into multiple computational units, and the finite difference method (FDM) was used for numerical solution. The mineralization rate of soil organic matter was calculated using first-order reaction kinetic equations, and the release and uptake of nutrients such as nitrogen, phosphorus, and potassium were calculated using a solute balance model. Subsequently, multiple calculations were performed for each unit using a time-step iteration method, and the results were corrected using an error convergence criterion.Finally, a spatial distribution map of nutrient loss within the soil profile was generated based on the numerical iteration results. Nutrient loss trends for different soil types were calculated using multivariate regression analysis, resulting in iterative data for soil profile nutrient loss. The expansion of nutrient sinking was assessed based on this iterative data using statistical analysis combined with fluid dynamics calculations. First, the Kriging interpolation method was used to spatially interpolate the soil profile nutrient loss data, calculating nutrient concentration gradients at different soil depths. Second, the settling rates of nutrient particles of varying sizes were calculated based on Stokes's settling law, and the lateral diffusion of nutrients was calculated using Fick's diffusion law. Subsequently, a solute transport model (CDE) was used to simulate nutrient sinking rates under different water conditions, and a one-dimensional vertical seepage model was used to calculate nutrient accumulation in different soil layers. Geostatistical methods were used to calculate the nutrient sinking range for different soil types, and partial least squares regression (PLSR) analysis was used to identify key factors influencing the expansion of nutrient sinking. Finally, spatial regression analysis was used to calculate nutrient diffusion coefficients for different soil layers. Combined with tomography, data on the expansion of nutrient sinking were generated, yielding the expansion of nutrient sinking. Nutrient leaching loss gradients were quantified using iterative nutrient loss data from soil profiles and the expansion of nutrient sinking, employing a multidimensional coupled modeling approach. First, a two-dimensional Darcy seepage model was used to calculate water migration rates in different soil layers based on hydrodynamic porosity data, and a solute transport model was used to calculate the loss rates of different nutrients. Second, the nutrient loss gradients were discretized using Lagrange interpolation, and geostatistical methods were used to calculate the distribution of nutrient leaching loss gradients across different regions. Principal component analysis (PCA) was used to identify key variables influencing nutrient leaching gradients, and K-means clustering was used to stratify soil regions into different leaching loss levels. Numerical integration was then used to calculate the cumulative nutrient loss at different time scales, and Fourier transforms were used to analyze the cyclical changes in nutrient loss. Finally, geographic information system (GIS) technology was used to draw a nutrient leaching loss gradient distribution map, and a multivariate regression model was used to predict the future nutrient leaching change trend under different irrigation conditions to obtain nutrient leaching loss gradient data.

[0111] Step S333 includes the following steps:

[0112] The hydrodynamic nonlinear pore expansion relationship is analyzed for the hydrodynamic soil porosity fluctuation data to obtain the hydrodynamic nonlinear pore expansion relationship;

[0113] The nonlinear pore expansion relationship of hydrodynamics is used to analyze the shear stress of water flow in soil profiles and obtain the pore water flow stress data of soil profiles.

[0114] The velocity difference of different directions of the pore water flow stress data of the soil profile is deduced to obtain the velocity difference of different directions of the pore water flow in the profile.

[0115] Based on the hydrodynamic nonlinear pore expansion relationship and the velocity difference of the cross-sectional pores at different seepage directions, the seepage convection pressure data of the cross-sectional pores were obtained.

[0116] According to the Richards equation, the velocity difference of different cross-sectional pore infiltration directions and the cross-sectional pore infiltration convection pressure data, the soil nutrient element content change data were numerically iterated to obtain the soil profile nutrient loss iterative data.

[0117] In this embodiment of the present invention, the hydrodynamic nonlinear pore expansion relationship is analyzed based on hydrodynamic soil porosity fluctuation data. The nonlinear expansion law under different pore states is determined through soil mechanics and seepage mechanics calculations. First, based on soil particle size distribution and porosity data, the particle contact mechanics method is used to calculate the particle stress state under different pore structures, and the Hertzian contact theory is combined to analyze the stress characteristics of the pore boundary. Second, the Biot solid-liquid coupling equation is used to solve the pore deformation of the soil under the action of water flow, and the elastic deformation of the soil skeleton is calculated in combination with the generalized Hooke's law. Subsequently, based on the changes in pore water pressure, the Navier-Stokes equation is used to simulate the dynamic distribution of pore water, and the Reynolds average equation is used to calculate the pore expansion effect under different water flow rates. A nonlinear relationship between the pore expansion coefficient and hydrodynamic variables (pore water pressure, shear stress, infiltration rate, etc.) is constructed through multivariate regression analysis. The Boussinesq equation is used to correct for the hydrodynamic nonlinearity to obtain the hydrodynamic nonlinear pore expansion relationship data. The analysis of water shear stress in soil profiles is based on the hydrodynamic nonlinear pore expansion relationship and utilizes shear mechanics to calculate the water flow forces in different soil layers. First, the Darcy-Weisbach equation is used to calculate the water flow resistance under different pore states, and the Mohr-Coulomb criterion is used to calculate the effective shear stress between soil particles. Second, the Reynolds stress model in fluid dynamics is used to analyze the water flow disturbance at different pore scales in a layered manner, and the two-dimensional turbulent transport equation is used to solve the temporal evolution of the hydrodynamic shear force. Subsequently, the Bagnold effect is used to calculate the shear slip rate between particles, taking into account the arrangement of soil particles. The Fourier transform is used to analyze the effect of different shear frequencies on pore stability. Boundary layer theory is used to calculate the shear stress gradient within the soil profile, and the pore water shear stress distribution at different soil depths is obtained through numerical integration. Pore water stress data for the soil profile are then obtained. The azimuthally varying velocity differences are derived based on the pore water stress data for the soil profile, and the flow velocity differences between different soil layers are calculated using seepage mechanics and heterogeneous fluid mechanics. First, based on the theory of nonuniform seepage, the Laplace equation was used to calculate the water pressure field in different soil layers. Combined with the anisotropic permeability of the soil, Darcy's law was used to deduce the spatial variation of the water flow rate. Secondly, considering the soil pore structure, the Kozeny-Carman equation was used to calculate the seepage resistance in different directions, and the second-order differential equation in fluid dynamics was used to analyze the gradient variation of flow velocity along the profile. Subsequently, the calculus of variations was used to solve the optimal flow velocity distribution for different seepage paths. Combined with the soil particle size and pore morphology, the Navier-Stokes equation was used to calculate the flow velocity differences at different locations in the soil profile.Finally, the finite element analysis (FEM) method was used to numerically simulate the velocity differences across the soil profile. Velocity deviations were corrected based on soil pore deformation patterns, resulting in velocity differences across the profile for different pore infiltration orientations. The infiltration convection pressure was assessed based on the hydrodynamic nonlinear pore expansion relationship and the velocity differences across the profile for different pore infiltration orientations. Fluid statics and osmotic dynamics were used to calculate the pressure distribution under different flow conditions. First, the osmotic pressure for different pore structures was calculated based on Poiseuille's law, and the hydrodynamic potential energy distribution for different soil layers was solved using the soil water potential formula. Second, the Bernoulli equation was used to calculate the fluid energy distribution under different hydrodynamic conditions. The Navier-Stokes equation was used to solve the spatiotemporal evolution of the infiltration convection pressure, combining the shear force of the flow and the pore permeability resistance. Subsequently, the pressure changes along different infiltration paths were calculated based on the mass conservation equation, and the generalized Stokes law was used to analyze the effect of pore deformation on pressure under the action of water flow. Numerical simulations were used to calculate the seepage convection pressure for different soil types. Pressure fluctuations under pore flow conditions were analyzed using boundary layer theory to obtain cross-sectional pore seepage convection pressure data. Nutrient loss from soil profiles was numerically iteratively solved using the Richards equation, velocity differences at different cross-sectional pore seepage locations, and cross-sectional pore seepage convection pressure data using a water-soil-solute coupled computational approach. First, the soil water potential at different moisture contents was calculated using the Richards equation, and dynamic equations for nutrient dissolution and migration were derived using the water infiltration rate. Second, a one-dimensional solute transport model (CDE) was used to calculate the distribution of nitrogen, phosphorus, and potassium nutrients in the profile, and Fick's law was used to solve for nutrient diffusion rates. Subsequently, nutrient transport was calculated under different velocity differences using the convection-diffusion coupled equation, and the nutrient retention capacity of soil particles was solved using the adsorption-desorption equilibrium model. The nutrient loss process was discretized using the finite volume method (FVM), and multiple rounds of numerical iterations were performed using an error convergence criterion. Finally, the multivariate regression analysis method was used to analyze the influencing factors of nutrient loss rate, and the Kriging interpolation method was combined to generate the spatial distribution map of nutrient loss in the soil profile, and the iterative data of nutrient loss in the soil profile were obtained.

[0118] Step S34 includes the following steps:

[0119] Step S341: performing soil oxygen content decrease gradient analysis based on soil volume water saturation to obtain soil oxygen content decrease gradient data;

[0120] Step S342: performing an anaerobic microbial activity enhancement function relationship analysis based on the nutrient leaching loss gradient data and the soil oxygen content decrease gradient data to obtain an anaerobic microbial activity enhancement relationship function;

[0121] Step S343: performing denitrification rate enhancement coupling according to the anaerobic microbial activity enhancement relationship function to obtain denitrification rate enhancement coupling data;

[0122] Step S344: performing asymptotic fitting estimation of nitrogen release on the denitrification rate enhancement coupling data to obtain nitrogen release fitting estimation data;

[0123] Step S345: Perform nitrogen cycle anomaly simulation and quantification based on the nitrogen release fitting estimation data to obtain soil nitrogen cycle anomaly data.

[0124] In this embodiment of the present invention, soil oxygen gradient analysis is based on soil volumetric water saturation data. The gradient of soil oxygen gradient is determined by analyzing the effect of water on soil oxygen distribution. First, Darcy's law is used to estimate water permeability and water velocity, combining soil aeration and water content, and water saturation at different soil depths. Then, based on soil water saturation, Fick's law of diffusion is used to describe the oxygen diffusion process in the soil, thereby estimating the oxygen concentration gradient under different water conditions. Oxygen concentrations in different soil layers are experimentally measured, and a mathematical model is developed to analyze the variation of oxygen concentration with water saturation, combining the relationship between water permeability and pore structure. In this model, increasing water saturation restricts gas exchange in soil pores, reducing oxygen diffusion capacity and leading to an intensified oxygen gradient. Numerical integration methods are used to simulate oxygen concentration variations in different soil types. The specific oxygen gradient is determined based on the soil type (e.g., clay, sand, etc.) and its gas diffusion characteristics. Finally, multiple regression analysis is performed to obtain soil oxygen gradient data. The enhancement function relationship between anaerobic microbial activity and water and oxygen conditions was derived by combining nutrient leaching loss gradient data and soil oxygen decline gradient data. First, based on the leaching loss gradients of nitrogen and phosphorus in the soil, Fick's law and the Advance-Dispersion model were used to simulate the nutrient flow and migration process, thereby determining the longitudinal distribution of nutrients in the soil. Then, combined with the soil oxygen decline gradient, the Monod equation was used to describe the growth and reproduction rate of anaerobic microorganisms in a hypoxic environment. The growth rate in the Monod equation exhibits a nonlinear relationship with oxygen concentration and nutrient concentration. Based on this, an enhancement function for anaerobic microbial activity was established by combining water, oxygen concentration, and nutrient loss data. By comparing anaerobic microbial activity under different water and oxygen conditions, the quantitative relationship between nutrient loss and microbial activity was determined. The function parameters were obtained by data fitting, and a nonlinear function was constructed to describe the interaction between water, oxygen, and nutrients. Finally, the function was optimized using polynomial regression or Lagrange interpolation to obtain the enhancement function for anaerobic microbial activity. Denitrification rate enhancement coupling is based on the anaerobic microbial activity enhancement relationship function to derive the rate changes during the denitrification process. First, based on the relationship between anaerobic microbial activity and water, oxygen, and nutrient concentrations, the Michaelis-Menten equation is used to describe the dependence of the denitrification rate on substrate concentration, thereby obtaining the initial denitrification rate. Subsequently, the relationship between the denitrification reaction rate and microbial activity is quantified based on the Hill equation, taking into account the metabolic characteristics of microorganisms in anaerobic environments. By introducing variables such as soil nitrogen source concentration, oxygen concentration, and temperature, a multi-factor coupled denitrification rate model is established.The denitrification rate in this model is not only limited by oxygen but also affected by soil water saturation and nutrient availability. To further enhance the simulation of the denitrification rate, the Newton method was used to optimize the parameters of the denitrification rate model, combined with experimental data. This resulted in a reinforced coupling relationship between the denitrification rate and anaerobic microbial activity. Finally, the denitrification rate was numerically simulated using the reinforced coupling data to obtain specific data for the reinforced coupling of the denitrification rate. Asymptotic fitting estimation of nitrogen release was used to estimate the amount of nitrogen released from the soil based on the reinforced coupling data of the denitrification rate. First, a cumulative nitrogen release model was used to simulate the nitrogen release process based on the denitrification rate data. By dynamically tracking soil nitrogen sources (such as nitrate) and combining experimental data on nitrogen release, the least squares method was used to fit the relationship between nitrogen release and the denitrification rate. The fitting process took into account the influence of factors such as soil temperature, moisture, and oxygen concentration on nitrogen release, resulting in a gradual process of nitrogen release. Next, an exponential decay model was used to describe the relationship between nitrogen release rate and time, analyzing the changing trends of nitrogen release over different time periods. Through data analysis, the cumulative amount of nitrogen release was determined, and a model was established to link the nitrogen release rate and denitrification process. Finally, through multiple fitting, asymptotic estimates of nitrogen release were obtained, which accurately reflect the nitrogen release amount under different soil conditions. The simulation and quantification of nitrogen cycle anomalies was based on the nitrogen release estimate data and combined with the soil nitrogen cycle process. First, the Richards equation was used to simulate the dynamic changes in soil moisture and gas. The nitrogen transformation and loss rates in the soil were calculated based on the distribution of soil nitrogen sources. Subsequently, the nitrogen cycle model was used to dynamically simulate the nitrogen release, absorption, and transformation processes, simulating nitrogen cycle anomalies under different soil environments. By analyzing the changing trends of the soil nitrogen release rate and combining factors such as soil nitrogen content, temperature, and humidity, nitrogen cycle anomalies were detected. For example, excessive nitrogen release indicates anaerobic soil conditions and abnormally intensified denitrification. The dynamic nitrogen cycle process was discretized using a difference equation to obtain quantitative data on nitrogen cycle anomalies. Finally, by comparing the nitrogen cycle data under different soil types and different management measures, the abnormal situation of soil nitrogen cycle was quantitatively evaluated and the abnormal data of soil nitrogen cycle were obtained.

[0125] Step S344 includes the following steps:

[0126] Based on the denitrification rate enhanced coupling data, nitrogen release is mapped proportionally to obtain rate-matched nitrogen release mapping data;

[0127] Based on the denitrification rate enhancement coupling data, the rate-matched nitrogen release mapping data were analyzed for multiple extreme values ​​of ammonia release, and the rate-matched ammonia release multiple extreme values ​​were obtained.

[0128] Perform time series delay calculation on multiple extreme values ​​of rate-matched ammonia release to obtain multiple extreme value time series delay data;

[0129] Based on the multiple extreme value time series delay data and the rate matching of multiple extreme values ​​of ammonia release, the mean difference of cyclic repeatability release is calculated to obtain the extreme value mean difference of ammonia release cycle;

[0130] The nitrogen release asymptotic fitting estimation was performed on the extreme mean difference of the ammonia release cycle according to the maximum likelihood estimation method to obtain the nitrogen release fitting estimation data.

[0131] In an embodiment of the present invention, based on the denitrification rate enhanced coupling data, a geometric mapping method is used to convert the denitrification rate data into nitrogen release data. In this process, first, based on the relationship between the denitrification rate and nitrogen release, a proportional model of the denitrification rate and nitrogen release is established. The model obtains the basic relationship between the denitrification rate and the nitrogen release rate by performing linear regression analysis on the experimental data. Then, through geometric mapping, for different denitrification rates, it is converted into the corresponding nitrogen release rate. The geometric mapping process is carried out in the following way: Assuming the denitrification rate and nitrogen release rate The relationship between can be expressed by the proportional coefficient k, that is, =k× . The proportional coefficient k is determined experimentally or obtained from relevant technical literature, and is corrected in combination with environmental factors such as different soil moisture and temperature. The denitrification rate data at different time points are geometrically mapped to obtain the nitrogen release rate data corresponding to each time point, and the data is smoothed by interpolation method to obtain rate-matched nitrogen release mapping data with high precision. Based on the rate-matched nitrogen release mapping data, multiple extreme value analysis of ammonia release is performed. In this step, the peak and valley values ​​of ammonia release are first identified based on the rate-matched nitrogen release mapping data. By solving the derivatives during the ammonia release process, the extreme points of the ammonia release rate over time can be identified. These extreme points represent the maximum release rate (peak value) and minimum release rate (valley value) during the ammonia release process. For the analysis of multiple extreme values, the second-order derivative method is used to determine multiple extreme points, and the existence and position of these extreme values ​​are further verified by the extreme value detection algorithm (such as solving the Lagrange multiplier method). During this process, special attention was paid to the periodic variations in the ammonia release rate. Periodic analysis methods, such as the Fast Fourier Transform (FFT), were used to analyze the frequency of ammonia release, identifying periodic peaks and valleys. By calculating the parameters of these periodic variations, multiple extreme value data for rate-matched ammonia release were obtained. Time delay calculations were performed based on these multiple extreme value data to obtain time series delay data during the ammonia release process. First, based on the multiple extreme value data, an autoregressive integrated moving average (ARIMA) model was used to analyze the temporal variation patterns of ammonia release. Through model fitting, the time delays between each extreme value point were calculated. During this process, the time series was smoothed using the differencing method to eliminate trend variations and ensure that the data was suitable for ARIMA modeling. Then, time delay analysis techniques were used to determine the time series delays between each extreme value point, i.e., the temporal offset of each extreme value point. To more accurately capture the dynamics of ammonia release, a cross-correlation function (CCF) was used to compare ammonia release rates at different time points and calculate the delay between peak and valley values. Time series delay calculations generated multiple extreme value time series delay data, allowing for further analysis of the temporal correlations within the ammonia release process. Combining these multiple extreme value time series delay data with multiple extreme value data from rate-matched ammonia release, the mean difference of extreme values ​​within the ammonia release cycle was calculated. First, based on the multiple extreme value time series delay data, the cycle period of the ammonia release process—that is, the time interval between extreme values—was calculated. Based on the known multiple extreme values, the mean difference within each cycle was calculated using the mean method. The cyclic stability of the ammonia release process was then reflected by calculating the mean difference across all cycles. Specifically, the range method was used to calculate the range of each cycle—the difference between the maximum and minimum extreme values ​​within a complete cycle. The ranges across all cycles were then averaged to obtain the mean difference of extreme values ​​within the ammonia release cycle.Statistical analysis of the extreme mean differences across multiple cycles was used to determine the cyclical pattern of ammonia release and quantify the degree of fluctuation within each cycle. The calculation of the extreme mean differences was based on the root mean square error (RMSE) method in statistics, ensuring that the differences between cycles accurately reflected the dynamic changes in ammonia release. Based on the extreme mean difference data, maximum likelihood estimation (MLE) was used to perform an asymptotic fit of nitrogen release. First, based on the distribution of the extreme mean differences, it was assumed that the asymptotic process of nitrogen release followed a certain statistical distribution (such as a normal distribution or a lognormal distribution). The maximum likelihood estimation (MLE) method was then used to determine the parameters of the fitting function. The MLE method estimates parameters by maximizing the likelihood function. In this step, the optimal parameters related to the nitrogen release process were calculated by deriving the MLE from the ammonia release time series data. Specifically, a likelihood function was constructed to represent the probability distribution of the parameters given the data. This likelihood function was then solved using an optimization method (such as the Newton method or the quasi-Newton method) to obtain the optimal fitting parameters. Ultimately, this estimation method matches the asymptotic behavior of the nitrogen release process with the actual data, resulting in a fitted estimate of nitrogen release. This fitted estimate provides a precise mathematical basis for subsequent quantification of nitrogen release.

[0132] Step S4 includes the following steps:

[0133] Step S41: performing convolution calculation on the soil nitrogen cycle abnormal data to obtain soil nitrogen cycle abnormal convolution data;

[0134] Step S42: performing soil water saturation control on the soil volume water saturation according to the soil nitrogen cycle abnormal convolution data and the nutrient leaching loss gradient data to obtain soil water saturation control data;

[0135] Step S43: Based on the soil moisture saturation control data and the soil nitrogen cycle abnormality convolution data, the leaf yellowing and curling degree data is subjected to soil water and fertilizer intelligent controlled release analysis to obtain soil water and fertilizer intelligent controlled release data, and the soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield.

[0136] In an embodiment of the present invention, a convolution calculation is first performed on the soil nitrogen cycle anomaly data, and the characteristic change trends of the data are analyzed through the convolution operation. The convolution calculation is achieved by convolving the soil nitrogen cycle anomaly data with a convolution kernel (or filter). Specifically, an appropriate data window size and convolution kernel are first selected based on the time series characteristics of the soil nitrogen cycle anomaly data. The convolution kernel is selected based on experimental data and research on similar data in relevant literature. By gradually sliding the convolution kernel and convolving the soil nitrogen cycle anomaly data, a weighted sum is obtained at each time point, that is, a smoothed result of the soil nitrogen cycle characteristics within each time period. The convolution calculation performs a weighted superposition of the soil nitrogen data at different time points to obtain smoother and more representative soil nitrogen cycle data, thereby revealing the overall trend of the nitrogen cycle. Ultimately, the soil nitrogen cycle anomaly convolution data obtained through the convolution calculation provides key input for the next step, has high time series characteristics, and can effectively reflect the main patterns of nitrogen changes in the soil. Soil water saturation is regulated based on the soil nitrogen cycle anomaly convolution data and nutrient leaching loss gradient data obtained in step S41. First, the soil nitrogen cycle anomaly convolution data is used to analyze the changing trends in soil nitrogen content. Combined with the nutrient leaching loss gradient data, the relationship between soil water and nitrogen is derived. During this process, a correlation between nitrogen loss and soil water saturation is assumed, and regression analysis is used to determine the relationship between soil nitrogen changes and water content. A mathematical model is developed to simulate the coupling mechanism between changes in soil water saturation and the nitrogen cycle, tailored to different soil types and environmental conditions. The model uses a series of weighting factors to reflect the impact of water saturation on nitrogen release rate and nutrient leaching. Specifically, soil water saturation is dynamically regulated by setting a soil moisture threshold and a critical point for nitrogen loss. In this step, water parameters are adjusted to reduce nitrogen loss caused by soil oversaturation, ensuring that the soil maintains a suitable nitrogen cycle under optimal water conditions. Model calculations during the regulation process are implemented using dynamic equilibrium equations, resulting in optimized soil water saturation regulation data, which provides a basis for analysis in subsequent steps. Based on the soil water saturation control data and the soil nitrogen cycle abnormal convolution data obtained in step S42, an intelligent controlled release analysis of soil water and fertilizer usage based on the leaf yellowing and curling degree data is performed. First, the soil water saturation control data is used in combination with the soil nitrogen cycle data to analyze the effects of water and nitrogen on crop growth, especially on the growth status of leaves. Through regression analysis, a mathematical model is established between soil water and nitrogen content and the yellowing and curling degree of crop leaves, and the influence of water and nitrogen on the state of crop leaves at different levels is obtained. The leaf yellowing and curling degree data is used to determine the water and fertilizer needs of crops at different growth stages, and the soil water and fertilizer usage is intelligently controlled and calculated based on these needs.This step uses a data-driven approach, combining the crop's growth needs with the actual soil conditions, and employs a dynamic optimization algorithm (such as particle swarm optimization (PSO)) to adjust the soil water and fertilizer release pattern. During the calculation process, fertilization and irrigation amounts are adjusted based on real-time feedback from the crop's leaf status to maintain an optimal growth environment. By calculating the difference between the amount of water and fertilizer in the soil and the crop's needs for water and fertilizer, the intelligent controlled-release system can dynamically adjust the release rate of water and fertilizer, thereby providing sufficient water and fertilizer support for the crop and reducing nutrient waste. Ultimately, the resulting intelligent controlled-release data on soil water and fertilizer usage is transmitted to the crop management terminal, enabling intelligent water and fertilizer analysis to improve crop yields and achieve precise fertilization and irrigation.

[0137] Step S43 includes the following steps:

[0138] Step S431: performing a nutrient deficiency long-term cumulative effect analysis on the soil nitrogen cycle abnormality convolution data to obtain nutrient deficiency long-term cumulative effect data;

[0139] Step S432: performing a nutrient absorption capacity degradation assessment on the leaf yellowing and curling degree data to obtain crop nutrient absorption capacity degradation data;

[0140] Step S433: analyzing the soil aeration enhancement ratio based on the soil water saturation control data to obtain the soil aeration enhancement ratio;

[0141] Step S434: performing a phased multi-objective optimization of the water-fertilizer ratio based on the long-term cumulative effect data of nutrient deficiency and the soil aeration enhancement ratio on the crop nutrient absorption capacity degradation data to obtain the multi-objective optimization data of the water-fertilizer ratio;

[0142] Step S435: Perform intelligent controlled-release analysis of soil water and fertilizer usage based on the multi-objective optimization data of water and fertilizer ratio, obtain soil water and fertilizer usage intelligent controlled-release data, and send the soil water and fertilizer usage intelligent controlled-release data to the terminal to perform intelligent water and fertilizer analysis for improving crop yield.

[0143] In an embodiment of the present invention, a long-term cumulative effect analysis of nutrient deficiency is performed based on the soil nitrogen cycle abnormal convolution data obtained in step S41. This step aims to analyze the long-term impact of long-term nutrient deficiency on soil and crop growth through changes in the nitrogen cycle in the soil. In order to achieve this goal, a multi-period cumulative analysis of the soil nitrogen cycle data is performed to establish a correlation between nitrogen loss in the soil and soil nutrient deficiency. The specific approach is to calculate the nitrogen loss value in each time period and accumulate it to evaluate the consumption trend of soil nutrients due to long-term deficiency. This process uses the cumulative effect analysis formula: ,in, Indicates the The cumulative effect of nutrient deficiency over time, For the Nitrogen concentration during the period, For the The loss rate of a certain period of time. Through the analysis of long-term cumulative effects, the cumulative effect data of nutrient deficiency can be obtained, that is, the gradual impact of nitrogen loss on soil nutrients over a long period of time, providing data support for subsequent nutrient regulation. On the basis of step S431, the data on the yellowing and curling of leaves are used to evaluate the degradation of crop nutrient absorption capacity. The health status of the leaves directly reflects the nutrient absorption capacity of the crop. The yellowing and curling of leaves are usually caused by nutrient deficiency or absorption disorders. First, the yellowing and curling of the leaves are quantitatively analyzed by image processing technology to obtain the yellowing area and curling angle of the leaves. On this basis, combined with the nutrient content such as nitrogen in the soil, the regression analysis method is used to construct a mathematical model between nutrient absorption capacity and leaf health status. The model estimates the degree of degradation of crop nutrient absorption capacity through multivariate regression of the visual characteristics of the leaves and soil nutrient data. Specifically, the following linear regression model is used to describe the degradation of nutrient absorption capacity:

[0144] ;

[0145] in, A score indicating the degradation of the crop's nutrient uptake capacity, is the area of ​​yellowed leaves, is the leaf curl angle, is the nitrogen content in the soil, , , , is the regression coefficient (e.g., =2.5, =0.3, =0.5, =0.1). Through this analysis, the degradation data of the nutrient absorption capacity of crops is obtained, which provides data support for the subsequent optimization of the water-fertilizer ratio. Based on the soil water saturation control data in step S42, the soil permeability enhancement ratio analysis is performed. The permeability of the soil plays a vital role in the respiration and nutrient absorption of the root system. Especially when the soil saturation is too high or too low, the permeability will drop significantly, affecting the growth of crops. Therefore, first, the current soil saturation is determined by real-time monitoring of the soil water saturation data, and the mathematical relationship model between soil permeability and water saturation is established for analysis. The physical and chemical model is used to calculate the soil permeability enhancement ratio, and the formula is as follows:

[0146] ;

[0147] in, is the soil aeration enhancement ratio, Ksat is the hydraulic conductivity of the soil under saturation, θ is the soil water saturation, and Kref is the reference soil aeration constant. Real-time calculation of the soil aeration enhancement ratio allows assessment of soil aeration capacity under different water conditions, providing effective data support for subsequent optimization of water-fertilizer ratios. This step performs a phased multi-objective optimization of the water-fertilizer ratio based on the long-term cumulative effect data of nutrient depletion, nutrient uptake capacity degradation data, and soil aeration enhancement ratio data obtained in steps S431, S432, and S433. The optimization goal is to dynamically adjust the water-fertilizer ratio to achieve optimal crop growth by comprehensively considering soil nutrient status, crop nutrient requirements, and soil aeration. A multi-objective particle swarm optimization (PSO) algorithm is used to solve the problem. The optimization objective function considers multiple objectives, including crop growth, nutrient use efficiency, and water and fertilizer losses. Through optimization, the optimal water-fertilizer ratio for each phase is obtained, providing a scientific basis for crop nutrient replenishment and water management. Based on the multi-objective optimization data of the water-fertilizer ratio obtained in step S434, an intelligent controlled-release analysis of the soil water and fertilizer usage is performed. The purpose of the intelligent controlled-release analysis is to dynamically adjust the time and amount of fertilization and irrigation according to the growth requirements of the crops and the moisture conditions of the soil to ensure the accurate supply of water and fertilizer. In this process, a fuzzy control algorithm is used to regulate the soil water and fertilizer usage in real time, and the irrigation and fertilization strategies are dynamically adjusted according to the real-time data. The specific control strategy adjusts the amount of fertilizer and irrigation in each time period through the control system algorithm to highly match it with the crop growth requirements. Finally, the intelligent controlled-release data of the soil water and fertilizer usage is sent to the terminal device so that the optimized water and fertilizer management strategy can be executed in real time to achieve accurate crop yield improvement.

[0148] The present invention also provides a water and fertilizer intelligent analysis system based on regionalized crop yield improvement in farmland, which is used to execute the water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland as described above. The water and fertilizer intelligent analysis system based on regionalized crop yield improvement in farmland includes:

[0149] The irrigation basin farmland status monitoring module is used to obtain farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and conduct real-time monitoring of the irrigation basin farmland status, thereby generating irrigation basin farmland status monitoring data;

[0150] The soil volumetric water saturation extraction module is used to collect crop morphological images based on farmland irrigation water flow data using electronic monitoring equipment to obtain crop morphological images in the water flow area; evaluate the degree of leaf yellowing and curling in the crop morphological images in the water flow area to obtain leaf yellowing and curling degree data; and extract the soil volumetric water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volumetric water saturation of crops with yellowing and curling leaves.

[0151] The soil nitrogen cycle anomaly analysis module is used to extract nutrient element content changes from irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantify the nutrient leaching loss gradient of soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulate and quantify nitrogen cycle anomalies based on nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data;

[0152] The intelligent analysis and control module for water and fertilizer usage is used to perform intelligent controlled release analysis of soil water and fertilizer usage based on the data on leaf yellowing and curling abnormalities in the soil nitrogen cycle, obtain intelligent controlled release data of soil water and fertilizer usage, and send the intelligent controlled release data of soil water and fertilizer usage to the terminal to perform intelligent water and fertilizer analysis to improve crop yield.

[0153] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland, characterized in that: The following steps are involved: Step S1: Acquire farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and perform real-time monitoring of the farmland status in the irrigation basin, thereby generating irrigation basin farmland status monitoring data; Step S2: Using electronic monitoring equipment and based on farmland irrigation water flow data, crop morphological images are collected to obtain crop morphological images in the water flow area; the degree of leaf yellowing and curling in the crop morphological images in the water flow area is assessed to obtain leaf yellowing and curling degree data; and based on the leaf yellowing and curling degree data, the soil volume water saturation of crops with yellowing and curling leaves in the irrigation basin is extracted from the farmland status monitoring data to obtain the soil volume water saturation of crops with yellowing and curling leaves. Step S3: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantifying the nutrient leaching loss gradient of the soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulating and quantifying nitrogen cycle anomalies based on the nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data; Step S4: Based on the soil nitrogen cycle abnormality data, the leaf yellowing and curling degree data is analyzed for soil water and fertilizer intelligent controlled release, and soil water and fertilizer intelligent controlled release data is obtained. The soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield. Step S4 includes: Step S41: performing convolution calculation on the soil nitrogen cycle abnormal data to obtain soil nitrogen cycle abnormal convolution data; Step S42: performing soil water saturation control on the soil volume water saturation according to the soil nitrogen cycle abnormal convolution data and the nutrient leaching loss gradient data to obtain soil water saturation control data; Step S43: Based on the soil water saturation control data and the soil nitrogen cycle abnormality convolution data, the leaf yellowing and curling degree data is subjected to soil water and fertilizer intelligent controlled release analysis to obtain soil water and fertilizer intelligent controlled release data, and the soil water and fertilizer intelligent controlled release data is sent to the terminal to perform water and fertilizer intelligent analysis for improving crop yield. Step S43 includes: Step S431: performing a nutrient deficiency long-term cumulative effect analysis on the soil nitrogen cycle abnormality convolution data to obtain nutrient deficiency long-term cumulative effect data; Step S432: performing a nutrient absorption capacity degradation assessment on the leaf yellowing and curling degree data to obtain crop nutrient absorption capacity degradation data; Step S433: analyzing the soil aeration enhancement ratio based on the soil water saturation control data to obtain the soil aeration enhancement ratio; Step S434: performing a phased multi-objective optimization of the water-fertilizer ratio based on the long-term cumulative effect data of nutrient deficiency and the soil aeration enhancement ratio on the crop nutrient absorption capacity degradation data to obtain the multi-objective optimization data of the water-fertilizer ratio; Step S435: Perform intelligent controlled-release analysis of soil water and fertilizer usage based on the multi-objective optimization data of water and fertilizer ratio, obtain soil water and fertilizer usage intelligent controlled-release data, and send the soil water and fertilizer usage intelligent controlled-release data to the terminal to perform intelligent water and fertilizer analysis for improving crop yield.

2. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: using electronic monitoring equipment and based on the farmland irrigation water flow data, crop morphology images are collected to obtain crop morphology images in the water flow area; Step S22: performing growth inhibition distribution recognition on the crop morphological image in the water flow area to obtain crop growth inhibition distribution data; Step S23: evaluating the degree of leaf yellowing and curling of the crop morphological images in the water flow area based on the crop growth inhibition distribution data to obtain leaf yellowing and curling degree data; Step S24: extracting the soil volume water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volume water saturation of crops with yellowing and curling leaves.

3. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: Calculating the soil particle dispersion density based on the soil volume water saturation to obtain soil particle dispersion density data; Step S32: extracting nutrient element content changes from the irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; Step S33: quantifying the nutrient leaching loss gradient of the soil nutrient element content change data based on the soil particle dispersion density data and the farmland irrigation water flow direction data to obtain nutrient leaching loss gradient data; Step S34: quantifying the nitrogen cycle anomaly simulation based on the nutrient leaching loss gradient data and the soil volumetric water saturation to obtain soil nitrogen cycle anomaly data.

4. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 3 is characterized in that: Step S33 includes the following steps: Step S331: performing soil particle size interval evaluation on the soil particle dispersion density data to obtain the soil particle size interval; Step S332: performing hydrodynamic soil porosity fluctuation calculation based on the soil particle size range and the farmland irrigation water flow direction data to obtain hydrodynamic soil porosity fluctuation data; Step S333: performing soil profile nutrient loss numerical iteration on the soil nutrient element content change data based on the hydrodynamic soil porosity fluctuation data to obtain soil profile nutrient loss iteration data; Step S334: evaluating the expansion of the nutrient sinking range on the iterative data of nutrient loss in the soil profile to obtain the expansion of the nutrient sinking range; Step S335: quantify the nutrient leaching loss gradient based on the soil profile nutrient loss iteration data and the nutrient sinking range expansion amount to obtain nutrient leaching loss gradient data.

5. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 3 is characterized in that: Step S333 includes the following steps: The hydrodynamic nonlinear pore expansion relationship is analyzed for the hydrodynamic soil porosity fluctuation data to obtain the hydrodynamic nonlinear pore expansion relationship; The nonlinear pore expansion relationship of hydrodynamics is used to analyze the shear stress of water flow in soil profiles and obtain the pore water flow stress data of soil profiles. The velocity difference of different directions of the pore water flow stress data of the soil profile is deduced to obtain the velocity difference of different directions of the pore water flow in the profile. Based on the hydrodynamic nonlinear pore expansion relationship and the velocity difference of the cross-sectional pores at different seepage directions, the seepage convection pressure data of the cross-sectional pores were obtained. According to the Richards equation, the velocity difference of different cross-sectional pore infiltration directions and the cross-sectional pore infiltration convection pressure data, the soil nutrient element content change data were numerically iterated to obtain the soil profile nutrient loss iterative data.

6. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 3 is characterized in that: Step S34 includes the following steps: Step S341: performing soil oxygen content decrease gradient analysis based on soil volume water saturation to obtain soil oxygen content decrease gradient data; Step S342: performing an anaerobic microbial activity enhancement function relationship analysis based on the nutrient leaching loss gradient data and the soil oxygen content decrease gradient data to obtain an anaerobic microbial activity enhancement relationship function; Step S343: performing denitrification rate enhancement coupling according to the anaerobic microbial activity enhancement relationship function to obtain denitrification rate enhancement coupling data; Step S344: performing asymptotic fitting estimation of nitrogen release on the denitrification rate enhancement coupling data to obtain nitrogen release fitting estimation data; Step S345: Perform nitrogen cycle anomaly simulation and quantification based on the nitrogen release fitting estimation data to obtain soil nitrogen cycle anomaly data.

7. The water and fertilizer intelligent analysis method based on regionalized crop yield improvement in farmland according to claim 6 is characterized in that: Step S344 includes the following steps: Based on the denitrification rate enhanced coupling data, nitrogen release ratio mapping was performed to obtain rate-matched nitrogen release mapping data; Based on the denitrification rate enhancement coupling data, the rate-matched nitrogen release mapping data were analyzed for multiple extreme values ​​of ammonia release, and the rate-matched ammonia release multiple extreme values ​​were obtained. Perform time series delay calculation on multiple extreme values ​​of rate-matched ammonia release to obtain multiple extreme value time series delay data; Based on the multiple extreme value time series delay data and the rate matching of multiple extreme values ​​of ammonia release, the mean difference of cyclic repeatability release is calculated to obtain the extreme value mean difference of ammonia release cycle; The nitrogen release asymptotic fitting estimation was performed on the extreme mean difference of the ammonia release cycle according to the maximum likelihood estimation method to obtain the nitrogen release fitting estimation data.

8. An intelligent water and fertilizer analysis system based on regionalized crop yield improvement in farmland, characterized by: The method for intelligent water and fertilizer analysis based on regionalized crop yield improvement in farmland according to claim 1 is used to implement the method. The system for intelligent water and fertilizer analysis based on regionalized crop yield improvement in farmland comprises: The irrigation basin farmland status monitoring module is used to obtain farmland regional structure data; extract irrigation water flow direction from the farmland regional structure data to obtain farmland irrigation water flow direction data; deploy multi-directional sensors based on the farmland irrigation water flow direction data, and conduct real-time monitoring of the irrigation basin farmland status, thereby generating irrigation basin farmland status monitoring data; The soil volumetric water saturation extraction module is used to collect crop morphological images based on farmland irrigation water flow data using electronic monitoring equipment to obtain crop morphological images in the water flow area; evaluate the degree of leaf yellowing and curling in the crop morphological images in the water flow area to obtain leaf yellowing and curling degree data; and extract the soil volumetric water saturation of crops with yellowing and curling leaves from the irrigation basin farmland status monitoring data based on the leaf yellowing and curling degree data to obtain the soil volumetric water saturation of crops with yellowing and curling leaves. The soil nitrogen cycle anomaly analysis module is used to extract nutrient element content changes from irrigation basin farmland status monitoring data based on soil volume water saturation to obtain soil nutrient element content change data; quantify the nutrient leaching loss gradient of soil nutrient element content change data to obtain nutrient leaching loss gradient data; and simulate and quantify nitrogen cycle anomalies based on nutrient leaching loss gradient data to obtain soil nitrogen cycle anomaly data; The intelligent analysis and control module for water and fertilizer usage is used to perform intelligent controlled release analysis of soil water and fertilizer usage based on the data on leaf yellowing and curling abnormalities in the soil nitrogen cycle, obtain intelligent controlled release data of soil water and fertilizer usage, and send the intelligent controlled release data of soil water and fertilizer usage to the terminal to perform intelligent water and fertilizer analysis to improve crop yield.

Images