A real-time intelligent soil pollution monitoring system and method

Abstract

The application discloses a kind of real-time intelligent soil pollution monitoring system and method, it is related to soil pollution monitoring field, the system can be through soil pollutant sensor real-time acquisition soil pollutant index data, and through data fusion, pollution modeling, pollution source tracing, pollution assessment and pollution early warning etc. Module, to soil environmental quality is comprehensively monitored, analyzed and evaluated, and the future pollution situation is predicted and early warned, to realize the effective protection and management of soil environment. The application has real-time, intelligence, accuracy and reliability and the like advantages, can provide a new technical scheme for soil environmental monitoring.

Description

Technical Field

[0001] This invention relates to the field of soil pollution monitoring, and specifically to a real-time intelligent soil pollution monitoring system and method. Background Technology

[0002] Soil is a vital resource for human survival and development, and its quality directly impacts agricultural production, the ecological environment, and human health. However, with the acceleration of industrialization, urbanization, and agricultural development, soil pollution has become increasingly serious, leading to the degradation of soil functions and increased ecological risks. To effectively prevent and control soil pollution, a comprehensive soil pollution monitoring system is needed to achieve real-time monitoring, intelligent analysis, and rapid disposal of soil pollutant indicators.

[0003] Currently, my country has established soil pollution monitoring systems of a certain scale in some areas, but these systems have some problems and shortcomings, mainly in the following aspects:

[0004] (1) The data fusion capability of the soil pollution monitoring system is not strong enough, and it is unable to effectively integrate soil pollutant index data from different sources, types, and times. At present, my country's soil pollution monitoring system mainly relies on single or a small number of monitoring points for data collection. This approach has disadvantages such as data silos, information gaps, and low utilization rates. Moreover, due to the lack of advanced data integration and preprocessing technologies, the collected data is difficult to effectively integrate and analyze, resulting in the failure to fully realize the value of the data.

[0005] (2) The modeling capabilities of soil pollution monitoring systems are not precise enough, making it impossible to accurately describe the relationship between soil pollutant index data and location, time, and environmental factors. Currently, my country's soil pollution monitoring systems mainly rely on traditional statistical analysis methods for data processing. This approach has drawbacks such as simple models, low accuracy, and poor applicability. Moreover, due to the lack of suitable regression analysis techniques and mathematical model support, current soil pollution monitoring systems struggle to accurately model soil pollutant index data at known locations, let alone fit and predict soil pollutant index data at unknown locations or future time points.

[0006] (3) The source tracing capability of the soil pollution monitoring system is not effective enough, and it is unable to accurately determine the possible source location and type of soil pollutant index data that exceeds the standard or changes abnormally. At present, my country's soil pollution monitoring system mainly relies on manual or simple software for data analysis, which has the disadvantages of low efficiency, low accuracy and poor timeliness. Moreover, due to the lack of effective mathematical models and source tracing technology support, the current soil pollution monitoring system is unable to accurately determine the location and type of pollution source, and calculate the spatial and temporal diffusion range and intensity of pollutants.

[0007] In conclusion, my country's current soil pollution monitoring system still has many problems and shortcomings, and urgently needs to be improved and innovated. Summary of the Invention

[0008] To address the above problems, this invention provides a real-time intelligent soil pollution monitoring system to solve the technical problem of the lack of real-time, intelligent, and comprehensive monitoring and analysis of soil pollution status in existing technologies. The system includes the following modules:

[0009] The data acquisition module is used to collect soil pollutant index data in real time through soil pollutant sensors, including organic matter concentration, heavy metal content, pH, temperature and humidity, and transmit them to the data center via wireless network;

[0010] Data centers are used to receive and store data;

[0011] The data fusion module is used to obtain the stored soil pollutant index data from the data center, and according to the source, type and time of the data, it uses the data integration method to preprocess the soil pollutant index data from different sources, types and times, and merge them into a soil pollutant index dataset, and then send the soil pollutant index dataset to the data center.

[0012] The pollution modeling module is used to retrieve the stored soil pollutant index dataset from the data center, and, based on the data fusion method, to model the soil pollutant index data at each known location using regression analysis, describing the relationship between the soil pollutant index data and location, time, and environmental factors, and to fit and predict the soil pollutant index data at unknown locations or future time points, and send the mathematical model and the fitting and prediction results to the data center.

[0013] The pollution source tracing module is used to determine the possible source location and type of soil pollutant index data exceeding the standard or abnormal changes by using mathematical models obtained from the data center, and to calculate the spatial and temporal diffusion range and intensity of pollutants, and send the determination results and calculation results to the data center.

[0014] The system employs a technical solution based on wireless sensor networks, data fusion, regression analysis, and source tracing analysis, which can effectively collect, process, model, and evaluate soil pollutant index data, and promptly detect and track pollution sources.

[0015] The system has produced the following beneficial effects: improved the accuracy, real-time performance, and intelligence of soil environmental quality monitoring; enhanced the ability to predict and warn of changes in soil environmental quality trends and risk factors; and promoted the timely detection and effective management of soil environmental quality problems.

[0016] Under optimal implementation conditions, this system addresses the technical problem of the lack of existing technologies for assessing and providing early warnings of soil environmental quality.

[0017] In a preferred implementation, the system also includes:

[0018] The pollution assessment module is used to assess the soil pollutant index data of all locations based on the fitting and prediction results of soil pollutant index data of unknown locations obtained from the data center, soil pollutant index data of known locations, and national or regional soil environmental quality standards, and to provide soil environmental quality level and evaluation report to reflect the actual status and changing trend of the soil environment, and send the soil environmental quality level and evaluation report to the data center.

[0019] The pollution early warning module is used to provide early warnings of exceeding the standards for soil pollutant index data at future time points based on the fitting and prediction results of soil pollutant index data for future time points obtained from the data center, as well as the preset soil pollutant index data thresholds or early warning values. When the predicted value exceeds the soil pollutant index data threshold or early warning value, the module sends early warning information to relevant departments or individuals so that countermeasures can be taken in advance, and sends the early warning results to the data center.

[0020] In optimal implementation, the system employs a technical solution that uses fitting and prediction results obtained from national or regional soil environmental quality standards, thresholds or warning values, and regression analysis methods to assess and warn of soil environmental quality at all locations and future time points, and provides corresponding reports and information.

[0021] Under optimal implementation conditions, the system has produced the following beneficial effects: improved objective understanding and scientific judgment of soil environmental quality status and risk level; enhanced the ability to proactively detect and promptly address soil environmental quality problems; and promoted the effective implementation of soil environmental quality protection and improvement work.

[0022] Under optimal implementation conditions, this system addresses the technical problem of the lack of effective integration and utilization of soil pollutant index data in existing technologies.

[0023] In a preferred implementation scenario, the data fusion module specifically includes:

[0024] The data quality assessment submodule is used to assess the quality of the acquired soil pollutant index data, including completeness, accuracy, consistency and timeliness. Based on the assessment results, the data is filtered, corrected or deleted, and the data is output to the data alignment submodule.

[0025] The data alignment submodule is used to align soil pollutant index data from different sources, types, and times, including spatial alignment, temporal alignment, and attribute alignment, to eliminate differences and conflicts between data.

[0026] The data fusion submodule is used to fuse the aligned soil pollutant index data, generate one or more soil pollutant index datasets based on the fusion results, and send them to the data center.

[0027] In a preferred implementation scenario, the system employs a technical approach based on data quality assessment, data alignment, and data fusion, which can improve the quality, consistency, and usability of soil pollutant index data and generate one or more soil pollutant index datasets suitable for subsequent analysis and modeling.

[0028] Under optimal implementation conditions, the system has produced the following beneficial effects: improved the reliability and efficiency of soil pollutant index data; enhanced the expressive power and information content of soil pollutant index data; and promoted the sharing and interaction of soil pollutant index data.

[0029] Under optimal implementation conditions, this system addresses the technical problem of the lack of effective modeling and prediction of soil pollutant index data in existing technologies.

[0030] In a preferred implementation scenario, the pollution modeling module specifically includes:

[0031] The feature extraction submodule is used to extract feature variables from the soil pollutant index dataset that help describe and predict the soil pollutant index data, including location, time, and environmental factors.

[0032] The model selection submodule is used to select the regression analysis method based on the number, type, and distribution of the feature variables, including linear regression, multiple regression, and nonlinear regression.

[0033] The model training submodule is used to establish a mathematical model based on the selected regression analysis method, using known soil pollutant index data and characteristic variables, and to adjust the model parameters through optimization algorithms before sending the optimized mathematical model to the data center.

[0034] The model evaluation submodule is used to evaluate the model's fit and predictive ability based on the established mathematical model, using unknown soil pollutant index data and characteristic variables, and to test the model's stability and reliability through error analysis and sensitivity analysis.

[0035] The model application submodule is used to fit and predict soil pollutant index data at unknown locations or future time points based on the established mathematical model, and send the fitting and prediction results to the data center.

[0036] In optimal implementation, the system employs a technical solution based on feature extraction, model selection, model training, model evaluation, and model application. It can utilize known soil pollutant index data and characteristic variables to establish a mathematical model describing the relationship between soil pollutant index data and location, time, and environmental factors, and to fit and predict soil pollutant index data at unknown locations or future time points.

[0037] Under optimal implementation conditions, the system has produced the following beneficial effects: improved the accuracy, sensitivity, and intelligence of modeling and predicting soil pollutant index data; enhanced the ability to dynamically monitor and analyze trends of soil pollutant index data; and promoted the effective implementation of control and optimization work for soil pollutant index data.

[0038] Under optimal implementation conditions, this system addresses the technical problem of the lack of effective identification and location of soil pollution sources in existing technologies.

[0039] In a preferred implementation scenario, the pollution source tracing module uses the following modules to determine the possible source location and type that leads to excessive or abnormal changes in soil pollutant index data:

[0040] The residual calculation submodule is used to calculate the residual of the soil pollutant index data at each location according to the mathematical model, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module.

[0041] The anomaly detection submodule is used to detect outliers in the residuals and filter out residuals that exceed a preset residual threshold as outliers.

[0042] The clustering analysis submodule is used to perform clustering analysis on outliers, grouping adjacent or close outliers into one class as an outlier region.

[0043] The pollution source type determination submodule is used to determine the possible pollution source type for each abnormal area based on the sign, size, and distribution of the residual, including point source, area source, and line source, and send it to the data center.

[0044] The pollution source location estimation module is used to estimate the possible pollution source locations for each anomalous area based on the direction and gradient of the residuals, including the center point, edge point, and intermediate point, and then send the results to the data center.

[0045] In optimal implementation, the system employs a technical solution based on residual analysis, anomaly detection, cluster analysis, and source tracing analysis. It can utilize the residual information obtained from mathematical models to identify and distinguish soil pollution sources of different types and locations, and calculate their potential hazards to the surrounding area.

[0046] Under optimal implementation conditions, the system has produced the following beneficial effects: improved the accuracy and efficiency of soil pollution source identification and location; enhanced the ability to control and remediate soil pollution sources; and promoted the effective implementation of soil environmental quality restoration and improvement work.

[0047] In optimal implementation, this system addresses the technical problem of the lack of effective calculation of the spatial and temporal diffusion range and intensity of soil pollutants in existing technologies.

[0048] Under optimal implementation conditions, the method for calculating the spatial and temporal diffusion range and intensity of pollutants is as follows:

[0049]

[0050] Where C(x, y, z, t) is the pollutant concentration at time t at a distance of (x, y, z) from the pollution source, Q is the emission amount from the pollution source, and D... x D y D z These are the diffusion coefficients of pollutants on the x, y, and z axes, respectively. When C(x, y, z, t) is greater than or equal to a preset concentration threshold, the location and time point are within the diffusion range of the pollutants. The value of C(x, y, z, t) reflects the diffusion intensity of the pollutants. The diffusion range is determined by the location and time point when the pollutant concentration equals the preset concentration threshold. That is, the boundary of the diffusion range is the surface or curve when the pollutant concentration equals the threshold.

[0051] In optimal implementation, the system employs a Gaussian diffusion model-based technical solution, which can calculate the soil pollutant concentration at any location and time point based on the emission amount, location, and diffusion coefficient of the pollution source, determine whether it is within the diffusion range of the pollutant, and reflect its diffusion intensity.

[0052] Under optimal implementation conditions, the system produces the following beneficial effects: it improves the accuracy and efficiency of calculating the spatial and temporal diffusion range and intensity of soil pollutants; it enhances the understanding and mastery of the spatial and temporal distribution characteristics and impact of soil pollutants; and it promotes the spatial and temporal monitoring and management of soil pollutants.

[0053] In optimal implementation, this system addresses the technical problem of the lack of comprehensive assessment and classification of soil environmental quality in existing technologies.

[0054] In a preferred implementation scenario, the pollution assessment module specifically includes:

[0055] The standard acquisition submodule is used to acquire national or regional soil environmental quality standards, including limit or allowable values ​​for soil pollutant index data.

[0056] The index acquisition submodule is used to obtain soil pollutant index data for all locations based on the fitting and prediction results of soil pollutant index data for unknown locations and soil pollutant index data for known locations.

[0057] The data comparison submodule is used to compare the soil pollutant index data of all locations with the soil environmental quality standards, and to determine whether the soil environmental quality at that location exceeds the standards based on the comparison results.

[0058] The grading submodule is used to grade the soil environmental quality at each location based on the comparison results obtained from the data comparison submodule and the national or regional soil environmental quality grading standards, including excellent, good, medium and poor.

[0059] The report generation submodule is used to generate a soil environmental quality assessment report based on the grading results obtained from the grading submodule, so as to reflect the actual status and changing trends of the soil environment, and send the soil environmental quality grade and assessment report to the data center.

[0060] In optimal implementation, the system employs a technical solution based on national or regional soil environmental quality standards, data comparison, and grading. It can compare soil pollutant index data at different locations with soil environmental quality standards, grade the soil environmental quality at each location based on the comparison results, and generate corresponding evaluation reports.

[0061] Under optimal implementation conditions, the system has produced the following beneficial effects: improved the objective understanding and scientific evaluation of the status and changing trends of soil environmental quality; enhanced the timely detection and effective management of soil environmental quality problems; and promoted the effective implementation of soil environmental protection and improvement work.

[0062] This monitoring method solves the technical problems of inaccurate, untimely, incomplete, and unpredictable data in existing soil pollution monitoring methods.

[0063] A real-time intelligent method for monitoring soil pollution includes the following steps:

[0064] Step 1: The data acquisition module collects soil pollutant index data in real time through soil pollutant sensors, including organic matter concentration, heavy metal content, pH, temperature and humidity, and transmits it to the data center via wireless network.

[0065] Step 2: The data center receives and stores soil pollutant index data;

[0066] Step 3: The data fusion module obtains the stored soil pollutant index data from the data center, and according to the source, type and time of the data, it uses the data integration method to preprocess the soil pollutant index data from different sources, types and times, and merges them into a soil pollutant index dataset, and sends the soil pollutant index dataset to the data center.

[0067] Step 4: The pollution modeling module retrieves the stored soil pollutant index dataset from the data center and, based on the data fusion method, uses regression analysis to model the soil pollutant index data at each known location, describing the relationship between the soil pollutant index data and location, time, and environmental factors, and fitting and predicting the soil pollutant index data at unknown locations or future time points.

[0068] Step 5: The pollution source tracing module uses the obtained mathematical model to determine the possible source location and type of soil pollutant index data that exceeds the standard or changes abnormally, and calculates the spatial and temporal diffusion range and intensity of pollutants.

[0069] This method has the following beneficial effects: it can monitor soil pollution status in real time, accurately and comprehensively, predict future soil pollution trends, trace the source of soil pollution, and issue early warning information in a timely manner, thereby effectively protecting and improving soil environmental quality.

[0070] Under optimal implementation conditions, this monitoring method addresses the technical problem of the lack of existing technologies for assessing and providing early warning of soil environmental quality.

[0071] In a preferred implementation, step 4 is followed by the following steps:

[0072] Step 4.1: Based on the fitting and prediction results of the soil pollutant index data for unknown locations, the soil pollutant index data for known locations, and the national or regional soil environmental quality standards, the pollution assessment module evaluates the soil pollutant index data for all locations and provides a soil environmental quality level and evaluation report to reflect the actual status and changing trends of the soil environment.

[0073] Step 4.2: Based on the fitting and prediction results of the soil pollutant index data for future time points, as well as the preset soil pollutant index data thresholds or warning values, the pollution early warning module issues warnings for exceeding the standards of soil pollutant index data for future time points. When the predicted value exceeds the soil pollutant index data threshold or warning value, it sends warning information to relevant departments or individuals so that countermeasures can be taken in advance.

[0074] Under optimal implementation conditions, this monitoring method produces the following beneficial effects: it enables more detailed assessment of the soil environmental quality level at each location and generates detailed evaluation reports; it enables more in-depth prediction of soil environmental quality trends at future time points and timely issuance of early warning information; it enables more effective improvement of soil environmental quality management and reduction of the risks posed by soil pollution.

[0075] Under optimal implementation conditions, this monitoring method addresses the technical problem of the lack of effective identification and location of soil pollution sources in existing technologies.

[0076] Under optimal implementation conditions, identifying the possible sources and types of soil pollutant index data that exceed standards or exhibit abnormal changes involves the following steps:

[0077] Step 51: Based on the mathematical model, calculate the residual of the soil pollutant index data at each location, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module.

[0078] Step 52: Perform outlier detection on the residuals and filter out residuals that exceed the preset residual threshold as outliers;

[0079] Step 53: Perform cluster analysis on the outliers, grouping adjacent or nearby outliers into one category as an outlier region;

[0080] Step 54: For each abnormal area, determine the possible pollution source type based on the sign, magnitude, and distribution of the residual, including point source, area source, and line source.

[0081] Step 55: For each anomalous region, estimate the possible location of the pollution source, including the center point, edge point, and intermediate point, based on the direction and gradient of the residual.

[0082] Under optimal implementation conditions, this monitoring method produces the following beneficial effects: it enables effective tracing of soil pollution sources; by calculating residuals, detecting outliers, performing cluster analysis, determining pollution source types, and estimating pollution source locations, it can identify deviations and inconsistencies between soil pollutant index data and mathematical models, thereby identifying the existence and characteristics of soil pollution sources; it can help determine the responsible parties and remediation measures for soil pollution sources, as well as prevent or reduce the impact of soil pollution sources on surrounding areas. Attached Figure Description

[0083] Figure 1 This is a schematic diagram of the system structure in Embodiment 1 of the present invention;

[0084] Figure 2 This is a schematic diagram of the system structure in Embodiment 2 of the present invention;

[0085] Figure 3 This is a flowchart of the method in Embodiment 3 of the present invention. Detailed Implementation

[0086] To enable those skilled in the art to better understand the technical solution, the technical solution will be described in detail below with reference to the embodiments. The description in this part is only exemplary and explanatory, and should not be used to limit the scope of protection of the present invention in any way.

[0087] Example 1:

[0088] This embodiment provides a real-time intelligent soil pollution monitoring system, such as... Figure 1 Specifically, it includes the following modules:

[0089] The data acquisition module is used to collect real-time soil pollutant index data, including organic matter concentration, heavy metal content, pH, temperature, and humidity, through soil pollutant sensors, and transmit this data to the data center via a wireless network. Specifically, this module deploys several soil pollutant sensors in the area, each with a unique number and location information. At regular time intervals (e.g., every 10 minutes), the sensors automatically collect and upload soil pollutant index data for their current location. For example, sensor number 001, located at longitude 120.1234°, latitude 30.5678°, collected and uploaded the following data at 10:00:00 on October 31, 2023:

[0090] organic matter 0.12 mg / kg Heavy metal (lead) 0.05 mg / kg pH level 6.8 temperature 25.6 ℃ humidity 45.3 ％

[0091] This module encapsulates the data in JSON format and sends it to the data center via wireless network (such as 4G or 5G).

[0092] A data center is used to receive and store data. A data center is a cloud server with large storage capacity and high-speed processing capabilities, capable of receiving and storing data from various sensors.

[0093] The data fusion module is used to acquire stored soil pollutant index data from the data center, and, based on the data's source, type, and time, employ data integration methods to preprocess the soil pollutant index data from different sources, types, and times, and then merge them into a soil pollutant index dataset, which is then sent back to the data center. Specifically, this module includes the following sub-modules:

[0094] The data quality assessment submodule evaluates the quality of acquired soil pollutant indicator data, including completeness, accuracy, consistency, and timeliness. Based on the assessment results, it filters, corrects, or deletes data and outputs the data to the data alignment submodule. For example, this submodule checks whether each data entry contains all necessary fields (such as ID, location, time, and various indicators), whether there are missing or outlier values ​​(such as null or negative values), whether there are conflicts or contradictions with other data (such as significant differences in data from the same location at different times), and whether it matches the current time (such as expired or future data). If problems are found, they are handled according to different situations, such as deleting invalid or erroneous data, filling in missing or outlier values ​​with averages or interpolation, and replacing conflicting or contradictory data with the latest or most reliable data.

[0095] The data alignment submodule is used to align soil pollutant indicator data from different sources, types, and times, including spatial alignment, temporal alignment, and attribute alignment, to eliminate differences and conflicts between data. For example, this submodule converts location information of different precision or formats into uniform latitude and longitude coordinates, converts time information of different frequencies or time zones into uniform timestamps, and converts indicator data of different units or ranges into uniform standard values.

[0096] The data fusion submodule is used to fuse the aligned soil pollutant indicator data, generate one or more soil pollutant indicator datasets based on the fusion results, and send them to the data center. For example, this submodule selects different data fusion methods, such as simple averaging, weighted averaging, maximum value method, and minimum value method, depending on different purposes or needs, to merge or discard multiple data points from the same or similar locations to obtain a more representative or reliable dataset. Simultaneously, this submodule also groups or classifies the data according to different dimensions or levels, resulting in multiple different datasets, such as dividing them into different regions or grids according to location, different time periods or cycles according to time, and different types or categories according to indicators.

[0097] The pollution modeling module retrieves the stored soil pollutant index dataset from the data center and, based on the data fusion format, uses regression analysis to model the soil pollutant index data at each known location. This model describes the relationship between the soil pollutant index data and location, time, and environmental factors. It also fits and predicts soil pollutant index data for unknown locations or future time points. Specifically, this module includes the following sub-modules:

[0098] The feature extraction submodule is used to extract characteristic variables from the soil pollutant index dataset that help describe and predict soil pollutant index data, including location, time, and environmental factors. For example, this submodule selects appropriate characteristic variables as independent variables (X) based on the characteristics and influencing factors of different indicators. For instance, organic matter is greatly affected by temperature and humidity, so temperature and humidity are selected as characteristic variables; heavy metals are greatly affected by location and pH, so location and pH are selected as characteristic variables, and so on. Simultaneously, this submodule also uses the corresponding soil pollutant index data as dependent variables (Y), such as organic matter and heavy metals.

[0099] The model selection submodule is used to select a regression analysis method based on the number, type, and distribution of the feature variables, including linear regression, multiple regression, and nonlinear regression. For example, this submodule will select the most suitable regression analysis method based on whether a linear relationship exists between the feature variables and the dependent variable, and whether multicollinearity exists among the feature variables. If a linear relationship exists between the feature variables and the dependent variable, a linear regression method is selected; if multicollinearity exists, a multiple regression method is selected; if no linear relationship exists between the feature variables and the dependent variable, a nonlinear regression method is selected, and so on.

[0100] The model training submodule is used to build a mathematical model based on the selected regression analysis method, using known soil pollutant index data and characteristic variables. It then adjusts the model parameters using optimization algorithms and sends the optimized mathematical model to the data center. For example, this submodule selects appropriate optimization algorithms, such as gradient descent, Newton's method, or genetic algorithms, to solve for the model parameters, ensuring the model fits the known data to the greatest extent possible and minimizes the error function. This submodule then sends the solved model parameters and error function values ​​as part of the mathematical model to the data center.

[0101] The model evaluation submodule is used to assess the goodness of fit and predictive ability of the established mathematical model using unknown soil pollutant index data and characteristic variables. It also verifies the model's stability and reliability through error analysis and sensitivity analysis. For example, this submodule uses different evaluation metrics, such as R-squared, mean squared error, and mean absolute error, to measure the model's fit to known data and its predictive accuracy for unknown data, and provides corresponding evaluation results. Simultaneously, this submodule also uses different analytical methods, such as residual distribution plots, homogeneity of variance tests, and correlation tests of characteristic variables, to examine whether the model has systematic biases or overfitting issues, and provides corresponding improvement suggestions.

[0102] The model application submodule is used to fit and predict soil pollutant index data at unknown locations or future time points based on the established mathematical model, and then send the fitting and prediction results to the data center. For example, this submodule will input the corresponding feature variable values ​​(such as location, time, etc.) according to the user's query or needs, use the mathematical model to calculate the corresponding dependent variable values ​​(such as soil pollutant index data, etc.), and display the calculation results to the user in the form of charts or text.

[0103] The pollution source tracing module uses the obtained mathematical model to determine the possible source location and type of pollutants causing excessive or abnormal changes in soil pollutant index data, and calculates the spatial and temporal diffusion range and intensity of pollutants. The determination and calculation results are then sent to the data center. Specifically, determining the possible source location and type of pollutants causing excessive or abnormal changes in soil pollutant index data is achieved through the following sub-modules:

[0104] The residual calculation submodule is used to calculate the residual of soil pollutant index data at each location based on the mathematical model, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module. For example, this submodule will use the mathematical model to predict the soil pollutant index data at each location and compare it with the actual collected data to obtain the residual value at each location.

[0105] The anomaly detection submodule is used to detect outliers in the residuals, filtering out residuals that exceed a preset residual threshold as outliers. For example, this submodule will determine a reasonable residual threshold based on different outlier detection methods, such as box plot method, 3σ method, Mahalanobis distance method, etc., and consider residuals exceeding the threshold as outliers.

[0106] The clustering analysis submodule is used to perform clustering analysis on outliers, grouping adjacent or similar outliers into a single class as an outlier region. For example, this submodule will determine an appropriate number of clusters and clustering criteria based on different clustering analysis methods, such as K-means, hierarchical clustering, and density clustering, and group outliers according to their location distance or similarity to obtain several outlier regions.

[0107] The pollution source type determination submodule is used to determine the possible pollution source type for each abnormal area based on the sign, magnitude, and distribution of the residuals. This includes point sources, area sources, and line sources, and the results are sent to the data center. For example, this submodule will make the determination based on the following rules:

[0108] If the residuals are positive and large, and distributed over a small range, they may be caused by a single point source.

[0109] If the residuals are positive and large, and distributed over a large range, they may be caused by a surface source.

[0110] If the residuals are positive and large, and distributed along a linear or curved shape, they are likely caused by a linear source.

[0111] The pollution source location estimation module estimates the possible pollution source locations (center, edge, and intermediate points) for each anomalous region based on the direction and gradient of the residuals, and then sends these estimates to the data center. For example, this module might estimate the source using the following method:

[0112] If the anomaly is caused by a point source, then the likely source of contamination is the point with the largest residual in that anomaly.

[0113] If the abnormal area is caused by a surface source, then the possible location of the pollution source is the center point of the abnormal area.

[0114] If the anomalous area is caused by a linear source, the likely location of the source of contamination is the edge or midpoint of the anomalous area.

[0115] The spatial and temporal diffusion range and intensity of pollutants are calculated using the following formula:

[0116]

[0117] Here, x, y, and z are variables on the coordinate axes, representing the distances along the horizontal direction of groundwater flow (i.e., the x-axis), the horizontal direction perpendicular to the groundwater flow direction (i.e., the y-axis), and the direction perpendicular to the ground surface (i.e., the z-axis).

[0118] t is a time variable, which represents the time interval from the start of pollution emissions from the pollution source to the observation time point;

[0119] C(x, y, z, t) is the pollutant concentration at time t at a distance of (x, y, z) from the pollution source;

[0120] Q represents the emissions from the pollution source;

[0121] D x D y and D z These are the diffusion coefficients of pollutants along the x, y, and z axes, respectively, which represent the squared distances diffused along the coordinate axes per unit time.

[0122] When C(x, y, z, t) is greater than or equal to the preset concentration threshold, the location and time point are within the diffusion range of the pollutant. The value of C(x, y, z, t) reflects the diffusion intensity of the pollutant. The diffusion range is determined by the location and time point when the pollutant concentration equals the preset concentration threshold. That is, the boundary of the diffusion range is the surface or curve when the pollutant concentration equals the threshold.

[0123] The diffusion range refers to the spatial and temporal distribution range of the diffusing substance, while the diffusion intensity refers to the spatial and temporal concentration or density of the diffusing substance. The diffusion range and diffusion intensity are represented by the same variable, namely the concentration or density of the diffusing substance at a certain location and at a certain time.

[0124] The diffusion range is determined by the diffusion intensity because the concentration or density of the diffusing substance is related to a preset threshold or standard. When the concentration or density of the diffusing substance is greater than or equal to the threshold or standard, it means that the location and time are within the diffusion range; when the concentration or density of the diffusing substance is less than the threshold or standard, it means that the location and time are not within the diffusion range.

[0125] For example, the diffusion range of soil pollutants can be determined based on national or regional soil environmental quality standards. When the concentration of pollutants exceeds the soil environmental quality standards, it indicates that the area is polluted; when the concentration of pollutants is below the soil environmental quality standards, it indicates that the area is not polluted. Therefore, the diffusion range is determined by the concentration of pollutants (i.e., the diffusion intensity).

[0126] Example 2:

[0127] like Figure 2 As shown, this embodiment, compared to embodiment 1, also provides a pollution assessment module and a standard acquisition submodule.

[0128] The pollution assessment module is used to assess soil pollutant index data for all locations based on the fitting and prediction results of soil pollutant index data for unknown locations obtained from the data center, soil pollutant index data for known locations, and national or regional soil environmental quality standards. It then provides a soil environmental quality level and assessment report to reflect the actual condition and changing trends of the soil environment. Specifically, this module includes the following sub-modules:

[0129] The standards acquisition submodule is used to obtain national or regional soil environmental quality standards, including limits or allowable values ​​for soil pollutant indicators. For example, this submodule will obtain the "Soil Environmental Quality Standard" (GB15618-2018) from relevant departments or websites, which specifies the limits or allowable values ​​for different indicators (such as organic matter and heavy metals) for different types of soil (such as agricultural land and construction land).

[0130] The index acquisition submodule is used to obtain soil pollutant index data for all locations based on the fitting and prediction results of soil pollutant index data at unknown locations and soil pollutant index data at known locations. For example, this submodule will obtain the actual collected data at known locations and the data calculated by mathematical models at unknown locations from the data center, and integrate these data into a complete dataset.

[0131] The data comparison submodule is used to compare soil pollutant index data from all locations with soil environmental quality standards and determine whether the soil environmental quality at a given location exceeds the standards based on the comparison results. For example, this submodule will compare each data point in the dataset with the corresponding limit or allowable value based on different soil types and different indicators, and provide an exceedance rate or exceedance multiple to reflect the gap between the data and the standard.

[0132] The grading submodule is used to grade the soil environmental quality at each location based on the comparison results obtained from the data comparison submodule and the national or regional soil environmental quality grading standards, including excellent, good, medium, and poor. For example, this submodule will classify the soil environmental quality at each location into four grades: excellent (Class I), good (Class II), medium (Class III), and poor (Class IV) based on the grade classification of the soil environmental quality standards and the exceedance rate or exceedance multiple obtained from the data comparison submodule.

[0133] The report generation submodule generates a soil environmental quality assessment report based on the grading results obtained from the grading submodule. This report reflects the actual condition and changing trends of the soil environment, and the soil environmental quality grade and assessment report are then sent to the data center. For example, this submodule will display the grading results to the user in a visual format, using different display methods such as maps, charts, and text, and provide corresponding evaluations and suggestions, such as the overall soil environmental quality, major pollutants, pollution levels, influencing factors, and improvement measures.

[0134] The pollution early warning module is used to provide early warnings of exceeding standards for soil pollutant indicators at future time points based on the fitting and prediction results of soil pollutant indicator data for future time points obtained from the data center, as well as preset soil pollutant indicator data thresholds or early warning values. When the predicted value exceeds the soil pollutant indicator data threshold or early warning value, it sends an early warning message to relevant departments or individuals to allow for proactive countermeasures. Specifically, this module includes the following steps:

[0135] Threshold or warning value setting is used to set a reasonable threshold or warning value for soil pollutant indicators based on national or regional soil environmental quality standards or user-defined needs, serving as the basis for determining whether a warning is needed. For example, this step can be based on the limits or allowable values ​​specified in the "Soil Environmental Quality Standard" (GB 15618-2018), or the user can set a stricter or more lenient value as the threshold or warning value according to their own needs.

[0136] The prediction result acquisition step involves obtaining soil pollutant index data for each location over a future period (such as a day, a week, or a month) based on the fitting and prediction results of soil pollutant index data for future time points. For example, this step can retrieve data for future time points that have already been calculated using mathematical models from a data center and integrate this data into a prediction dataset.

[0137] The exceedance judgment step compares each data point in the prediction dataset with a threshold or warning value, and determines whether an exceedance will occur at that location and time point based on the comparison result. For example, this step can compare each data point in the prediction dataset with the corresponding threshold or warning value based on different soil types and different indicators, and give an exceedance rate or exceedance multiple to reflect the gap between the data and the threshold or warning value.

[0138] The early warning information generation step is used to generate corresponding early warning information based on the results of the exceedance judgment, and to send early warning information to relevant departments or individuals when necessary, so that countermeasures can be taken in advance. For example, this step can display the exceedance judgment results to users in a visual form, such as maps, charts, and text, and provide corresponding prompts and suggestions, such as the location of exceedance, the time of exceedance, the degree of exceedance, possible causes, and emergency measures.

[0139] Example 3:

[0140] This embodiment provides a real-time intelligent method for monitoring soil pollution, such as... Figure 3 As shown, the method includes the following steps:

[0141] Step 1: The data acquisition module collects real-time soil pollutant index data, including organic matter concentration, heavy metal content, pH, temperature, and humidity, using soil pollutant sensors, and transmits this data to the data center via a wireless network. For example, the data acquisition module can install soil pollutant sensors at several locations within a farmland area (such as field ridges, irrigation ditches, and crops) to detect indicators such as organic matter, heavy metals, pH, temperature, and humidity in the soil at regular time intervals (e.g., every 10 minutes) and convert them into electrical signals. The data acquisition module can also transmit these electrical signals to the data center via a wireless communication module.

[0142] Step 2: The data center receives and stores soil pollutant indicator data. For example, the data center can receive electrical signals from the data acquisition module via a wireless receiving module and convert them into digital signals. The data center can also store the received digital signals locally or remotely via a data storage module (such as a database or cloud storage) and provide query and retrieval functions for other modules.

[0143] Step 3: The data fusion module retrieves stored soil pollutant index data from the data center. Based on the data source, type, and time, it employs a data integration method to preprocess the soil pollutant index data from different sources, types, and times, and then merges them into a soil pollutant index dataset. This dataset is then sent to the data center. For example, the data fusion module can retrieve stored soil pollutant index data from the data center and, based on the data source (e.g., different sensor nodes), type (e.g., different pollutant indicators), and time (e.g., different collection times), use a weighted average-based data integration method to preprocess the soil pollutant index data from different sources, types, and times (e.g., noise removal, missing value imputation, normalization, etc.), and merge them into a soil pollutant index dataset containing several sample points, several feature variables (organic matter, heavy metals, pH, temperature, and humidity), and one response variable (soil environmental quality level). This dataset is then sent to the data center.

[0144] Step 4: The pollution modeling module retrieves the stored soil pollutant index dataset from the data center and, based on the data fusion format, uses regression analysis to model the soil pollutant index data at each known location. This model describes the relationship between the soil pollutant index data and location, time, and environmental factors. It also fits and predicts the soil pollutant index data at unknown locations or future time points. For example, the pollution modeling module can retrieve the stored soil pollutant index dataset from the data center and, based on the data fusion format, use a multiple linear regression analysis method to model the soil environmental quality level at each known location. This model describes the relationship between the soil environmental quality level and location, time, and environmental factors (such as organic matter, heavy metals, pH, temperature, and humidity). It also fits and predicts the soil environmental quality level at unknown locations or future time points.

[0145] Step 4.1: The pollution assessment module evaluates the soil pollutant index data for all locations based on the fitting and prediction results of the soil pollutant index data for unknown locations, the soil pollutant index data for known locations, and the national or regional soil environmental quality standards. It then provides the soil environmental quality level and an evaluation report to reflect the actual condition and changing trends of the soil environment. For example, the pollution assessment module can evaluate the soil environmental quality level data for all locations (e.g., all locations within a farmland area) based on the fitting and prediction results of the soil environmental quality level for unknown locations (e.g., locations within a farmland area where sensor nodes are installed), the soil environmental quality level for known locations (e.g., locations within a farmland area where sensor nodes are installed), and the national or regional (e.g., China's) soil environmental quality standards (e.g., "Soil Environmental Quality Standard" GB15618-2018). It then provides the soil environmental quality level (e.g., Level I, Level II, Level III, Level IV, or Level V) and an evaluation report (e.g., Excellent, Good, Slightly Polluted, Moderately Polluted, or Severely Polluted) to reflect the actual condition and changing trends of the soil environment.

[0146] Step 4.2: Based on the fitting and prediction results of soil pollutant index data for future time points (e.g., the next day, week, or month), and preset soil pollutant index data thresholds or warning values, the pollution early warning module issues warnings for exceeding standards for soil pollutant index data at future time points. When the predicted value exceeds the soil pollutant index data threshold or warning value, it sends warning information to relevant departments or individuals to allow for proactive countermeasures. For example, the pollution early warning module can, based on the fitting and prediction results of soil environmental quality levels for future time points (e.g., the next day), and preset soil environmental quality level thresholds or warning values ​​(e.g., Level III or Level IV), issue warnings for exceeding standards for soil environmental quality levels at future time points (e.g., tomorrow). When the predicted value exceeds the soil environmental quality level threshold or warning value, it sends warning information to relevant departments or individuals to allow for proactive countermeasures. For example, the pollution early warning module can send warning information to agricultural departments, environmental protection departments, farmers, or other relevant personnel via SMS, telephone, WeChat, and other communication methods and mobile networks or wireless networks.

[0147] Step 5: The pollution source tracing module uses the obtained mathematical model to determine the possible source locations and types that cause soil pollutant index data to exceed standards or exhibit abnormal changes, and calculates the spatial and temporal diffusion range and intensity of pollutants. Specifically, the possible source locations and types that cause soil pollutant index data to exceed standards or exhibit abnormal changes are determined through the following steps:

[0148] Step 51: Based on the mathematical model, calculate the residual of the soil pollutant index data at each location, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module. For example, the pollution source tracing module can calculate the residual of the soil environmental quality level at each location within the farmland area based on the above mathematical model, and use it as input for anomaly detection.

[0149] Step 52: Perform outlier detection on the residuals and filter out residuals that exceed a preset residual threshold as outliers. For example, the pollution source tracing module can use an outlier detection method based on the Local Outlier Factor (LOF) to detect outliers in the residuals and identify residuals that exceed a preset residual threshold (e.g., 0.5) as outliers.

[0150] Step 53: Perform cluster analysis on the outliers, grouping adjacent or nearby outliers into a single cluster as an outlier region. For example, the pollution source tracing module can use a density-based clustering method to cluster outliers and group adjacent or nearby outliers (e.g., within 10 meters) into a single outlier region. For instance, if there are two locations within a farmland area with residuals of 0.6 and 0.7 respectively, and the distance between these two locations is less than 10 meters, the pollution source tracing module can group the outliers at these two locations into one outlier region and use it as input for pollution source tracing. If there are three locations within a farmland area with residuals of 0.8, 0.9, and 1.0 respectively, and the distance between these three locations is greater than 10 meters, the pollution source tracing module can group the outliers at these three locations into three separate outlier regions and use each as input for pollution source tracing.

[0151] Step 54: For each abnormal region, determine its possible pollution source type based on the sign, magnitude, and distribution of the residuals, including point sources, area sources, and line sources. For example, the pollution source tracing module can determine the type of pollution source that may have caused each abnormal region based on the sign, magnitude, and distribution of the residuals. If the residual is positive and large, and distributed over a small area, it may be caused by a point source; if the residual is positive and large, and distributed over a large area, it may be caused by an area source; if the residual is positive and large, and distributed along a linear or curved shape, it may be caused by a line source. For example, if there is an anomalous area in a farmland area with a positive and large residual, and it is distributed within a circular area with a diameter of about 5 meters, the pollution source tracing module can determine that the anomalous area may be caused by a point source; if there is an anomalous area in a farmland area with a positive and large residual, and it is distributed within a rectangular area with a length of about 50 meters and a width of about 10 meters, the pollution source tracing module can determine that the anomalous area may be caused by a area source; if there is an anomalous area in a farmland area with a positive and large residual, and it is distributed along a curved shape with a length of about 100 meters and a width of about 5 meters, the pollution source tracing module can determine that the anomalous area may be caused by a line source.

[0152] Step 55: For each anomalous region, estimate its possible pollution source location based on the direction and gradient of the residuals, including the center point, edge points, and intermediate points. For example, the pollution source tracing module can estimate the possible pollution source location for each anomalous region based on the direction and gradient of the residuals. If the anomalous region is caused by a point source, the possible pollution source location is the point with the largest residual in the anomalous region; if the anomalous region is caused by a surface source, the possible pollution source location is the center point of the anomalous region; if the anomalous region is caused by a line source, the possible pollution source location is the edge point or intermediate point of the anomalous region. For example, if there is an anomalous area in a farmland area with a positive and large residual, and it is distributed within a circular area with a diameter of about 5 meters, the pollution source tracing module can estimate that the possible pollution source location of the anomalous area is the point with the largest residual within the circular area. If there is an anomalous area in a farmland area with a positive and large residual, and it is distributed within a rectangular area with a length of about 50 meters and a width of about 10 meters, the pollution source tracing module can estimate that the possible pollution source location of the anomalous area is the center point within the rectangular area. If there is an anomalous area in a farmland area with a positive and large residual, and it is distributed along a curved shape with a length of about 100 meters and a width of about 5 meters, the pollution source tracing module can estimate that the possible pollution source location of the anomalous area is the edge point or the middle point on the curved shape.

[0153] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the present invention. These examples are merely for the purpose of helping to understand the method and core ideas of the present invention. The above descriptions are only preferred embodiments of the present invention. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or variations without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, variations, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of the present invention.

Claims

1. A real-time intelligent soil pollution monitoring system, characterized in that, include: The data acquisition module is used to collect soil pollutant index data in real time through soil pollutant sensors, including organic matter concentration, heavy metal content, pH, temperature and humidity, and transmit them to the data center via wireless network; Data centers are used to receive and store data; The data fusion module is used to obtain the stored soil pollutant index data from the data center, and according to the source, type and time of the data, it uses the data integration method to preprocess the soil pollutant index data from different sources, types and times, and merge them into a soil pollutant index dataset, and then send the soil pollutant index dataset to the data center. The pollution modeling module is used to retrieve the stored soil pollutant index dataset from the data center, and, based on the data fusion method, to model the soil pollutant index data at each known location using regression analysis, describing the relationship between the soil pollutant index data and location, time, and environmental factors, and to fit and predict the soil pollutant index data at unknown locations or future time points, and send the mathematical model and the fitting and prediction results to the data center. The pollution source tracing module is used to determine the possible source location and type of soil pollutant index data exceeding the standard or abnormal changes by using mathematical models obtained from the data center, and to calculate the spatial and temporal diffusion range and intensity of pollutants, and send the determination results and calculation results to the data center. The pollution source tracing module determines the possible source location and type of soil pollutant index data exceeding the standard or showing abnormal changes through the following modules: The residual calculation submodule is used to calculate the residual of the soil pollutant index data at each location according to the mathematical model, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module. The anomaly detection submodule is used to detect outliers in the residuals and filter out residuals that exceed a preset residual threshold as outliers. The clustering analysis submodule is used to perform clustering analysis on outliers, grouping adjacent or close outliers into one class as an outlier region. The pollution source type determination submodule is used to determine the possible pollution source type for each abnormal area based on the sign, size, and distribution of the residual, including point source, area source, and line source, and send it to the data center. The pollution source location estimation module is used to estimate the possible pollution source locations for each abnormal area based on the direction and gradient of the residual, including the center point, edge point and intermediate point, and send them to the data center. The method for calculating the spatial and temporal diffusion range and intensity of the pollutants is as follows: ;in, Distance from pollution source Location, in time t The concentration of pollutants at that time Q The amount of emissions from pollution sources. These are the diffusion coefficients of the pollutants along the x-axis, y-axis, and z-axis, respectively; when... When the concentration is greater than or equal to a preset concentration threshold, then the location and time point are within the diffusion range of the pollutant, and The value reflects the diffusion intensity of pollutants. The diffusion range is determined by the location and time point when the pollutant concentration equals the preset concentration threshold. That is, the boundary of the diffusion range is the surface or curve when the pollutant concentration equals the threshold.

2. The real-time intelligent soil pollution monitoring system according to claim 1, characterized in that, The system also includes: The pollution assessment module is used to assess the soil pollutant index data of all locations based on the fitting and prediction results of soil pollutant index data of unknown locations obtained from the data center, soil pollutant index data of known locations, and national or regional soil environmental quality standards, and to provide soil environmental quality level and evaluation report to reflect the actual status and changing trend of the soil environment, and send the soil environmental quality level and evaluation report to the data center. The pollution early warning module is used to provide early warnings of exceeding the standards for soil pollutant index data at future time points based on the fitting and prediction results of soil pollutant index data for future time points obtained from the data center, as well as the preset soil pollutant index data thresholds or early warning values. When the predicted value exceeds the soil pollutant index data threshold or early warning value, the module sends early warning information to relevant departments or individuals so that countermeasures can be taken in advance, and sends the early warning results to the data center.

3. The real-time intelligent soil pollution monitoring system according to claim 1, characterized in that, The data fusion module specifically includes: The data quality assessment submodule is used to assess the quality of the acquired soil pollutant index data, including completeness, accuracy, consistency and timeliness. Based on the assessment results, the data is filtered, corrected or deleted, and the data is output to the data alignment submodule. The data alignment submodule is used to align soil pollutant index data from different sources, types, and times, including spatial alignment, temporal alignment, and attribute alignment, to eliminate differences and conflicts between data. The data fusion submodule is used to fuse the aligned soil pollutant index data, generate one or more soil pollutant index datasets based on the fusion results, and send them to the data center.

4. The real-time intelligent soil pollution monitoring system according to claim 1, characterized in that, The pollution modeling module specifically includes: The feature extraction submodule is used to extract feature variables from the soil pollutant index dataset that help describe and predict the soil pollutant index data, including location, time, and environmental factors. The model selection submodule is used to select the regression analysis method based on the number, type, and distribution of the feature variables, including linear regression, multiple regression, and nonlinear regression. The model training submodule is used to establish a mathematical model based on the selected regression analysis method, using known soil pollutant index data and characteristic variables, and to adjust the model parameters through optimization algorithms before sending the optimized mathematical model to the data center. The model evaluation submodule is used to evaluate the model's fit and predictive ability based on the established mathematical model, using unknown soil pollutant index data and characteristic variables, and to test the model's stability and reliability through error analysis and sensitivity analysis. The model application submodule is used to fit and predict soil pollutant index data at unknown locations or future time points based on the established mathematical model, and send the fitting and prediction results to the data center.

5. The real-time intelligent soil pollution monitoring system according to claim 2, characterized in that, The pollution assessment module specifically includes: The standard acquisition submodule is used to acquire national or regional soil environmental quality standards, including limit or allowable values ​​for soil pollutant index data. The index acquisition submodule is used to obtain soil pollutant index data for all locations based on the fitting and prediction results of soil pollutant index data for unknown locations and soil pollutant index data for known locations. The data comparison submodule is used to compare the soil pollutant index data of all locations with the soil environmental quality standards, and to determine whether the soil environmental quality at that location exceeds the standards based on the comparison results. The grading submodule is used to grade the soil environmental quality at each location based on the comparison results obtained from the data comparison submodule and the national or regional soil environmental quality grading standards, including excellent, good, medium and poor. The report generation submodule is used to generate a soil environmental quality assessment report based on the grading results obtained from the grading submodule, so as to reflect the actual status and changing trends of the soil environment, and send the soil environmental quality grade and assessment report to the data center.

6. A real-time intelligent soil pollution monitoring method based on the soil pollution monitoring system of any one of claims 1-5, characterized in that, Includes the following steps: Step 1: The data acquisition module collects soil pollutant index data in real time through soil pollutant sensors, including organic matter concentration, heavy metal content, pH, temperature and humidity, and transmits it to the data center via wireless network. Step 2: The data center receives and stores soil pollutant index data; Step 3: The data fusion module obtains the stored soil pollutant index data from the data center, and according to the source, type and time of the data, it uses the data integration method to preprocess the soil pollutant index data from different sources, types and times, and merges them into a soil pollutant index dataset, and sends the soil pollutant index dataset to the data center. Step 4: The pollution modeling module retrieves the stored soil pollutant index dataset from the data center and, based on the data fusion method, uses regression analysis to model the soil pollutant index data at each known location, describing the relationship between the soil pollutant index data and location, time, and environmental factors, and fitting and predicting the soil pollutant index data at unknown locations or future time points. Step 5: The pollution source tracing module uses the obtained mathematical model to determine the possible source location and type of soil pollutant index data that exceeds the standard or changes abnormally, and calculates the spatial and temporal diffusion range and intensity of pollutants.

7. The real-time intelligent soil pollution monitoring method according to claim 6, characterized in that, Step 4 is followed by the following steps: Step 4.1: Based on the fitting and prediction results of the soil pollutant index data for unknown locations, the soil pollutant index data for known locations, and the national or regional soil environmental quality standards, the pollution assessment module evaluates the soil pollutant index data for all locations and provides a soil environmental quality level and evaluation report to reflect the actual status and changing trends of the soil environment. Step 4.2: Based on the fitting and prediction results of the soil pollutant index data for future time points, as well as the preset soil pollutant index data thresholds or warning values, the pollution early warning module issues warnings for exceeding the standards of soil pollutant index data for future time points. When the predicted value exceeds the soil pollutant index data threshold or warning value, it sends warning information to relevant departments or individuals so that countermeasures can be taken in advance.

8. The real-time intelligent soil pollution monitoring method according to claim 6, characterized in that, The determination of the possible sources and types of soil pollutant index data exceeding standards or showing abnormal changes specifically includes: Step 51: Based on the mathematical model, calculate the residual of the soil pollutant index data at each location, that is, the difference between the soil pollutant index data collected by the data acquisition module and the soil pollutant index data predicted by the pollution modeling module. Step 52: Perform outlier detection on the residuals and filter out residuals that exceed the preset residual threshold as outliers; Step 53: Perform cluster analysis on the outliers, grouping adjacent or nearby outliers into one category as an outlier region; Step 54: For each abnormal area, determine the possible pollution source type based on the sign, magnitude, and distribution of the residual, including point source, area source, and line source. Step 55: For each anomalous region, estimate the possible location of the pollution source, including the center point, edge point, and intermediate point, based on the direction and gradient of the residual.

Images