Farmland pollutant monitoring system and method based on Internet of Things

Abstract

The invention relates to the technical field of farmland pollutant monitoring, in particular to a farmland pollutant monitoring system and method based on the Internet of Things, and the method comprises the steps: collecting soil environment data to obtain real-time monitoring data, and analyzing the water carbon and phosphorus circulation process of a farmland region to obtain the spatial distribution characteristics of soil pollutants; determining a pollutant migration path based on the spatial distribution characteristics, and performing fixed-point scanning on the high-risk pollution accumulation area based on a high-precision spectral sensor to obtain pollutant spectral characteristic data; performing matching analysis on the pollutant spectral characteristic data and standard pollutant characteristics in a preset pollutant traceability database to obtain pollutant types and pollution source directions; comparing the pollution diffusion result with a preset safety diffusion threshold value to obtain a multi-level early warning instruction; and formulating a differentiated pollution control scheme based on a pollution diffusion result. Through the Internet of Things technology and automatic monitoring and analysis, farmland pollution monitoring and management are more intelligent.

Description

Technical Field

[0001] This invention relates to the field of farmland pollutant monitoring technology, specifically to an Internet of Things-based farmland pollutant monitoring system and method. Background Technology

[0002] Traditional methods typically rely on periodic manual sampling and laboratory analysis, failing to provide real-time data. This means that pollution may go undetected and unresponsive when it occurs, leading to its spread or exacerbation. Traditional methods also typically provide data from limited sampling points, failing to comprehensively reflect the spatial distribution of pollutants across the entire farmland area. Furthermore, they are often ineffective at tracing pollutant migration paths, especially in complex farmland environments. Traditional methods rely heavily on manual inspections or large-scale sampling to infer pollution sources, which is easily limited by local samples and may lack accuracy in the tracing process. Pollution detection in traditional methods is mostly post-hoc analysis, lacking effective early warning systems, making it impossible to identify and prevent pollution risks in advance. Finally, traditional methods often adopt a uniform approach to pollution control, without developing differentiated control plans for different pollution sources and regional characteristics. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide an agricultural pollutant monitoring system and method based on the Internet of Things.

[0004] The technical solution adopted to solve the above-mentioned technical problems is: a method for monitoring farmland pollutants based on the Internet of Things, including: Preferably, the monitoring data includes soil moisture content, soil temperature, soil organic carbon concentration, greenhouse gas emission flux, soil pH, electrical conductivity, bulk density, porosity, heavy metal concentration, pesticide residue content, nitrogen and phosphorus nutrient concentration, meteorological data, and agricultural activity data.

[0005] Preferably, the water, carbon, and phosphorus cycle processes in the farmland area are analyzed based on the real-time monitoring data to obtain the spatial distribution characteristics of soil pollutants, including: The real-time monitoring data is standardized to obtain standardized monitoring data. The standardized monitoring data is then input into a pre-constructed water-carbon-phosphorus cycle coupling model to simulate the dynamic processes of water movement, carbon conversion, and phosphorus migration within the farmland area, thereby obtaining simulation results of water movement trajectory, carbon conversion rate, and phosphorus migration path. The water-carbon-phosphorus cycle coupling model incorporates a meteorological driving module to correct model parameters based on precipitation intensity, wind speed, and sunshine duration. Based on the simulation results and the topographic data of the farmland area, the spatial distribution characteristics of the soil pollutants at different depths and horizontal directions were extracted.

[0006] Preferably, the water-carbon-phosphorus cycle coupling model includes a water movement module, a carbon transformation module, and a phosphorus migration module. The water movement module is used to simulate the infiltration, evaporation, and runoff processes of water in the soil based on soil moisture content, topographic data, and meteorological data. The carbon transformation module calculates the carbon mineralization rate and carbon emission intensity by combining soil organic carbon concentration, soil temperature data, and sunshine duration. The phosphorus migration module simulates the adsorption and desorption of phosphorus on the soil colloidal surface and its dynamic changes with water migration based on soil moisture content, initial phosphorus content, and soil pH. The water movement module, carbon transformation module, and phosphorus migration module achieve parameter interaction and process coupling through a data interface to collaboratively output the comprehensive simulation results of the water-carbon-phosphorus cycle. The core formula for water simulation in the water movement module is as follows: ; in, Soil moisture content ( t is time (seconds), Z is soil depth (m). Soil water diffusivity ( ), Where is the unsaturated hydraulic conductivity of the soil (m / s), P is the precipitation intensity (m / s), and E is the evaporation rate (m / s). Soil saturation water content ; The core formula for simulating carbon element conversion in the carbon element conversion module is as follows: (2) in, The carbon mineralization rate is expressed as milligrams of carbon per kilogram of soil per day. Baseline carbon mineralization rate (mg carbon / (kg soil·day)). For temperature sensitivity coefficient, Soil temperature (°C). The reference temperature is 20℃. Let be the function of moisture content influence. Let the influence function of sunshine duration be , The rate of organic matter decomposition (mg C / (kg soil·day)). Carbon fixation rate (mg carbon / (kg soil·day)); The core formula for simulating phosphorus migration in the phosphorus migration module is as follows: ; Where t is the time of phosphorus migration (hours (h)) and x is the migration time of phosphorus (meters (m)). Where is the porosity of the medium, v is the pore water flow velocity (m / h), S is the source-sink term, and C is the effective phosphorus concentration in the soil solution. Preferably, based on the simulation results and the topographic data of the farmland area, the spatial distribution characteristics of the soil pollutants at different depths and horizontal directions are extracted, including: The farmland area where the simulation results are located is divided into several grid cells of equal area; based on the elevation, slope and aspect of the topographic data, the convergence and diffusion trends of water, carbon and phosphorus elements in each grid cell are analyzed; Based on the associated relationships between pollutants and water, carbon, and phosphorus elements, and using spatial interpolation, pollutant concentration values ​​are assigned to each grid cell to generate concentration distribution maps of pollutants at different soil depths in the vertical direction and at different geographical locations in the horizontal direction. Based on the concentration distribution maps, the spatial distribution characteristics of the soil pollutants at different depths and in the horizontal direction are determined. The spatial interpolation method used is Kriging interpolation. The mathematical expression for the Kriging interpolation method is as follows: ; in, The estimated pollutant concentration at the interpolation point is [value]. Let i be the pollutant concentration value at the i-th known sampling point. Here, n represents the weight coefficient for the i-th sampling point, and n is the number of sampling points. The semivariogram values ​​between sampling point i and sampling point j Points to be interpolated The semivariogram value between sampling point j, The weighting coefficients are obtained by solving the system of equations, which are Lagrange multipliers. Then, the pollutant concentration at the interpolation point is calculated, and finally a continuous pollutant spatial distribution surface is generated.

[0007] Preferably, the high-risk pollution cluster area is scanned at specific points using a high-precision spectral sensor to obtain pollutant spectral characteristic data, including: Within the high-risk pollution cluster area, several scanning sampling points are set up according to a preset grid division rule. The high-precision spectral sensor is controlled to move sequentially to each sampling point, and the scanning angle and focal length of the high-precision spectral sensor are adjusted to ensure full coverage scanning of the soil surface and shallow soil at the sampling points. Reflectance spectral data of each sampling point within a preset wavelength range is collected, and the reflection spectral data is preprocessed to obtain standardized pollutant spectral characteristic data. The preprocessing includes dark current correction, spectral smoothing and denoising, and baseline drift elimination.

[0008] Preferably, the pollution diffusion results of the farmland area are obtained based on the pollutant type, pollution source location, and spatial distribution characteristics, including: The pollutant types, the diffusion direction of the pollution source, the topographic slope, soil physicochemical properties, and meteorological data in the spatial distribution characteristics are input into the pre-trained pollution diffusion prediction model. The pollution diffusion prediction model is used to simulate the diffusion range and concentration trends of pollutants in the longitudinal profile and transverse region of the soil at different time points to obtain the pollution diffusion results; wherein, the pollution diffusion prediction model is a three-dimensional diffusion model that considers soil stratification characteristics and pollutant degradation rate.

[0009] Preferably, the pollution diffusion prediction model is constructed based on an improved Gaussian diffusion model. This model, in addition to the traditional Gaussian diffusion model, incorporates soil porosity, moisture content, and clay content as correction parameters. By establishing a nonlinear mapping relationship between the pollutant diffusion coefficient and soil physicochemical properties, the model's adaptability to different soil types is improved. The pollution diffusion prediction model determines the initial diffusion direction vector based on the pollution source location and topographic slope, calculates the water-driven lateral diffusion rate using soil moisture content data, determines the vertical permeability coefficient based on soil porosity and pollutant molecular weight, and simulates the three-dimensional diffusion trajectory of pollutants in the soil using multi-parameter coupling. Finally, it outputs dynamic prediction results including the location of the diffusion front, changes in concentration gradient, and diffusion rate. The mathematical expression of the improved three-dimensional Gaussian diffusion model is as follows: ; in, At time t The location represents the pollutant concentration, M represents the total amount of pollutant released from the source, and k represents the pollutant degradation rate. The standard deviations of diffusion in the x, y, and z directions. Let x represent the diffusion rates in the x, y, and z directions. The coordinates of the pollution source are shown.

[0010] Preferably, the pollution diffusion results are compared with a preset safe diffusion threshold to obtain multi-level early warning instructions, including: The pollutant concentration, diffusion area, diffusion rate, pollutant accumulation, and distance to ecologically sensitive points in the pollution diffusion results are compared with preset safe diffusion thresholds in multiple dimensions. The safe diffusion thresholds include a first-level warning threshold, a second-level warning threshold, and a third-level warning threshold. When the pollution diffusion result is lower than the first-level warning threshold, a no-warning instruction is generated; When the pollution diffusion result exceeds the first-level warning threshold but does not reach the second-level warning threshold, a yellow warning instruction is generated; When the pollution diffusion result reaches the secondary warning threshold but does not exceed the tertiary warning threshold, an orange warning instruction is generated; When the pollution diffusion result exceeds the three-level warning threshold, a red warning instruction is generated. The first-level warning threshold includes a pollutant diffusion area of ​​1 hectare, a pollutant diffusion rate of 50 meters / day, a pollutant accumulation of 1.5 times the background value, and a distance of 500 meters from the ecologically sensitive point. The second-level warning threshold includes a pollutant diffusion area of ​​5 hectares, a pollutant diffusion rate of 200 meters / day, a pollutant accumulation of 2 times the background value, and a distance of 100 to 500 meters from the ecologically sensitive point. The third-level warning threshold includes a pollutant diffusion area of ​​20 hectares, a pollutant diffusion rate of 500 meters / day, a pollutant accumulation of 3 times the background value, and a distance of 100 meters from the ecologically sensitive point.

[0011] The technical solution adopted to solve the above-mentioned technical problems is: an Internet of Things (IoT)-based farmland pollutant monitoring system, which is applicable to the aforementioned IoT-based farmland pollutant monitoring method, comprising: The data analysis unit is used to collect soil environmental data based on IoT sensors deployed in farmland areas to obtain real-time monitoring data, and to analyze the water, carbon, and phosphorus cycle processes in the farmland areas based on the real-time monitoring data to obtain the spatial distribution characteristics of soil pollutants. The spectral scanning unit is used to determine the pollutant migration path based on the spatial distribution characteristics, determine the high-risk pollution accumulation area based on the pollutant migration path, and perform fixed-point scanning of the high-risk pollution accumulation area based on a high-precision spectral sensor to obtain pollutant spectral characteristic data. The pollution analysis unit is used to match and analyze the spectral characteristic data of the pollutants with the standard pollutant characteristics in the preset pollutant source tracing database to obtain the pollutant type and the location of the pollution source. The early warning generation unit is used to obtain the pollution diffusion results of the farmland area based on the pollutant type, pollution source location and spatial distribution characteristics, and compare the pollution diffusion results with the preset safe diffusion threshold to obtain multi-level early warning instructions; The pollution control unit is used to send the multi-level early warning instructions to the corresponding management personnel terminals, formulate differentiated pollution control plans based on the pollution diffusion results, and carry out pollution control on the farmland area based on the pollution control plans.

[0012] The beneficial effects of this invention are as follows: 1. This invention, by deploying physical sensors in farmland areas, can collect soil environmental data in real time and perform instant analysis. This real-time capability ensures that pollution problems are detected and dealt with promptly.

[0013] 2. This invention matches pollutant spectral characteristic data with a pollutant source tracing database to accurately identify the type and source of pollutants. This automated source tracing method can improve the accuracy and efficiency of pollutant identification and reduce errors from human judgment.

[0014] 3. This invention utilizes Internet of Things (IoT) technology, automated monitoring and analysis, and remote terminal control to make farmland pollution monitoring and management more intelligent, thereby improving work efficiency and accuracy. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the overall method steps in one embodiment of the present invention; Detailed Implementation

[0016] Example 1, as Figure 1 As shown, the present invention proposes an Internet of Things-based method for monitoring farmland pollutants, comprising: S1. Soil environmental data is collected based on IoT sensors deployed in farmland areas to obtain real-time monitoring data. Based on the real-time monitoring data, the water, carbon, and phosphorus cycle processes in farmland areas are analyzed to obtain the spatial distribution characteristics of soil pollutants. S2. Determine pollutant migration paths based on spatial distribution characteristics, identify high-risk pollution clusters based on pollutant migration paths, and perform point-to-point scanning of high-risk pollution clusters using high-precision spectral sensors to obtain pollutant spectral characteristic data. S3. Match the pollutant spectral characteristic data with the standard pollutant characteristics in the preset pollutant source tracing database to obtain the pollutant type and pollution source location; S4. Based on the types of pollutants, the location of pollution sources, and the spatial distribution characteristics, obtain the pollution diffusion results in farmland areas, compare the pollution diffusion results with the preset safe diffusion thresholds, and obtain multi-level early warning instructions; S5. Send multi-level early warning instructions to the corresponding management personnel terminals, formulate differentiated pollution control plans based on the pollution diffusion results, and carry out pollution control in farmland areas based on the pollution control plans.

[0017] In this invention, the water-carbon-phosphorus cycle refers to the cycling of water, carbon, and phosphorus in farmland soil. These elements are nutrients required for plant growth, and the cycling process affects soil fertility and health. Analyzing the cycling process of these elements helps to determine the ecological state of the soil. The spatial distribution characteristics of soil pollutants refer to the spatial distribution of different pollutants (such as heavy metals, chemical pesticides, and fertilizers) in the soil, that is, their distribution patterns and concentration differences within the farmland area. The pollutant migration path refers to the migration path of pollutants in the soil or between the soil and water. The pollutant migration path helps to predict how pollution spreads and which areas are at higher risk of pollution. High-risk pollution accumulation areas refer to areas that, based on the pollutant migration path, may become high-risk areas for pollutant accumulation due to topography, climate, or other factors. Identifying these areas helps to carry out more targeted pollution control. The high-precision spectral sensor refers to the analysis of the composition and concentration of pollutants in the soil by measuring the reflection of light. The spectral sensor can acquire spectral characteristic data of pollutants without direct contact, which is very helpful for pollution source identification and pollutant tracking. Pollutant spectral characteristic data refers to light reflectance data at specific wavelengths acquired through high-precision spectral sensors. This data reflects the types and concentrations of pollutants in the soil. The pollutant source database is a database storing pollutant spectral characteristic data. This data is created based on the spectral characteristics and source information of known pollutants. By matching and analyzing the spectral characteristic data, the types of pollutants and possible pollution source locations can be determined. Pollution source location refers to the location of the pollutant's origin. By analyzing the spectral characteristics of the pollutants and the information stored in the database, the spatial location of the pollution source can be determined. Pollution diffusion results refer to the analysis results based on pollutant types, pollution source locations, and spatial distribution characteristics, predicting how pollutants will diffuse in farmland areas. The safe diffusion threshold is used to assess the safe range of pollution diffusion. When pollution diffusion exceeds this threshold, it may pose a threat to the environment and human health, requiring corresponding countermeasures. Multi-level early warning instructions refer to the system generating multi-level early warning instructions based on the comparison between the pollution diffusion results and the safe diffusion threshold. These instructions are graded according to the severity of pollution, reminding managers to take different levels of emergency measures.

[0018] Example 2: The present invention proposes an IoT-based method for monitoring farmland pollutants. Compared with Example 1, this example further includes: A1. Analyze the water, carbon, and phosphorus cycle processes in farmland areas based on real-time monitoring data to obtain the spatial distribution characteristics of soil pollutants, including: A2. Standardize the real-time monitoring data to obtain standardized monitoring data; A3. Standardized monitoring data are input into a pre-constructed water-carbon-phosphorus cycle coupling model to simulate the dynamic processes of water movement, carbon conversion, and phosphorus migration in farmland areas, so as to obtain simulation results of water movement trajectory, carbon conversion rate, and phosphorus migration path; among them, the water-carbon-phosphorus cycle coupling model introduces a meteorological driving module to correct model parameters based on precipitation intensity, wind speed, and sunshine duration. A4. Based on simulation results and topographic data of farmland areas, extract the spatial distribution characteristics of soil pollutants at different depths and horizontal directions.

[0019] In this embodiment, soil moisture content refers to the proportion of water in the soil, usually expressed as the percentage of water mass to total soil mass; water movement trajectory refers to the path of water movement in farmland soil, which is usually affected by factors such as precipitation, irrigation, evaporation, and soil permeability. Through simulation, the flow and distribution of water can be predicted; carbon conversion rate refers to the speed at which carbon is converted from one form to another in the soil, for example, the rate at which organic carbon is converted into carbon dioxide. The carbon conversion rate is affected by soil temperature, humidity, and organic carbon concentration; phosphorus migration path refers to the migration path of phosphorus in the soil, usually referring to the diffusion or dissolution process of phosphate. Phosphorus migration is affected by factors such as soil moisture, soil pH, and soil organic matter.

[0020] In an optional embodiment, the water-carbon-phosphorus cycle coupled model includes a water movement module, a carbon transformation module, and a phosphorus migration module. The water movement module simulates the infiltration, evaporation, and runoff processes of water in the soil based on soil moisture content and topographic data. The carbon transformation module calculates the carbon mineralization rate and carbon emission intensity by combining soil organic carbon concentration and soil temperature data. The phosphorus migration module simulates the adsorption and desorption of phosphorus on soil colloid surfaces and its dynamic changes with water migration based on soil moisture content and initial phosphorus content. The three modules interact and couple their parameters through a data interface to collaboratively output comprehensive simulation results of the water-carbon-phosphorus cycle. The core formula for water simulation in the water movement module is as follows: ; in, Soil moisture content t is time (seconds), z is soil depth (meters). Soil water diffusivity , Unsaturated hydraulic conductivity of soil P is the precipitation intensity (m / s), and E is the evaporation rate (m / s). Soil saturation water content ; The core formula for simulating carbon element conversion in the carbon element conversion module is as follows: ; in, The carbon mineralization rate is expressed as milligrams of carbon per kilogram of soil per day. The baseline carbon mineralization rate is (mg carbon / (kg soil·day)). T is the temperature sensitivity coefficient, where T is the soil temperature (°C). The reference temperature is 20℃. Let be the function of moisture content influence. Let the influence function of sunshine duration be , The rate of organic matter decomposition (mg C / (kg soil·day)). Carbon fixation rate (mg carbon / (kg soil·day)); The core formula for simulating phosphorus migration in the phosphorus migration module is as follows: ; Where t is the time of phosphorus migration (hours (h)) and x is the migration time of phosphorus (meters (m)). denoted as medium porosity, v as pore water flow velocity (m / h), S as source-sink term, and C as effective phosphorus concentration in soil solution.

[0021] It should be noted that the comprehensive simulation results refer to the overall phenomena obtained through the interaction and coupling calculation of the three modules. These results can help us understand how water, carbon, and phosphorus interact in farmland, as well as their distribution and changes in the soil. The comprehensive simulation results provide an overall view of the water, carbon, and phosphorus cycles, and can provide a scientific basis for agricultural management, soil pollution control, and other related fields.

[0022] In an optional embodiment, the spatial distribution characteristics of soil pollutants at different depths and horizontal directions are extracted based on simulation results and topographic data of the farmland area, including: B1. Divide the farmland area where the simulation results are located into several grid cells of equal area; B2. Based on the elevation, slope, and aspect data in the topographic data, analyze the convergence and diffusion trends of water, carbon, and phosphorus in each grid cell; B3. Based on the co-occurrence relationship between pollutants and water, carbon and phosphorus elements, and spatial interpolation, pollutant concentration values ​​are assigned to each grid cell to generate concentration distribution maps of pollutants at different soil depths in the vertical direction and at different geographical locations in the horizontal direction. Based on the concentration distribution maps, the spatial distribution characteristics of soil pollutants at different depths and in the horizontal direction are determined. Among them, the spatial interpolation method adopts Kriging interpolation. The mathematical expression for Kriging interpolation is as follows: ; in, The estimated pollutant concentration at the interpolation point is [value]. Let i be the pollutant concentration value at the i-th known sampling point. Here, n represents the weight coefficient for the i-th sampling point, and n is the number of sampling points. The semivariogram values ​​between sampling point i and sampling point j Points to be interpolated The semivariogram value between sampling point j, The weighting coefficients are obtained by solving the system of equations, which are Lagrange multipliers. Then, the pollutant concentration at the interpolation point is calculated, and finally a continuous pollutant spatial distribution surface is generated.

[0023] It should be noted that the co-occurrence relationship between pollutants and elements such as water, carbon, and phosphorus refers to the spatial and temporal interdependence between pollutants and elements such as water, carbon, and phosphorus. For example, pollutants may flow in the soil together with water, or react chemically with organic carbon or phosphorus in the soil, leading to the migration or transformation of pollutants. Spatial interpolation is a method of predicting data values ​​of unknown points (such as the location of other grid cells) based on known data points (such as the concentration of pollutants at certain specific locations).

[0024] In an optional embodiment, a high-risk pollution accumulation area is scanned at specific points using a high-precision spectral sensor to obtain pollutant spectral characteristic data, including: C1. In high-risk pollution cluster areas, set up several scanning sampling points according to the preset grid division rules, control the high-precision spectral sensor to move to each sampling point in sequence, and adjust the scanning angle and focal length of the high-precision spectral sensor to ensure full coverage scanning of the soil surface and shallow soil at the sampling points. C2. Collect reflectance spectral data of each sampling point within a preset wavelength range, and preprocess the reflectance spectral data to obtain standardized pollutant spectral characteristic data. The preprocessing includes dark current correction, spectral smoothing and denoising, and baseline drift elimination.

[0025] It should be noted that the preset wavelength range refers to the specific wavelength range set by the sensor during spectral scanning. Different pollutants and soil components will exhibit different reflection characteristics at specific wavelengths. The preset wavelength range usually covers the absorption or reflection characteristics of these pollutants in order to accurately detect them. Dark current correction refers to the fact that spectral sensors may generate a certain current even without light. This current is called dark current, which can affect the accuracy of spectral data. Therefore, correction is needed to eliminate this unnecessary interference and ensure the reliability of the measurement results. Baseline drift elimination refers to the possibility that the baseline (i.e., the zero point of the reflectance spectrum) in the spectral data may drift due to changes in equipment or environment. Baseline drift can lead to inaccurate measurements. Therefore, data processing methods are needed to eliminate this drift, making the spectral data more stable and reliable.

[0026] In an optional embodiment, the pollution diffusion results of farmland areas are obtained based on pollutant type, pollution source location, and spatial distribution characteristics, including: D1. Input the type of pollutant, the direction of diffusion of the pollution source, and the topographic slope in the spatial distribution characteristics into the pre-trained pollution diffusion prediction model. D2. Based on the pollution diffusion prediction model, simulate the diffusion range and concentration change trend of pollutants in the longitudinal profile and transverse region of soil at different time points to obtain the pollution diffusion results.

[0027] In an optional embodiment, the pollution diffusion prediction model is built upon an improved Gaussian diffusion model. This model incorporates soil porosity, moisture content, and clay content as correction parameters, building upon the traditional Gaussian diffusion model. By establishing a nonlinear mapping relationship between the pollutant diffusion coefficient and soil physicochemical properties, the model's adaptability to different soil types is enhanced. The pollution diffusion prediction model determines the initial diffusion direction vector based on the pollution source location and topographic slope, calculates the water-driven lateral diffusion rate using soil moisture content data, determines the vertical permeability coefficient based on soil porosity and pollutant molecular weight, and simulates the three-dimensional diffusion trajectory of pollutants in the soil using multi-parameter coupling. Finally, it outputs dynamic prediction results including the location of the diffusion front, changes in concentration gradient, and diffusion rate. The mathematical expression for the improved three-dimensional Gaussian diffusion model is as follows: ; in, At time t The location represents the pollutant concentration, M represents the total amount of pollutant released from the source, and k represents the pollutant degradation rate. The standard deviations of diffusion in the x, y, and z directions. Let x represent the diffusion rates in the x, y, and z directions. The coordinates of the pollution source are shown.

[0028] It should be noted that the Gaussian diffusion model is a classic mathematical model used to describe the diffusion process of pollutants in air or water. Based on the Gaussian distribution assumption, this model assumes that the concentration of pollutants exhibits a symmetrical bell-shaped distribution with increasing distance diffusion. It assumes that pollutants diffuse from a point source to the surrounding environment, and the concentration gradually decreases with increasing distance. In soil pollution diffusion prediction, this model is usually used as the basis to estimate the diffusion path and concentration of pollutants. The diffusion coefficient is a parameter used to describe the rate and intensity of pollutant diffusion in a medium. The larger the diffusion coefficient, the faster the pollutant diffuses. There is a non-linear relationship between the pollutant diffusion coefficient and the physical and chemical properties of the soil (such as porosity and water content). The model uses this relationship to more accurately predict the diffusion behavior of pollutants in the soil. The lateral diffusion rate refers to the diffusion speed of pollutants along the horizontal plane driven by water. This rate is determined by factors such as soil water content. Water helps pollutants spread laterally along the soil surface. The longitudinal permeability coefficient refers to the infiltration rate of pollutants in the vertical direction (i.e., the direction of soil layers).

[0029] In an optional embodiment, the pollution diffusion result is compared with a preset safe diffusion threshold to obtain multi-level early warning instructions, including: E1. Compare the pollutant concentration, diffusion area, and diffusion rate in the pollution diffusion results with the preset safe diffusion threshold in multiple dimensions. The safe diffusion threshold includes a first-level warning threshold, a second-level warning threshold, and a third-level warning threshold. E2. When the pollution diffusion result is lower than the first-level warning threshold, a no-warning instruction is generated; E3. When the pollution diffusion result exceeds the Level 1 warning threshold but does not reach the Level 2 warning threshold, a yellow warning instruction is generated; E4. When the pollution diffusion result reaches the level 2 warning threshold but does not exceed the level 3 warning threshold, an orange warning instruction is generated; E5. When the pollution diffusion results exceed the Level 3 warning threshold, a red warning order will be generated. The Level 1 warning threshold includes a pollutant diffusion area of ​​1 hectare, a pollutant diffusion rate of 50 meters / day, a pollutant accumulation of 1.5 times the background value, and an ecologically sensitive point distance of 500 meters. The Level 2 warning threshold includes a pollutant diffusion area of ​​5 hectares, a pollutant diffusion rate of 200 meters / day, a pollutant accumulation of 2 times the background value, and an ecologically sensitive point distance of 100 to 500 meters. The Level 3 warning threshold includes a pollutant diffusion area of ​​20 hectares, a pollutant diffusion rate of 500 meters / day, a pollutant accumulation of 3 times the background value, and an ecologically sensitive point distance of 100 meters.

[0030] Example 3: The present invention proposes an IoT-based farmland pollutant monitoring system, applicable to an IoT-based farmland pollutant monitoring method, comprising: Data analysis unit 1 is used to collect soil environmental data based on IoT sensors deployed in farmland areas to obtain real-time monitoring data, and to analyze the water, carbon and phosphorus cycle process in farmland areas based on the real-time monitoring data to obtain the spatial distribution characteristics of soil pollutants. The spectral scanning unit 2 is used to determine the pollutant migration path based on spatial distribution characteristics, identify high-risk pollution accumulation areas based on the pollutant migration path, and perform fixed-point scanning of the high-risk pollution accumulation areas based on a high-precision spectral sensor to obtain pollutant spectral characteristic data. Pollution analysis unit 3 is used to match and analyze the spectral characteristics of pollutants with the standard pollutant characteristics in the preset pollutant source tracing database to obtain the types of pollutants and the location of pollution sources. The early warning generation unit 4 is used to obtain the pollution diffusion results of farmland areas based on the type of pollutants, the location of pollution sources and spatial distribution characteristics, and compare the pollution diffusion results with the preset safe diffusion threshold to obtain multi-level early warning instructions. Pollution control unit 5 is used to send multi-level early warning instructions to the corresponding management personnel terminals, formulate differentiated pollution control plans based on the pollution diffusion results, and carry out pollution control in farmland areas based on the pollution control plans.

[0031] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. An Internet of Things based method for monitoring of agricultural field pollutants, characterized in that, The method comprises the following steps: Collecting soil environment data based on Internet of Things sensors deployed in farmland areas to obtain real-time monitoring data, analyzing the water-carbon-phosphorus cycle process of the farmland areas based on the real-time monitoring data to obtain the spatial distribution characteristics of soil pollutants; Determine the pollutant migration path based on the spatial distribution characteristics, determine the high-risk pollution accumulation area based on the pollutant migration path, and perform fixed-point scanning on the high-risk pollution accumulation area based on a high-precision spectral sensor to obtain pollutant spectral characteristic data; Match and analyze the pollutant spectral characteristic data and the standard pollutant characteristics in the preset pollutant tracing database to obtain the pollutant species and the pollution source position; Obtain the pollution diffusion result of the farmland area based on the pollutant species, the pollution source position and the spatial distribution characteristics, compare the pollution diffusion result with the preset safe diffusion threshold to obtain a multi-level early warning instruction; Send the multi-level early warning instruction to the corresponding management personnel terminal, develop a differentiated pollution control scheme based on the pollution diffusion result, and control the pollution of the farmland area based on the pollution control scheme. 2.The farmland pollutant monitoring method based on the Internet of Things according to claim 1, characterized in that, The monitoring data includes soil moisture content, soil temperature, soil organic carbon concentration, greenhouse gas emission flux, soil pH value, electrical conductivity, bulk density, porosity, heavy metal concentration, pesticide residue content, nitrogen and phosphorus nutrient salt concentration, meteorological data and agricultural activity data. 3.The farmland pollutant monitoring method based on the Internet of Things according to claim 2, characterized in that, The method comprises the following steps: Standardize the real-time monitoring data to obtain standardized monitoring data, input the standardized monitoring data into a pre-constructed water-carbon-phosphorus cycle coupling model, simulate the dynamic process of water movement, carbon element transformation and phosphorus element migration in the farmland area to obtain simulation results of water movement trajectory, carbon element transformation rate and phosphorus element migration path, wherein the water-carbon-phosphorus cycle coupling model introduces a meteorological driving module to correct model parameters based on precipitation intensity, wind speed and sunshine duration; Extract the spatial distribution characteristics of the soil pollutants at different depths and horizontal directions based on the simulation results and the topographic and geomorphic data of the farmland area. 4.The farmland pollutant monitoring method based on the Internet of Things according to claim 3, characterized in that, The water-carbon-phosphorus cycle coupling model comprises a water movement module, a carbon element transformation module and a phosphorus element migration module, wherein the water movement module is used to simulate the infiltration, evaporation and runoff process of water in the soil according to soil moisture content, topographic data and meteorological data, the carbon element transformation module calculates the carbon mineralization rate and carbon emission intensity in combination with soil organic carbon concentration, soil temperature data and sunshine duration, and the phosphorus element migration module simulates the dynamic changes of adsorption and desorption of phosphorus on the surface of soil colloids and migration with water based on soil moisture content, initial phosphorus content and soil pH value, and the water movement module, the carbon element transformation module and the phosphorus element migration module realize parameter interaction and process coupling through a data interface to cooperatively output the comprehensive simulation results of the water-carbon-phosphorus cycle; The water simulation core formula in the water movement module is as follows: in, Soil moisture content ( ), Time (seconds). Soil depth (m), Soil water diffusivity ( ), The unsaturated hydraulic conductivity of the soil (m / s) Precipitation intensity (m / s), Evaporation rate (m / s) Soil saturation water content ( ) The simulation carbon element conversion core formula in the carbon element conversion module is as follows: （2） wherein, is the carbon mineralization rate (mg carbon / (kg soil・day)), is the baseline carbon mineralization rate (mg carbon / (kg soil・day)), is the temperature sensitivity coefficient, is the soil temperature (°C), is the baseline temperature (20 °C), is the water content influence function, is the sunlight duration influence function, is the organic matter decomposition rate (mg carbon / (kg soil・day)), is the carbon sequestration rate (mg carbon / (kg soil・day)). The simulation phosphorus element migration core formula in the phosphorus element migration module is as follows: where t is the time of phosphorus transport (hours (h)), is the phosphorus transport (meters (m)), is the medium porosity, v is the pore water velocity (meters / hour), S is the source-sink term, and C is the effective concentration of phosphorus in the soil solution. 5.The farmland pollutant monitoring method based on the Internet of Things according to claim 4, characterized in that, Based on the simulation results and the topographic data of the farmland area, the spatial distribution characteristics of the soil pollutants in different depths and horizontal directions are extracted, including: The farmland area where the simulation results are located is divided into several grid cells with equal area; based on the elevation, slope and aspect in the topographic data, the convergence and diffusion trends of water, carbon elements and phosphorus elements in each grid cell are analyzed; Based on the concomitant relationship of pollutants and water, carbon and phosphorus elements and the spatial interpolation method, the pollutant concentration values are distributed to each grid cell to generate a concentration distribution map of pollutants in different soil depths in the vertical direction and different geographical positions in the horizontal direction, and based on the concentration distribution map, the spatial distribution characteristics of the soil pollutants in different depths and horizontal directions are determined, wherein the spatial interpolation method adopts the Kriging interpolation method; The mathematical expression of the Kriging interpolation method is as follows: wherein, is the pollutant concentration estimate value of the point to be interpolated, is the pollutant concentration value of the ith known sampling point, is the weight coefficient of the ith sampling point, and n is the number of sampling points, is the semi-variogram value between the sampling point i and the sampling point j, is the semi-variogram value between the point to be interpolated and the sampling point j, is the Lagrange multiplier, and the weight coefficient is obtained by solving the equation set, and then the pollutant concentration of the point to be interpolated is calculated, and finally the continuous pollutant spatial distribution surface is generated. 6.The farmland pollutant monitoring method based on the Internet of Things according to claim 5, characterized in that, Based on the high-precision spectral sensor, the high-risk pollution aggregation area is scanned at fixed points to obtain pollutant spectral feature data, including: According to the preset grid division rule, a plurality of scanning sampling points are arranged in the high-risk pollution aggregation area, the high-precision spectral sensor is controlled to move to each sampling point in turn, and the scanning angle and focal length of the high-precision spectral sensor are adjusted to ensure full coverage scanning of the soil surface and shallow soil at the sampling point; The reflectance spectrum data of each sampling point in the preset wavelength range is collected, and the reflectance spectrum data is preprocessed to obtain standardized pollutant spectral feature data, wherein the preprocessing includes dark current correction, spectral smoothing and denoising, and baseline drift elimination. 7.The farmland pollutant monitoring method based on the Internet of Things according to claim 6, characterized in that, Based on the pollutant type, pollution source direction and spatial distribution characteristics, the pollution diffusion result of the farmland area is obtained, including: The pollutant type, pollution source direction, spatial distribution characteristics, topographic slope, soil physical and chemical properties and meteorological data are input into a pre-trained pollution diffusion prediction model; Based on the pollution diffusion prediction model, the diffusion range and concentration change trend of pollutants in the soil vertical profile and horizontal area at different time nodes are simulated to obtain the pollution diffusion result; wherein the pollution diffusion prediction model is a three-dimensional diffusion model considering the soil layering characteristics and the degradation rate of pollutants. 8.The farmland pollutant monitoring method based on the Internet of Things according to claim 7, characterized in that, The pollution diffusion prediction model is constructed based on an improved Gaussian diffusion model, modified parameters of the pollution diffusion prediction model include soil porosity, water content, clay content, pollutant solubility, adsorption coefficient, degradation rate and meteorological data, the pollution diffusion prediction model is used to establish a nonlinear mapping relationship between a pollutant diffusion coefficient and soil physicochemical properties, the pollution diffusion prediction model determines an initial diffusion direction vector according to a pollution source orientation and a terrain slope, calculates a horizontal diffusion rate under water driving in combination with soil water content data, determines a vertical permeation coefficient based on soil porosity and pollutant molecular weight, simulates a three-dimensional diffusion trajectory of the pollutant in the soil based on multi-parameter coupling, and finally outputs dynamic prediction results including diffusion front position, concentration gradient change and diffusion rate; A mathematical expression of the improved three-dimensional Gaussian diffusion model is specifically as follows: wherein, is the instant ) the concentration of the pollutant at the location, is the total amount released by the pollution source, is the degradation rate of the pollutant, is the standard deviation of the diffusion in the x, y, z directions, is the diffusion rate in the x, y, z directions, is the coordinate of the pollution source. 9.The farmland pollutant monitoring method based on the Internet of Things according to claim 8, characterized in that, The pollution diffusion results are compared with preset safe diffusion threshold values to obtain multi-level early warning instructions, including: The pollutant concentration value, diffusion area, diffusion rate, pollutant cumulative amount and ecological sensitive point distance in the pollution diffusion results are compared with preset safe diffusion threshold values in multiple dimensions, wherein the safe diffusion threshold values include a first-level early warning threshold value, a second-level early warning threshold value and a third-level early warning threshold value; When the pollution diffusion results are lower than the first-level early warning threshold value, no early warning instruction is generated; When the pollution diffusion results exceed the first-level early warning threshold value but do not reach the second-level early warning threshold value, a yellow early warning instruction is generated; When the pollution diffusion results reach the second-level early warning threshold value but do not exceed the third-level early warning threshold value, an orange early warning instruction is generated; When the pollution diffusion results exceed the third-level early warning threshold value, a red early warning instruction is generated, wherein the first-level early warning threshold value includes a pollutant diffusion area of 1 hectare, a pollutant diffusion rate of 50 meters per day, a pollutant cumulative amount of 1.5 times a background value, and an ecological sensitive point distance of 500 meters, the second-level early warning threshold value includes a pollutant diffusion area of 5 hectares, a pollutant diffusion rate of 200 meters per day, a pollutant cumulative amount of 2 times a background value, and an ecological sensitive point distance of 100 meters to 500 meters, and the third-level early warning threshold value includes a pollutant diffusion area of 20 hectares, a pollutant diffusion rate of 500 meters per day, a pollutant cumulative amount of 3 times a background value, and an ecological sensitive point distance of 100 meters.

10. An Internet of Things based farmland pollutant monitoring system, which is suitable for an Internet of Things based farmland pollutant monitoring method as claimed in any one of claims 1-9, characterized in that, It includes: A data analysis unit (1) is configured to collect soil environment data based on Internet of Things sensors deployed in a farmland area to obtain real-time monitoring data, and analyze a water-carbon-phosphorus cycle process of the farmland area based on the real-time monitoring data to obtain spatial distribution characteristics of soil pollutants. A spectrum scanning unit (2) is configured to determine a pollutant migration path based on the spatial distribution characteristics, determine a high-risk pollution aggregation area based on the pollutant migration path, and perform fixed-point scanning on the high-risk pollution aggregation area based on a high-precision spectrum sensor to obtain pollutant spectrum characteristic data. A pollution analysis unit (3) is configured to perform matching analysis on the pollutant spectrum characteristic data and standard pollutant characteristics in a preset pollutant traceability database to obtain a pollutant species and a pollution source orientation. The early warning generation unit (4) is configured to acquire a pollution diffusion result of the farmland area based on the pollutant type, the pollution source position, and the spatial distribution characteristics, compare the pollution diffusion result with a preset safe diffusion threshold, and obtain a multi-level early warning instruction; The pollution control unit (5) is configured to send the multi-level early warning instruction to a corresponding management personnel terminal, formulate a differentiated pollution control scheme based on the pollution diffusion result, and control pollution of the farmland area based on the pollution control scheme.

Images