A method and system for determining the content of organic pollutants in soil

Abstract

The present application relates to the technical field of soil data analysis, and particularly relates to a soil organic pollutant content determination method and system; comprising: step S1: a three-dimensional sampling grid is established through the soil to be measured, a plurality of sampling points are set, and monitoring data of physical matrix parameters and multi-dimensional organic matter concentration are collected correspondingly; step S2: a percolation reference sampling point set corresponding to each sampling point is constructed, the physical matrix parameters and multi-dimensional organic matter concentration of each sampling point in the current sampling point and the corresponding percolation reference sampling point set are analyzed, and a sampling point dynamic diffusion influence degree based on percolation topology is obtained; step S3: a high-dimensional joint index vector set is established and input into an initialized neural network, a penalty function is embedded and the neural network is updated through the diffusion influence degree, all sampling points are distinguished by using the updated neural network, it is judged whether the sampling points are stable or change, and the organic pollutant concentration data of each sampling point is output correspondingly; false positive determination errors are overcome.

Description

Technical Field

[0001] This invention relates to the field of soil data analysis technology, specifically to a method and system for determining the content of organic pollutants in soil. Background Technology

[0002] Soil is a core component of the ecosystem, and its health directly affects agricultural production and environmental safety. Soil organic pollutants mainly include volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), petroleum hydrocarbons, and polycyclic aromatic hydrocarbons. Due to their characteristics of being easily volatile, easily migrating in groundwater and soil pores, and easily adsorbed by soil organic matter, these pollutants exhibit strong three-dimensional heterogeneity and dynamic temporal variability in their spatial distribution in the underground environment. In other words, organic pollutants pose a serious threat to soil ecology due to their persistence, bioaccumulation, and toxicity.

[0003] Therefore, accurate and efficient determination of organic pollutants in soil has become a core requirement of environmental monitoring. With the increasing demand for real-time detection of soil organic pollutants, portable detection devices have become widely used. While existing portable detection devices have improved the timeliness of data acquisition to some extent, they often face some shortcomings in practical applications. The measurement and identification systems configured in these devices often only focus on the total concentration at a specific moment, ignoring the dynamic distribution of organic pollutants between the gas phase (soil gas) and the solid phase (soil particle adsorption) and their natural degradation over time. This results in the inability to identify the dynamic evolution characteristics of soil conditions at historical sampling points, leading to deviations and low accuracy in the measurement results of organic pollutant content obtained by portable detection devices. Summary of the Invention

[0004] To address the technical problem of measurement bias and reduced accuracy in existing methods for determining soil organic pollutant content, the present invention aims to provide a method for determining soil organic pollutant content, the specific technical solution of which is as follows:

[0005] Step S1: Establish a three-dimensional sampling grid using the soil to be tested, set multiple sampling points, and collect monitoring data on physical matrix parameters and multidimensional organic matter concentrations accordingly;

[0006] Step S2: Construct a permeation reference sampling point set corresponding to each sampling point, analyze the physical matrix parameters and multidimensional organic matter concentration of each sampling point in the current sampling point and the corresponding permeation reference sampling point set, and obtain the degree of dynamic diffusion influence of the sampling point based on the permeation topology;

[0007] Step S3: Integrate all sampling points to establish a high-dimensional joint index vector set and input it into the initialized neural network. Construct a penalty function based on the degree of diffusion influence and embed it into the neural network to update it. Use the updated neural network to distinguish all sampling points, determine whether the sampling points are stable or changing, and output the organic pollutant concentration data for each sampling point accordingly.

[0008] Preferably, step S1 includes:

[0009] Step S11: Establish a three-dimensional sampling grid on the soil to be tested to obtain multiple sampling points;

[0010] Step S12: Perform in-situ and ex-situ sampling at each sampling point to obtain historical baseline data and determine physical matrix parameters; obtain the VOCs concentration in the gas phase of the soil through detection equipment and determine the SVOCs concentration in the solid phase of the soil.

[0011] Preferably, the physical matrix parameters include, but are not limited to, total organic carbon content, soil porosity, and permeability parameters.

[0012] Preferably, step S2 includes:

[0013] Step S21: Construct a set of permeability reference sampling points for each sampling point based on the three-dimensional hydrogeological model of the site;

[0014] Step S22: Form a sequence from all sampling points in the permeability reference sampling point set, and evaluate the similarity of soil state between the current sampling point and any sampling point in the sequence;

[0015] Step S23: Obtain historical periodic monitoring data for each sampling point multiple times, and output a comprehensive anomaly disturbance degree by combining soil condition similarity.

[0016] Step S24: Filter all abnormal upstream sampling points from the sequence by comprehensively analyzing the abnormal perturbation degree;

[0017] Step S25: Analyze the historical monitoring data corresponding to the abnormal upstream sampling point to determine the degree of influence of the current sampling point on the upstream strong seepage diffusion.

[0018] Preferably, step S22 includes:

[0019] Step S221: Sort all sampling points in the infiltration reference sampling point set according to the groundwater flow direction and the elevation gradient of gas phase diffusion in the soil to be tested, forming a sequence;

[0020] Step S222: Analyze the physical matrix parameters of the current sampling point and any sampling point in the sequence to obtain the physical matrix similarity;

[0021] Step S223: Integrate the multidimensional organic matter concentration of each sampling point to establish an organic matter concentration feature vector, analyze the organic matter concentration feature vector of the current sampling point and any sampling point in the sequence, and obtain the similarity of organic matter components;

[0022] Step S224: By combining the similarity of physical matrix and the similarity of organic components, the soil state similarity between the current sampling point and any sampling point in the sequence is obtained.

[0023] Preferably, step S23 includes:

[0024] Step S231: Obtain historical periodic monitoring data for each sampling point multiple times, determine historical similarity, and combine soil condition similarity to obtain the fluctuation range of soil consistency;

[0025] Step S232: Set a dynamic threshold for organic matter diffusion, compare it with the fluctuation amplitude consistent with soil, and count the number of diffusion peaks in the sequence;

[0026] Step S233: Combine the fluctuation amplitude of soil consistency and the number of diffusion peaks to obtain the comprehensive abnormal disturbance degree of the current sampling point and any sampling point in the sequence.

[0027] Preferably, step S24 specifically includes:

[0028] A preset screening threshold is compared with the comprehensive abnormal disturbance degree. The current sampling point and any sampling point in the sequence are defined as having a strong abnormal correlation, and the comprehensive abnormal disturbance degree is greater than the screening threshold. All abnormal sampling points are counted, and all sampling points upstream of the current sampling point are selected from the sequence. All abnormal upstream sampling points are counted.

[0029] Preferably, step S25 includes:

[0030] Step S251: Analyze the historical multidimensional organic matter concentrations corresponding to the abnormal upstream sampling points to determine the diffusion degree of any abnormal upstream sampling point to the downstream at the current sampling point;

[0031] Step S252: Combine the number of abnormal sampling points and abnormal upstream sampling points, and with the diffusion degree of the abnormal upstream sampling points of the current sampling point to the downstream, determine the degree of influence of the upstream strong seepage diffusion on the current sampling point.

[0032] Preferably, step S3 includes:

[0033] Step S31: Normalize the physical matrix parameters and multidimensional organic matter concentrations corresponding to all sampling points, splice them to form a high-dimensional joint index vector set, and input it into the initialized neural network;

[0034] Step S32: Construct a penalty function based on the degree of diffusion influence, embed and update the neural network, and cluster the high-dimensional joint index vector set to divide the sampling points into background area, high true content area and high measurement error area;

[0035] Step S33: Sampling points in the background area and high true content area are stable, and the corresponding monitoring data are defined as organic pollutant concentration data; Sampling points in the high measurement error area change, and the monitoring data are corrected by the degree of diffusion influence, and the organic pollutant concentration data of the corresponding sampling points are output.

[0036] To solve the above-mentioned technical problems, the present invention also provides a soil organic pollutant content determination system, the system comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus, and the processor calls logical instructions in the memory to execute the soil organic pollutant content determination method described in any of the preceding claims.

[0037] The present invention has the following beneficial effects:

[0038] 1. A three-dimensional sampling grid is established based on the soil to be tested to monitor the physical matrix parameters and multidimensional organic matter concentrations at each sampling point. By constructing a permeability reference sampling point set, the monitoring data corresponding to the sampling points are analyzed to obtain the dynamic diffusion influence of the sampling points based on the permeability topology, i.e., the diffusion degree is quantified, and the comprehensive influence of upstream strong seepage diffusion on the sampling points is obtained. Finally, the influence degree is transformed into a penalty function and embedded into a neural network to distinguish all sampling points. This overcomes the defect of traditional static clustering that easily misidentifies passive plume areas as sources. It can intelligently remove free flow interference from sensor readings and accurately divide primary and secondary pollution areas. Based on this, dynamic concentration stripping compensation is performed on the initial monitoring data, and the high-fidelity true content after dynamic compensation is output, providing a reliable basis for accurate remediation and accurately restoring the true concentration data to output the organic pollutant concentration data of each sampling point. This effectively overcomes the false positive measurement error caused by the dynamic migration and diffusion of organic pollutants in groundwater and pores in portable devices.

[0039] 2. The soil organic pollutant content determination system provided by the present invention has the same beneficial effects as the soil organic pollutant content determination method provided by the present invention, and will not be described in detail here. Attached Figure Description

[0040] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating the steps of a method for determining the content of organic pollutants in soil, as provided in an embodiment of the present invention. Detailed Implementation

[0042] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method and system for determining the content of organic pollutants in soil according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0044] The following description, in conjunction with the accompanying drawings, details the specific scheme of the method and system for determining the content of organic pollutants in soil provided by the present invention.

[0045] To better illustrate this point, soil organic matter is one of the core indicators for measuring soil fertility. It has a significant impact on the physical properties of soil and is an important parameter for assessing the health of soil ecosystems and biodiversity. Measuring soil organic matter concentration is a fundamental and crucial test in the fields of soil science, agricultural production, and environmental protection. It helps to comprehensively understand the soil fertility status, ecological functions, and potential environmental risks. It is also of great significance to environmental protection and climate change research. In other words, accurate measurement of soil organic matter concentration not only provides a scientific basis for the rational use and protection of soil resources, but also forms the foundation for achieving sustainable soil use and green agricultural development.

[0046] Existing methods for determining and identifying soil organic matter concentration typically focus only on the total concentration at any given moment, neglecting the dynamic distribution of organic pollutants between the gas and solid phases. Furthermore, organic pollutants undergo natural degradation over time, failing to identify the dynamic evolution of soil conditions at historical sampling points, thus affecting the accuracy of organic matter concentration measurements. Specifically, the soil gas phase, or soil air, refers to the total gases present in soil pores, primarily including atmospheric oxygen and nitrogen, as well as small amounts of carbon dioxide, water vapor, and inert gases. The soil solid phase, or soil particle adsorption, refers to the adsorption of substances such as water, gases, ions, and organic molecules by the surface and internal pores of soil solid particles such as mineral and organic matter particles.

[0047] Therefore, a method for determining the content of soil organic pollutants and a system for determining the content of soil organic pollutants are proposed. When operating, the system requires the use of a method for determining the content of soil organic pollutants. Therefore, whether the system and program data are integrated or different hardware is configured to produce functions with similar effects to those achieved by this invention, they all fall within the protection scope of this invention.

[0048] Please see Figure 1 The diagram illustrates a flowchart of a method for determining the content of organic pollutants in soil according to an embodiment of the present invention, the method comprising:

[0049] Step S1: Establish a three-dimensional sampling grid using the soil to be tested, set multiple sampling points, and collect monitoring data on physical matrix parameters and multidimensional organic matter concentrations accordingly;

[0050] Step S2: Construct a permeation reference sampling point set corresponding to each sampling point, analyze the physical matrix parameters and multidimensional organic matter concentration of each sampling point in the current sampling point and the corresponding permeation reference sampling point set, and obtain the degree of dynamic diffusion influence of the sampling point based on the permeation topology;

[0051] Step S3: Integrate all sampling points to establish a high-dimensional joint index vector set and input it into the initialized neural network. Construct a penalty function based on the degree of diffusion influence and embed it into the neural network to update it. Use the updated neural network to distinguish all sampling points, determine whether the sampling points are stable or changing, and output the organic pollutant concentration data for each sampling point accordingly.

[0052] It is explained that while existing portable testing devices improve the timeliness of data acquisition to some extent, they affect the accuracy of organic pollutant content measurement in practical applications. Portable testing devices typically refer to small, lightweight, and easy-to-operate instruments that can be carried to the testing site by non-professionals or on-site personnel for rapid analysis. Therefore, implementing a method for determining soil organic pollutant content will yield more accurate organic matter content data.

[0053] Further, step S1 includes:

[0054] Step S11: Establish a three-dimensional sampling grid on the soil to be tested to obtain multiple sampling points.

[0055] Specifically, according to existing technical guidelines for soil pollution investigation, such as the Technical Guidelines for Soil Pollution Investigation of Construction Land, a three-dimensional sampling grid is established at the site to be tested. The intersections between the grids are used as sampling points to obtain multiple three-dimensional spatial sampling points. The number of sampling points is denoted as... It integrates all sampling points to form a sampling point set, and records the three-dimensional spatial coordinates of each sampling point based on a three-dimensional sampling grid, outputting the precise position of each sampling point.

[0056] Step S12: Perform in-situ and ex-situ sampling at each sampling point to obtain historical baseline data and determine physical matrix parameters; obtain the VOCs concentration in the gas phase of the soil through detection equipment and determine the SVOCs concentration in the solid phase of the soil.

[0057] To clarify, in-situ sampling refers to the process of directly sampling and preliminary analysis in the original environment where pollutants or target substances exist; ex-situ sampling is a method of collecting environmental media such as soil, water sediments, and rocks from sampling points and transporting them to laboratories or other specialized analytical sites for detailed measurement and analysis.

[0058] Specifically, historical baseline data is obtained through pre-set groundwater monitoring wells or soil gas monitoring probes. That is, monitoring equipment is pre-deployed in the soil to be tested to collect basic data in the soil in a long-term and systematic manner and measure physical matrix parameters. The concentration of VOCs in the soil gas phase is obtained using a portable detector in the field, and the concentration of SVOCs in the solid phase soil, such as benzene series compounds and polycyclic aromatic hydrocarbons, is measured by laboratory GC-MS (Gas Chromatography-Mass Spectrometry). The data corresponding to VOCs concentration and SVOCs concentration are integrated and recorded as multidimensional organic matter concentration.

[0059] Furthermore, physical matrix parameters include, but are not limited to, total organic carbon content, soil porosity, and permeability parameters.

[0060] It can be explained that total organic carbon (TOC) refers to the total carbon content in all organic matter in the soil, usually expressed as a percentage of the soil's dry weight; soil porosity refers to the percentage of pore volume in the soil to the total soil volume, used to reflect the soil's aeration, permeability, and water retention; and infiltration parameters are quantitative indicators describing the soil's ability to allow water or gas to pass through its pore structure, used to reflect the ease with which water or gas flows through the soil under gravity.

[0061] It should be further explained that although the preset groundwater monitoring wells or soil gas monitoring probes can collect real-time data, the TOC, soil porosity, and multidimensional concentration indicators are all historical baseline data. In other words, the physical matrix parameters of the soil being tested are the inherent basic properties of the soil, and their changes are usually relatively slow and stable. Historical baseline data can more realistically reflect the long-term average state of the soil, while real-time data is affected by instantaneous environmental factors, which makes it impossible to accurately represent the normal physical characteristics of the soil. The concentration data collected by the portable detector is the real-time data at the current moment. In the subsequent comparative calculations, the target pollutant dimensions shared by both the gaseous and solid phases of the soil will be aligned.

[0062] As explained, in subsequent related method analyses, the sampling point of the current analysis will be denoted as the [number]. The determination of relevant data for each sampling point is related to the first sampling point. The same applies to each sampling point.

[0063] Further, step S2 includes:

[0064] Step S21: Construct a set of permeability reference sampling points for each sampling point based on the three-dimensional hydrogeological model of the site.

[0065] Specifically, for the first At each sampling point, due to the migration of organic matter downstream or along pores, the local hydraulic gradient and permeability tensor are calculated based on the three-dimensional hydrogeological model of the site. The local hydraulic gradient reflects the driving force of groundwater flow, and the permeability tensor describes the differences in permeability of porous media in different directions. Then, from all sampling points, i.e. Select from the sampling points and the first The sampling points are located in the same hydraulic connectivity zone, meaning that there is a pore network between the locations of the two underground sampling points that can influence each other through water flow, or they are connected to the first sampling point. The sampling points were also located at other sampling points along high-permeability priority paths such as gravel belts and sewage pipelines, integrating data based on the first sampling point. All sampling points selected from the nth sampling point form the nth sampling point. A set of penetration reference sampling points for each sampling point.

[0066] Step S22: Form a sequence of all sampling points in the infiltration reference sampling point set, and evaluate the similarity of soil state between the current sampling point and any sampling point in the sequence.

[0067] It is explained that soil condition similarity includes two aspects: the degree of similarity of physical matrix and the degree of similarity of organic matter composition. It refers to the level of similarity or consistency between sampling points in terms of physical structure characteristics and organic matter composition for the soil to be tested, which provides a basis for subsequent analysis of the degree of diffusion influence.

[0068] Further, step S22 includes:

[0069] Step S221: Sort all sampling points in the infiltration reference sampling point set according to the groundwater flow direction and the elevation gradient of gaseous diffusion in the soil to be tested, forming a sequence; wherein, the groundwater flow direction is determined by the local hydraulic gradient obtained in step S21 above, and its direction reflects the main path of pollutant migration with groundwater; gaseous diffusion is affected by soil pore structure, temperature, pressure, and elevation changes, and the elevation gradient, i.e., the height change per unit distance, determines the vertical and horizontal components of gaseous pollutant diffusion; for the first... The penetration rate of each sampling point is used to generate a sequence by directing the sorting of all sampling points in the sampling point set, denoted as . Through sequence Potential pollution diffusion channels were established in physical space.

[0070] Step S222: Analyze the physical matrix parameters of the current sampling point and any sampling point in the sequence to obtain the physical matrix similarity.

[0071] It is explained that organic pollutants are affected by the total organic carbon content of the soil and dynamically distributed between the gas and solid phases. The total organic carbon in the soil acts as an important adsorption site and solvent, and its content directly determines the distribution coefficient of organic pollutants between the soil particle surface (solid phase) and soil pore water (liquid phase) or air (gas phase). The remaining physical matrix parameters also have a certain impact on the soil. Therefore, a comprehensive evaluation is conducted to improve the accuracy of the analysis.

[0072] Preferably, in this embodiment, a range standardization method is used for any physical matrix parameter extracted in step S1, including total organic carbon content, soil porosity, and permeability parameters, to map each indicator to a value range. This generates a dimensionless feature vector. For example, regarding total organic carbon content, the range is the difference between the maximum and minimum values ​​in the historical baseline data. During range standardization, if outliers occur during the real-time analysis where the total organic carbon content exceeds the maximum or falls below the minimum, truncation is performed. Normalized results exceeding the maximum are truncated to 1, and normalized results below the minimum are truncated to 0. Soil porosity and permeability parameters are processed similarly based on total organic carbon content.

[0073] Specifically, with the first A sequence is formed from each sampling point. The first in The analysis of each sampling point determines the degree of physical matrix similarity, and the corresponding calculation formula is as follows:

[0074]

[0075] in, Indicates the first Each sampling point and sequence The first in The degree of physical matrix similarity among the sampling points; The total number of dimensions representing the physical matrix parameters; Dimension index representing physical matrix parameters; , They represent the first Each sampling point and sequence The first in Physical matrix parameters of each sampling point; This represents absolute value operations.

[0076] It can be explained that the dimensional index of the physical matrix parameters These correspond to total organic carbon content, soil porosity, and permeability parameters, respectively. For example, the first dimension represents total organic carbon content, the second dimension represents soil porosity, etc. Therefore, in this embodiment, the total number of dimensions of the physical matrix parameters is... The dimension index and total number of dimensions can be adjusted according to the actual situation.

[0077] It should be noted that when the denominator in the calculation formula is 0, it means that all three physical matrix parameters (total organic carbon, porosity, and permeability parameters) are zero after normalization. This indicates that the physical properties of the current sampling point are equal to or lower than the minimum value of the historical baseline. In order to avoid data processing anomalies, data can be collected again. If this anomaly occurs multiple times (e.g., 3 times), it indicates that the historical baseline data may not be applicable to the current area or that there is a systemic failure in the instrument. This triggers a conservative risk warning, terminates the analysis, and prompts the user to check the historical baseline data or check the instrument.

[0078] Step S223: Integrate the multidimensional organic matter concentration of each sampling point to establish an organic matter concentration feature vector, analyze the organic matter concentration feature vector of the current sampling point and any sampling point in the sequence, and obtain the similarity of organic matter components.

[0079] It is explained that while the total concentration of pollutants in the soil decreases as they diffuse downstream, the proportion of organic components in the leak source often remains consistent, such as the ratio of benzene to toluene. This conforms to the general law of pollutant migration, that is, physical diffusion and biodegradation mainly affect the total amount of pollutants and have little impact on the relative proportions between components. Therefore, it is necessary to analyze the similarity of organic components at the two sampling points to facilitate subsequent verification of the continuous diffusion path of pollutants.

[0080] Specifically, based on the integration of VOCs concentration in the gas phase and SVOCs concentration in the solid phase of the soil corresponding to each sampling point, an organic matter concentration feature vector is formed for the corresponding sampling point. The similarity of organic matter components is then determined, and the corresponding calculation formula is as follows:

[0081]

[0082] in, Indicates the first Each sampling point and sequence The first in The degree of similarity of organic components at each sampling point; , They represent the first Each sampling point and sequence The first in Organic matter concentration feature vector of each sampling point; The L2 norm represents the eigenvector of organic matter concentration.

[0083] It can be explained that the L2 norm of the organic matter concentration eigenvector That is, the modulus of the vector is calculated to measure the overall load of organic matter concentration at the corresponding sampling point; in addition, based on the analysis of the soil to be tested, The calculated result is usually impossible to be zero. To ensure the stability of the calculation, theoretically, only in ideal pure soil that is completely inorganic and unaffected by any biological activity, where the concentration of all organic matter is zero, will the eigenvector of organic matter concentration be zero and the L2 norm be zero. However, such soil properties are almost non-existent in the natural environment. Even soil formed by the weathering of original rocks will contain trace amounts of simple organic matter produced by the decomposition of microbial residues or plant debris.

[0084] Step S224: By combining the similarity of physical matrix and the similarity of organic components, the soil state similarity between the current sampling point and any sampling point in the sequence is obtained.

[0085] Specifically, the formula for determining soil state similarity is as follows:

[0086]

[0087] in, Indicates the first Each sampling point and sequence The first in Soil condition similarity at each sampling point; Indicates the weighting coefficient; Indicates the first Each sampling point and sequence The first in The degree of physical matrix similarity among the sampling points; Indicates the first Each sampling point and sequence The first in The degree of similarity of organic components at each sampling point.

[0088] It can be noted that, in this embodiment, the weighting coefficient It can be specifically set according to the actual situation; since the physical matrix parameters and multidimensional organic matter concentrations collected and measured in the soil reflect two different physical evolution processes, that is, they have independent response characteristics in the soil formation and evolution process, the weight values ​​are determined by relying on the experience of environmental experts and the previous hydrogeological survey report, which more objectively reflects the similarity of soil state, takes into account both subjective judgment and objective cognition, and can avoid the limitations caused by subjective judgment, as well as the insufficient depth of the survey report in describing the dynamic process of soil micro-organic matter.

[0089] Step S23: Obtain historical periodic monitoring data for each sampling point multiple times, and output the comprehensive abnormal disturbance degree by combining soil condition similarity.

[0090] The explanation is as follows: the comprehensive abnormal disturbance degree is determined by combining historical periodic monitoring data with the current soil condition similarity to judge whether there is consistent fluctuation, and outputs the quantitative information of abnormal changes that exceed normal fluctuations.

[0091] Further, step S23 includes:

[0092] Step S231: Obtain historical periodic monitoring data for each sampling point multiple times, determine historical similarity, and combine soil condition similarity to obtain the fluctuation range of soil consistency;

[0093] Specifically, because organic matter has natural degradation characteristics, it is also necessary to combine historical data feedback to determine the change in similarity with the currently calculated soil state. Through a dynamic feedback mechanism of long-term historical data, the feedback of current data can be effectively calibrated to prevent overlooking the lag and cumulative effects of organic matter degradation and avoid errors in anomaly identification. The number of historical data collections is recorded as follows: Next, the historical sequence number is determined similarly based on the aforementioned steps S21-S22. Each sampling point and its corresponding sequence The first in The similarity of soil conditions at each sampling point is denoted as historical similarity. In this embodiment, the similarity is based on the historical similarity of the first sampling point. The analysis will be conducted in the next step. This represents the index of a historical sequence, indicating the first index in the sequence. The second Each sampling point and sequence The first in The historical similarity of each sampling point is denoted as Its similarity to soil conditions The absolute value of the difference represents the fluctuation range of soil consistency, i.e. , Indicates the first Each sampling point and sequence The first in Similarity of current soil conditions at each sampling point , and history Historical similarities The fluctuation range of soil consistency; This represents absolute value operations.

[0094] Step S232: Set a dynamic threshold for organic matter diffusion, compare it with the fluctuation amplitude consistent with soil, and count the number of diffusion peaks in the sequence.

[0095] Specifically, the dynamic threshold for organic matter diffusion is denoted as... , It can be specifically set according to the actual situation; in this embodiment, when the fluctuation range of soil consistency When, explain the first Each sampling point and sequence The first in The sampling points experienced dramatic organic matter migration during historical evolution, resulting in poor consistency in soil conditions. This indicates that the current analysis of the historical data point... A diffusion peak appeared, indicating an anomaly at the sampling point; conversely, a peak appeared at the sampling point at the next peak. At this point, the soil consistency is good, and there are no abnormal fluctuations. To avoid computational redundancy, the analysis can be omitted. Next, the statistics for the [number]th [item] are [calculated]. Sequence of sampling points The first in Each sampling point in history All of these values ​​are greater than the dynamic threshold for organic matter diffusion. The number of fluctuations in soil consistency, i.e., the number of statistical diffusion peaks, is denoted as... .

[0096] Step S233: Combine the fluctuation amplitude of soil consistency and the number of diffusion peaks to obtain the comprehensive abnormal disturbance degree of the current sampling point and any sampling point in the sequence.

[0097] Specifically, in the process of counting the number of diffusion peaks, a corresponding sequence was obtained. The first in The number of times an anomaly occurred at each sampling point was recorded as follows: The total number of anomalies was aggregated and recorded as... This represents the total number of anomalies; the comprehensive anomaly disturbance degree is determined by the following formula:

[0098]

[0099] in, Indicates the first Each sampling point and sequence The first in The overall abnormal disturbance degree of each sampling point; This indicates the total number of historical sampling and monitoring sessions; Indicates the first Each sampling point and sequence The first in The number of diffusion peaks that have occurred at each sampling point in history; Represents the normalization function; The index indicating the subindex where the exception occurred; Indicates the total number of times an anomaly occurred; Indicates the first When the anomaly occurs for the first time Each sampling point and sequence The first in The fluctuation range of soil consistency at each sampling point.

[0100] Preferably, in this embodiment, the normalization function Max-min normalization is performed using the maximum and minimum values. The dataset is a set of data representing the fluctuation range of soil consistency between the sampled points and the sampled points in the sequence when global anomalies occur. The maximum and minimum values ​​in the dataset are used for normalization to eliminate the influence of units and make different features comparable. If the range is 0, the default output normalization result is 1.

[0101] Step S24: Filter all abnormal upstream sampling points from the sequence by comprehensively analyzing the abnormal perturbation degree.

[0102] An explanation is provided based on the aforementioned steps, combined with the first... Each sampling point and sequence The first in Similarly, the sampling point is determined to be the first... Each sampling point and its corresponding sequence The overall abnormal disturbance degree of each sampling point.

[0103] Further, in step S24, specifically:

[0104] A preset screening threshold is compared with the comprehensive abnormal disturbance degree. The current sampling point and any sampling point in the sequence are defined as having a strong abnormal correlation, and the comprehensive abnormal disturbance degree is greater than the screening threshold. All abnormal sampling points are counted, and all sampling points upstream of the current sampling point are selected from the sequence. All abnormal upstream sampling points are counted.

[0105] Specifically, the screening threshold is denoted as , and the overall anomaly disturbance degree When a comparison is performed, When, it indicates the first Each sampling point and sequence The first in There is a strong correlation between the sampling points; conversely, it indicates that the anomalies between the two points are not caused by mutual influence, but may be due to other unrelated reasons leading to misjudgment. Based on this screening mechanism, the same comparison is made with the sampling points... The combined anomalous perturbation degree of each sampling point and the remaining sampling points in the sequence is used to filter out all sampling points with strong correlation anomalies, and the number is denoted as . Sampling points exhibiting strong correlation anomalies are defined as anomalous sampling points; then, from... From the nth abnormal sampling point, the one located at the nth All sampling points upstream of each sampling point, totaling [number] One abnormal upstream sampling point; preferably, in this embodiment, It can be configured according to the actual situation.

[0106] Step S25: Analyze the historical monitoring data corresponding to the abnormal upstream sampling point to determine the degree of influence of the current sampling point on the upstream strong seepage diffusion.

[0107] The explanation is that the degree of influence of upstream strong seepage diffusion on the sampling point refers to the quantitative or qualitative description of the impact of pollutants carried by strong seepage, such as rapid rise in groundwater level caused by heavy rain, groundwater backflow caused by artificial pumping, or strong infiltration of surface runoff, on the monitoring data of downstream sampling points through the flow and diffusion process of groundwater or surface water, under the current soil conditions being analyzed.

[0108] Further, step S25 includes:

[0109] Step S251: Analyze the historical multidimensional organic matter concentrations corresponding to the abnormal upstream sampling points to determine the diffusion degree of any abnormal upstream sampling point to the downstream.

[0110] An explanation is provided based on the first For an abnormal upstream sampling point identified by a sampling point, the faster the concentration of organic matter surges and the larger the leakage, the greater the driving force pushing it outward, and thus the greater the impact on downstream sampling points; therefore, the degree of diffusion from upstream to downstream is determined by analyzing the changes in the concentration of organic matter.

[0111] Specifically, taking the first of the abnormal upstream sampling points Each sampling point is used to explain the extent of diffusion of the anomaly from the upstream sampling point to the downstream. The corresponding calculation formula is as follows:

[0112]

[0113] in, Indicates the first The corresponding sampling point of the th sampling point The extent of diffusion of an abnormal upstream sampling point to the downstream; Represents the normalization function; Indicates the quantity of different types of organic matter; Indicates an index of organic species; Indicates the first The mobility coefficient of organic matter; Indicates the first The first abnormal upstream sampling point The rising slope of the historical concentration curve of a certain organic compound; Indicates the first The first abnormal upstream sampling point The historical average concentration of various organic compounds.

[0114] Preferably, the normalization function We still use the maximum and minimum values ​​for max-min normalization, that is, for the th... The reference set constructed from the organic matter corresponding to the sampling points upstream of the anomalies is normalized. If the elements of the set are unique or the range is 0, the normalization result is recorded as 1.

[0115] It can be noted that the organic matter type refers to all organic matter in the multidimensional organic matter concentration obtained in step S1 above, namely the VOCs concentration in the soil gas phase and the SVOCs concentration in the solid phase soil, which usually refers to organic matter categories such as benzene compounds, pesticides, polycyclic aromatic hydrocarbons, and phthalate plasticizers; mobility coefficient In existing technologies, soil column leaching experiments or batch adsorption-desorption experiments are used, combined with theoretical calculations, to reflect the migration capacity of organic matter in soil; the slope of the historical concentration curve is also used. It is obtained from the rising phase of the concentration curve plotted based on historical monitoring data of organic matter. (The rising slope is the slope obtained by linearly fitting the rising phase. It should be understood that if there are multiple rising phases, the average of the slopes of each rising phase can be used as the rising slope.) In particular, in actual operation, if the concentration curve shows a downward or stable trend globally, without a rising slope, it will... It is recorded as 0; in addition, depending on the location of the abnormal upstream sampling point in the soil, the upward slope of the historical concentration curve of organic matter or the average historical concentration of organic matter will change with infiltration and diffusion, so as to more accurately reflect the degree of diffusion of the abnormal upstream sampling point to the downstream through organic matter concentration related data.

[0116] Step S252: Combine the number of abnormal sampling points and abnormal upstream sampling points, and with the diffusion degree of the abnormal upstream sampling points of the current sampling point to the downstream, determine the degree of influence of the upstream strong seepage diffusion on the current sampling point.

[0117] Specifically, the formula for determining the degree of influence of upstream strong seepage diffusion on the current sampling point is as follows:

[0118]

[0119] in, Indicates the first The degree to which each sampling point is affected by the upstream strong seepage diffusion; Indicates based on the first The number of abnormal upstream sampling points selected by each sampling point; Indicates the index of the upstream sampling point of the anomaly; Indicates based on the first The number of abnormal sampling points selected from each sampling point; Indicates the first The corresponding sampling point of the th sampling point The extent of diffusion of an abnormal upstream sampling point to the downstream.

[0120] It can be explained that, This indicates the percentage of abnormal upstream sampling points. The smaller the value, the more likely the abnormal sampling points are in the first position. Downstream of the sampling point, indicating the... Each sampling point itself is the source, indicating that the point is less affected by upstream strong seepage diffusion; conversely, the greater the impact, the more significant the impact. Therefore, the degree of diffusion downstream from an abnormal upstream sampling point is significant. The larger the value, the more upstream sampling points are abnormal. More indicates the number of times. The higher the degree to which organic pollutants at each sampling point are passively transported through severe upstream leaks and groundwater or soil pores, the better.

[0121] It should be further noted that, based on the actual analysis of the soil being tested, due to factors such as natural degradation or strong seepage diffusion, the sequence corresponding to the potential pollution diffusion channels formed based on any sampling point will inevitably contain anomalous sampling points, i.e., the number of anomalous sampling points... It is almost impossible for the value to approach 0; however, during the analysis, there may be no abnormal upstream sampling points upstream of each currently analyzed sampling point. That is, when the sampling point itself is a source of pollution, the number of abnormal upstream sampling points is... At this point, the degree of influence of the upstream strong seepage diffusion on the sampling point is directly determined. Skip the formula calculation.

[0122] As an alternative implementation method, the neural network adopts SOM (Self-organizing feature map) neural network, which is an unsupervised learning neural network model that can map high-dimensional input data to a low-dimensional output space while preserving the topological structure and statistical properties of the input data, thereby realizing data visualization, clustering and feature extraction.

[0123] Further, step S3 includes:

[0124] Step S31: Normalize the physical matrix parameters and multidimensional organic matter concentrations corresponding to all sampling points, splice them into a high-dimensional joint index vector set, and input it into the initialized neural network; that is, in this embodiment, the monitoring data corresponding to all sampling points are normalized to eliminate the influence of dimensional differences and numerical ranges on subsequent neural network training, and spliced ​​into a high-dimensional joint index vector set to characterize the comprehensive information of the environmental characteristics of the sampling points, which can more realistically reflect the comprehensive state of media such as soil or water; input it into the initialized SOM neural network to facilitate subsequent data clustering processing.

[0125] Step S32: Construct a penalty function based on the degree of diffusion influence, embed and update the neural network, and cluster the high-dimensional joint index vector set to divide the sampling points into background area, high true content area and high measurement error area.

[0126] To clarify, in traditional SOM (Search Engine Analysis), all sampling points have the same traction weight for acquiring neurons. However, this is fatal in organic pollution identification. Because the migration and transformation of pollutants are influenced by multiple factors such as hydrogeological conditions, water flow velocity, and diffusion coefficient, they exhibit significant spatial heterogeneity and dynamic changes. If a consistent traction weight is still used, high-concentration plume areas of passive diffusion will be mistakenly identified as cluster centers, thus masking truly bio- or chemically significant pollution hotspots. Therefore, the degree of influence of upstream strong seepage diffusion on sampling points calculated using the aforementioned steps is crucial. A custom penalty function is used in the SOM neural network to adjust the weights, thereby more realistically reflecting the spatial distribution characteristics of organic pollution and the location of potential pollution sources.

[0127] Specifically, construct the penalty function, i.e. , Indicates the first The penalty factor corresponding to each sampling point is embedded as a multiplication coefficient in the learning rate update function of the SOM neural network, representing the neuron and its neighboring nodes. That is, the actual update learning rate for each sampling point is equal to the initial learning rate assigned by the neural network plus the penalty factor. The product of; in actual operation, if the first The degree of influence of upstream strong seepage diffusion on each sampling point The maximum value indicates that the sampling point is passively affected by the diffusion of upstream pollution, and the corresponding penalty factor is... The size will be drastically reduced to significantly suppress the weight update magnitude of that sampling point in the neural network; conversely, if the organic matter concentration at a sampling point is extremely high and the influence of strong upstream seepage diffusion is significant... If the value is extremely small, it indicates that the source is unaffected by other sampling points and is considered an absolute leakage source. Therefore, the corresponding penalty factor... Increase its size, thus granting it the power to dominate the updating of the neural network topology.

[0128] Next, after iteratively running the SOM neural network embedded with the constructed penalty function, all sampling points are clustered. After the clustering converges, multiple clusters are obtained, and the influence of upstream strong seepage diffusion on all sampling points in each cluster is determined. Determine the average influence level corresponding to each cluster, denoted as . ; First and second thresholds are preset respectively, which can be specifically set based on relevant data obtained from soil environmental quality standards and traditional measurement methods; average impact degree Clusters below the first threshold are labeled as background areas and high true content areas, while clusters above the second threshold are labeled as high measurement error areas.

[0129] It should be noted that the average degree of impact Clustering within the range of the first and second thresholds indicates that there may be errors, but the errors are small and can be ignored. The measured monitoring data can be used.

[0130] Step S33: Sampling points in the background area and high true content area are stable, and the corresponding monitoring data are defined as organic pollutant concentration data; Sampling points in the high measurement error area change, and the monitoring data are corrected by the degree of diffusion influence, and the organic pollutant concentration data of the corresponding sampling points are output.

[0131] Specifically, for the background zone and the high true concentration zone, since there is no upstream flow interference and physical adsorption is stable, all sampling points in this zone are in a stable state. Therefore, the initial measured concentration of the portable measuring device is directly determined, meaning the multidimensional organic matter concentration corresponding to each sampling point is the true organic pollutant concentration data. For the high measurement error zone, where the apparent high concentration includes a large amount of temporarily flowing free gas / liquid, measurement compensation is performed to determine the compensated measured concentration value. The corresponding calculation formula is: , Indicates the first The measured content values ​​after compensation at each sampling point; Indicates the first Initial measured content values ​​at each sampling point; This represents the compensation limit coefficient determined based on the soil phase carrying capacity of the site. Represents the normalization function; Indicates the first The degree of influence of upstream strong seepage diffusion on each sampling point.

[0132] Preferably, the normalization function Max-min normalization is performed using the maximum and minimum values; this indicates that the compensated measured content value reflects the degree of influence of upstream strong seepage diffusion on the sampling point. The initial measured content values ​​were corrected, and the initial multidimensional organic matter concentration was optimized to obtain the true measured content values ​​for each sampling point in the high measurement error zone, so as to output the organic pollutant concentration data of the sampling points in this area measured by portable measuring equipment; the compensation limit coefficient was determined based on the soil phase carrying capacity of the site. This refers to a quantitative index determined by assessing the composition of different phases in the soil and their impact on the pollutant carrying capacity, with a corresponding value range of [missing value]. In this embodiment, It can be adjusted according to the actual situation.

[0133] Understandably, a three-dimensional sampling grid is established based on the soil to be tested to monitor the physical matrix parameters and multidimensional organic matter concentrations at each sampling point. By constructing a permeability reference sampling point set, the monitoring data corresponding to the sampling points are analyzed to obtain the degree of dynamic diffusion influence of the sampling points based on permeability topology, i.e., the degree of diffusion is quantified, and the comprehensive influence of upstream strong seepage diffusion on the sampling points is obtained. Finally, the degree of influence is transformed into a penalty function and embedded into a neural network to distinguish all sampling points. This overcomes the defect of traditional static clustering that easily misidentifies passive plume areas as sources. It can intelligently remove free flow interference from sensor readings and accurately divide primary and secondary pollution areas. Based on this, dynamic concentration stripping compensation is performed on the initial monitoring data, and the high-fidelity true content after dynamic compensation is output, providing a reliable basis for accurate remediation and accurately restoring the true concentration data to output the organic pollutant concentration data of each sampling point. This effectively overcomes the false positive measurement error caused by the dynamic migration and diffusion of organic pollutants in groundwater and pores in portable devices.

[0134] The second embodiment of the present invention provides a soil organic pollutant content determination system. The system includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The processor calls logical instructions in the memory to execute a soil organic pollutant content determination method as described in any of the foregoing embodiments.

[0135] It can be noted that in this system, the various components work together to perform a method for determining the content of soil organic pollutants as described in any embodiment of the present invention, which has the same beneficial effects as the aforementioned method for determining the content of soil organic pollutants, and will not be elaborated here.

[0136] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0137] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for determining the content of organic pollutants in soil, characterized in that, The method includes: Step S1: Establish a three-dimensional sampling grid using the soil to be tested, set multiple sampling points, and collect monitoring data on physical matrix parameters and multidimensional organic matter concentrations accordingly; Step S2: Construct a permeation reference sampling point set corresponding to each sampling point, analyze the physical matrix parameters and multidimensional organic matter concentration of each sampling point in the current sampling point and the corresponding permeation reference sampling point set, and obtain the degree of dynamic diffusion influence of the sampling point based on the permeation topology; Step S3: Integrate all sampling points to establish a high-dimensional joint index vector set and input it into the initialized neural network. Construct a penalty function based on the degree of diffusion influence and embed it into the neural network to update it. Use the updated neural network to distinguish all sampling points, determine whether the sampling points are stable or changing, and output the organic pollutant concentration data for each sampling point accordingly.

2. The method for determining the content of organic pollutants in soil according to claim 1, characterized in that, Step S1 includes: Step S11: Establish a three-dimensional sampling grid on the soil to be tested to obtain multiple sampling points; Step S12: Perform in-situ and ex-situ sampling at each sampling point to obtain historical baseline data and determine physical matrix parameters; obtain the VOCs concentration in the gas phase of the soil through detection equipment and determine the SVOCs concentration in the solid phase of the soil.

3. The method for determining the content of organic pollutants in soil according to claim 2, characterized in that, The physical matrix parameters include, but are not limited to, total organic carbon content, soil porosity, and permeability parameters.

4. The method for determining the content of organic pollutants in soil according to claim 1, characterized in that, Step S2 includes: Step S21: Construct a set of permeability reference sampling points for each sampling point based on the three-dimensional hydrogeological model of the site; Step S22: Form a sequence from all sampling points in the infiltration reference sampling point set, and evaluate the similarity of soil state between the current sampling point and any sampling point in the sequence; Step S23: Obtain historical periodic monitoring data for each sampling point multiple times, and output a comprehensive anomaly disturbance degree by combining soil condition similarity. Step S24: Filter all abnormal upstream sampling points from the sequence by comprehensively analyzing the abnormal perturbation degree; Step S25: Analyze the historical monitoring data corresponding to the abnormal upstream sampling point to determine the degree of influence of the current sampling point on the upstream strong seepage diffusion.

5. The method for determining the content of organic pollutants in soil according to claim 4, characterized in that, Step S22 includes: Step S221: Sort all sampling points in the infiltration reference sampling point set according to the groundwater flow direction and the elevation gradient of gas phase diffusion in the soil to be tested, forming a sequence; Step S222: Analyze the physical matrix parameters of the current sampling point and any sampling point in the sequence to obtain the physical matrix similarity; Step S223: Integrate the multidimensional organic matter concentration of each sampling point to establish an organic matter concentration feature vector, analyze the organic matter concentration feature vector of the current sampling point and any sampling point in the sequence, and obtain the similarity of organic matter components; Step S224: By combining the similarity of physical matrix and the similarity of organic components, the soil state similarity between the current sampling point and any sampling point in the sequence is obtained.

6. The method for determining the content of organic pollutants in soil according to claim 4, characterized in that, Step S23 includes: Step S231: Obtain historical periodic monitoring data for each sampling point multiple times, determine historical similarity, and combine soil condition similarity to obtain the fluctuation range of soil consistency; Step S232: Set a dynamic threshold for organic matter diffusion and compare it with the fluctuation amplitude consistent with soil, and count the number of diffusion peaks from the sequence; Step S233: Combine the fluctuation amplitude of soil consistency and the number of diffusion peaks to obtain the comprehensive abnormal disturbance degree of the current sampling point and any sampling point in the sequence.

7. The method for determining the content of organic pollutants in soil according to claim 4, characterized in that, In step S24, specifically: A preset screening threshold is compared with the comprehensive abnormal disturbance degree. The current sampling point and any sampling point in the sequence are defined as having a strong abnormal correlation, and the comprehensive abnormal disturbance degree is greater than the screening threshold. All abnormal sampling points are counted, and all sampling points upstream of the current sampling point are selected from the sequence. All abnormal upstream sampling points are counted.

8. The method for determining the content of organic pollutants in soil according to claim 7, characterized in that, Step S25 includes: Step S251: Analyze the historical multidimensional organic matter concentrations corresponding to the abnormal upstream sampling points to determine the diffusion degree of any abnormal upstream sampling point to the downstream at the current sampling point; Step S252: Combine the number of abnormal sampling points and abnormal upstream sampling points, and with the diffusion degree of the abnormal upstream sampling points of the current sampling point to the downstream, determine the degree of influence of the upstream strong seepage diffusion on the current sampling point.

9. The method for determining the content of organic pollutants in soil according to claim 1, characterized in that, Step S3 includes: Step S31: Normalize the physical matrix parameters and multidimensional organic matter concentrations corresponding to all sampling points, splice them to form a high-dimensional joint index vector set, and input it into the initialized neural network; Step S32: Construct a penalty function based on the degree of diffusion influence, embed and update the neural network, and cluster the high-dimensional joint index vector set to divide the sampling points into background area, high true content area and high measurement error area; Step S33: Sampling points in the background area and high true content area are stable, and the corresponding monitoring data are defined as organic pollutant concentration data; Sampling points in the high measurement error area change, and the monitoring data are corrected by the degree of diffusion influence, and the organic pollutant concentration data of the corresponding sampling points are output.

10. A system for determining the content of organic pollutants in soil, characterized in that, The system includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The processor calls logical instructions in the memory to execute the method for determining the content of soil organic pollutants as described in any one of claims 1 to 9.

Images