Groundwater pollution source identification method and device, computer equipment and storage medium

Abstract

This application provides a method, apparatus, computer equipment, and storage medium for identifying groundwater pollution sources, relating to the field of water environment technology, and is used to improve the accuracy of groundwater pollution source identification. The method mainly includes: acquiring sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data; calculating the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vectors; performing a weighted calculation on the Euclidean distance and clustering distance to determine the winning neuron; updating the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and output neuron weight vector of the winning neuron; and when the number of updates to the output neuron weight vector of the water pollution neural network reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

Description

Technical Field

[0001] This application relates to the field of water environment technology, and in particular to a method, apparatus, computer equipment and storage medium for identifying groundwater pollution sources. Background Technology

[0002] Because groundwater has a wide variety of indicators and a very large sample size, methods such as dimensionality reduction and clustering can be used to process the data. Principal component analysis (PCA) and factor analysis (FA) are commonly used methods, but they are linear methods. When processing environmental data, the nonlinear relationships between explanatory variables often have limitations, resulting in a low accuracy rate in identifying groundwater pollution sources. Summary of the Invention

[0003] This application provides a method, apparatus, computer equipment, and storage medium for identifying groundwater pollution sources, which can improve the accuracy of groundwater pollution source identification.

[0004] This invention provides a method for identifying groundwater pollution sources, the method comprising:

[0005] Obtain sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data;

[0006] Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector;

[0007] The winning neuron is determined by weighted calculation of the Euclidean distance and the clustering distance.

[0008] The output neuron weight vector of the water pollution neural network is updated based on the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0009] When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

[0010] This invention provides a groundwater pollution source identification device, the device comprising:

[0011] The acquisition module is used to acquire sample data for groundwater pollution source detection. The sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data.

[0012] The calculation module is used to calculate the Euclidean distance and clustering distance between the sample data and the weight vector of the corresponding output neuron;

[0013] The calculation module is also used to perform a weighted calculation on the Euclidean distance and the clustering distance to determine the winning neuron;

[0014] The update module is used to update the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0015] The determination module is used to determine the groundwater pollution source in the target area when the number of times the output neuron weight vector of the water pollution neural network is updated reaches a preset value.

[0016] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described groundwater pollution source identification method.

[0017] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for identifying groundwater pollution sources.

[0018] A computer program product comprising a computer program that, when executed by a processor, implements the aforementioned groundwater pollution source identification method.

[0019] This invention provides a method, apparatus, computer equipment, and storage medium for identifying groundwater pollution sources. First, sample data for groundwater pollution source detection is acquired. This sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data. Then, the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vectors are calculated. The Euclidean distance and clustering distance are weighted to determine the winning neuron. Finally, the output neuron weight vector of the water pollution neural network is updated based on the input and output neuron weight vectors of the winning neuron. When the number of updates to the output neuron weight vector of the water pollution neural network reaches a preset value, the finally updated water pollution neural network determines the groundwater pollution source in the target area. Because this invention updates the output neuron weight vector of the water pollution neural network based on both Euclidean distance and clustering distance, the training of the water pollution neural network is more accurate, thereby improving the accuracy of groundwater pollution source identification through the water pollution neural network. Attached Figure Description

[0020] Figure 1 A flowchart of a groundwater pollution source identification method provided in this application;

[0021] Figure 2Mapping feature map of multidimensional data indicators in groundwater provided in this application;

[0022] Figure 3 Clustering results of different locations in SOM provided for this application;

[0023] Figure 4 A schematic diagram of the groundwater pollution source identification device provided in this application;

[0024] Figure 5 A schematic diagram of the computer device provided in this application. Detailed Implementation

[0025] To better understand the above technical solutions, the technical solutions of the embodiments of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of this application and the specific features in the embodiments are detailed descriptions of the technical solutions of the embodiments of this application, rather than limitations on the technical solutions of this application. In the absence of conflict, the embodiments of this application and the technical features in the embodiments can be combined with each other.

[0026] Please see Figure 1 As shown, this is an embodiment of the groundwater pollution source identification method provided by the present invention, used to perform steps S101-S105:

[0027] Step S101: Obtain sample data for groundwater pollution source detection. The sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data.

[0028] The sample data from groundwater pollution source detection were preprocessed. Data-related information was categorized by characteristics. Pollutant concentrations and latitude / longitude coordinates were directly represented using the original data, and the data was standardized using the `data_struct` and `normalize` commands provided by the `somtoolbox` in MATLAB. Data-related information, such as pollution source conditions and surface water system conditions, was reasonably numbered or digitized based on actual conditions. Surface water systems were represented using binary encoding to indicate their presence (1 for presence, 0 for absence), and subsequent binary encodings followed the same principle. Flow direction was set using planar coordinate vectors, with the location of the data point as the origin and the flow angle set accordingly. Pollution sources were categorized into industrial pollution sources, mining areas, landfills, gas stations, agricultural pollution sources, and surface wastewater, using binary encoding.

[0029] The monitoring indicators include pH, total hardness (calculated as CaCO3), sulfate, chloride, permanganate index, nitrate, nitrite nitrogen, ammonia nitrogen, fluoride, cyanide, manganese, cadmium, lead, iron, hexavalent chromium, zinc, and total dissolved solids. Table 1 shows the statistical characteristics of the site monitoring indicator data, namely the concentration data of water chemical indicators.

[0030] Table 1

[0031]

[0032]

[0033] Based on the collected data, the monitoring indicators and latitude and longitude data of the monitoring points directly use the original data. The surface water system and enterprise types in the study area are represented by binary codes to indicate their existence. Industrial pollution sources, mining areas, landfills, gas stations, agricultural pollution sources, and surface water system conditions are distinguished according to "1 for existence and 0 for non-existence" (Table 2). That is, the data in Table 2 are the data labels corresponding to the sample data. In other words, the water pollution sources corresponding to the sample data can be industrial pollution sources, mining areas, landfills, gas stations, agricultural pollution sources, etc. This embodiment does not make specific limitations on this.

[0034] Table 2

[0035]

[0036] Step S102: Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector.

[0037] In this embodiment, based on the principle of self-organizing map (SOM), it is assumed that the sample data is n-dimensional, and the weight vector between input unit i and neuron j in the computational layer is W = {Wi,j,j=1,...N}, where N is the number of sample data. Further, the SOM output size, i.e., the SOM grid size, is constructed. The size of the two-dimensional SOM grid is determined by m*n. To obtain a suitable mn value, quantization error (QE) and topology graph error (TE) need to be considered. This process requires iteratively calculating the QE and TE values ​​for each SOM size. By setting different grid sizes, QE and TE are used to evaluate the performance of the SOM topology structure with different grid sizes. Different sizes can also be set according to specific needs, typically set to [size missing]. After saving the results of QE and TE, plotting instructions are written to obtain the SOM output model. The self-organizing learning process of the self-organizing map neural network can be summarized into the following sub-processes.

[0038] (1) Initialize the SOM neural network. Assign values ​​to the initial neurons of the output layer using weight vectors.

[0039] (2) Input the processed groundwater-related data table into the input layer. Use the `data` function to sequentially set the indicator data items, indicator name items, and total items. By scanning the weight vectors corresponding to the indicator data items and the neurons in the output layer, calculate the sum of squares of the differences between the connection weights and the corresponding input values. For each data point in the indicator data item, find the output layer neuron with the closest Euclidean distance; this output layer neuron is determined as the winning neuron. The winning neuron obtained through competition is the result of dimensionality reduction of the groundwater-related data obtained by SOM in one training iteration. This result is visually displayed in the output layer neurons. The calculation formula for the competition process is shown below.

[0040] In an optional embodiment of the present invention, calculating the Euclidean distance between the sample data and the corresponding output neuron weight vector includes:

[0041] The Euclidean distance between the sample data and the output neuron is calculated using the following formula:

[0042]

[0043] Where d(x) j Let w be the Euclidean distance between the j-th sample data and the output neuron, where j∈[1,N] and N is the number of sample data. ji To output the neuron weight vector, x ji Let n be the input neuron weight vector corresponding to the i-th data dimension in the j-th sample data, where n is the data dimension of the sample data.

[0044] In an optional embodiment of the present invention, calculating the clustering distance between the sample data and the corresponding output neuron weight vector includes:

[0045] The clustering distance between the sample data and the output neuron is calculated using the following formula:

[0046] dist(x j ,w j )=||x j -w j ||

[0047] Where, dist(x) j ,w j ) represents the clustering distance between the sample data and the output neuron, j∈[1,N], where N is the number of sample data, x j For the j-th sample data, w ji Let j be the j-th output neuron.

[0048] In this embodiment, the clustering distance between the sample data and the output neuron is calculated based on the CURE algorithm, and then the convergence efficiency of the SOM result is optimized. Initially, each point in the CURE algorithm is a cluster. Samples are randomly selected from the initial clustering data, and the clustering distance between the sample data and the output neuron is calculated. Random sampling removes isolated points. A convergence coefficient α (which can be set to 0.2) is defined to move the sample points towards the centroid O by a fixed proportion. The locally generated clusters are further clustered, and the calculation is iterative. Finally, the clustering results are labeled, and the clustering results will spatially approximate the overall shape of the groundwater multidimensional data.

[0049] Step S103: Perform a weighted calculation on the Euclidean distance and the clustering distance to determine the winning neuron.

[0050] In an optional embodiment of the present invention, the step of determining the winning neuron by weighted calculation of the Euclidean distance and the clustering distance includes: obtaining the minimum Euclidean distance and the minimum clustering distance; and determining the winning neuron by weighted calculation of the minimum Euclidean distance and the minimum clustering distance.

[0051] Step S104: Update the output neuron weight vector of the water pollution neural network according to the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0052] In an optional embodiment of the present invention, updating the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and the output neuron weight vector of the winning neuron includes: obtaining the influence range of the winning neuron; and updating the output neuron weight vector of the water pollution neural network based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector.

[0053] The influence range of the winning neuron is calculated using the following formula:

[0054]

[0055] Where σ(t) is the influence range of the winning neuron, σ0 is the initial influence range of the winning neuron, t is the time length, and τ0 is a fixed decay coefficient.

[0056] The output neuron weight vector of the water pollution neural network is updated using the following formula:

[0057] Δw p =η(t)*T(t)*(x p -w p )

[0058]

[0059]

[0060] Where, Δw p The output neuron weight vector of the updated water pollution neural network is given, where η(t) is the learning coefficient that decays over time, T(t) is the neighborhood influence coefficient, η0 is the initial value of the learning coefficient, and x... p w is the input neuron weight vector of the winning neuron. p Let d be the output neuron weight vector of the winning neuron. pq τ is the Euclidean distance between the winning neuron and its nearest neighbor neurons. η The learning coefficient is fixed, and σ(t) represents the influence range of the winning neuron.

[0061] Step S105: When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

[0062] Repeat the above calculations and iterate multiple times to obtain a trained water pollution neural network. Then, based on the trained water pollution neural network, the groundwater pollution sources in the target area are determined. Specifically, the water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data of the target area are used as input neuron weight vectors and input into the trained water pollution neural network to obtain the groundwater pollution sources corresponding to the target area. The specific types of groundwater pollution sources are the pollution types in Table 2.

[0063] In this embodiment, a water pollution neural network (SOM-CURE model) is constructed based on multidimensional data. This fully considers the factors that may affect the source-sink relationship in the study area, explores the spatial distribution relationship and patterns between points, and comprehensively analyzes the possible situations affecting groundwater pollution. This provides certain theoretical support and practical value for the diagnosis of groundwater pollution source-sink relationship and pollution prevention and control.

[0064] This invention provides a method for identifying groundwater pollution sources. First, sample data for groundwater pollution source detection is acquired. This sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data. Then, the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vectors are calculated. The Euclidean distance and clustering distance are weighted to determine the winning neuron. Finally, the output neuron weight vector of the water pollution neural network is updated based on the input and output neuron weight vectors of the winning neuron. When the number of updates to the output neuron weight vector of the water pollution neural network reaches a preset value, the finally updated water pollution neural network determines the groundwater pollution source in the target area. Because this invention updates the output neuron weight vector of the water pollution neural network based on both Euclidean distance and clustering distance, the training of the water pollution neural network is more accurate, thereby improving the accuracy of groundwater pollution source identification through the water pollution neural network.

[0065] In one application scenario provided by an embodiment of the present invention, the calculation result obtained based on the groundwater pollution source identification method provided in this embodiment is as follows: Figure 2 and Figure 3 As shown, the points D21, D23, D18, D17, and D24 in the lower right corner represent high response intensities of nitrate, nitrite, ammonia nitrogen, potassium permanganate, arsenic, and manganese in the characteristic spectrum; medium intensity of sulfate and chloride; and relatively low levels of fluoride, cadmium, iron, and chromium. The groundwater pollution sources at these points are mainly industrial pollution sources and gas stations, and are also affected by agricultural pollution sources, but are almost unaffected by mining pollution.

[0066] The locations D16, D19, and D20 in the lower left corner are areas in the study area significantly affected by mining operations. Higher lead levels were concentrated at these locations, along with higher levels of arsenic and manganese. Furthermore, these locations were also affected by agricultural and industrial pollution sources, resulting in relatively high levels of nitrates and chlorides.

[0067] The overall response at points D06, D10, and D27 above was low, mainly due to industrial pollution sources and water-rock interactions. Compared to the points clustered below, most pollution indicators were detected at the points above, indicating a smaller impact from industrial pollution.

[0068] This method can uncover potential relationships between monitoring data and data that are difficult to quantify, such as latitude and longitude coordinates, surface water system conditions, and enterprise types. It can obtain multidimensional data reduction and clustering results that include pollution index data. The high correlation between pollutant indicators indicates that they may have similar source characteristics. The high correlation between monitoring indicators and pollution sources can, to some extent, explain the source of pollution.

[0069] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0070] In one embodiment, a groundwater pollution source identification device is provided, which corresponds one-to-one with the groundwater pollution source identification method described in the above embodiments. For example... Figure 4 As shown, the functional modules of the device are described in detail below:

[0071] The acquisition module 21 is used to acquire sample data for groundwater pollution source detection. The sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data.

[0072] Calculation module 22 is used to calculate the Euclidean distance and clustering distance between the sample data and the weight vector of the corresponding output neuron;

[0073] The calculation module 22 is also used to perform a weighted calculation on the Euclidean distance and the clustering distance to determine the winning neuron;

[0074] Update module 23 is used to update the output neuron weight vector of the water pollution neural network according to the input neuron weight vector and the output neuron weight vector of the winning neuron;

[0075] The determination module 24 is used to determine the groundwater pollution source in the target area when the number of times the output neuron weight vector of the water pollution neural network is updated reaches a preset value.

[0076] In an optional embodiment provided by the present invention, the calculation module 22 is specifically used for:

[0077] The Euclidean distance between the sample data and the output neuron is calculated using the following formula:

[0078]

[0079] Where d(x) j Let w be the Euclidean distance between the j-th sample data and the output neuron, where j∈[1,N] and N is the number of sample data. ji To output the neuron weight vector, x ji Let n be the input neuron weight vector corresponding to the i-th data dimension in the j-th sample data, where n is the data dimension of the sample data.

[0080] In an optional embodiment provided by the present invention, the calculation module 22 is specifically used for:

[0081] The clustering distance between the sample data and the output neuron is calculated using the following formula:

[0082] dist(x j ,w j )=||x j -w j ||

[0083] Where, dist(x) j ,w j ) represents the clustering distance between the sample data and the output neuron, j∈[1,N], N is the number of sample data, c j For the j-th sample data, w ji Let j be the j-th output neuron.

[0084] In an optional embodiment provided by the present invention, the calculation module 22 is specifically used for:

[0085] The step of determining the winning neuron by weighting the Euclidean distance and the clustering distance includes:

[0086] Find the minimum Euclidean distance and the minimum cluster distance;

[0087] The winning neuron is determined by a weighted calculation of the minimum Euclidean distance and the minimum cluster distance.

[0088] In an optional embodiment provided by the present invention, the updating module 23 is specifically used for:

[0089] Obtain the influence range of the winning neuron;

[0090] The output neuron weight vector of the water pollution neural network is updated based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector.

[0091] In an optional embodiment of the present invention, the acquisition module 21 is further configured to calculate the influence range of the winning neuron using the following formula:

[0092]

[0093] Where σ(t) is the influence range of the winning neuron, σ0 is the initial influence range of the winning neuron, t is the time length, and τ0 is a fixed decay coefficient.

[0094] In an optional embodiment provided by the present invention, the updating module 23 is specifically used for:

[0095] The output neuron weight vector of the water pollution neural network is updated using the following formula:

[0096] Δw p=η(t)*T(t)*(x p -w p )

[0097]

[0098]

[0099] Where, Δw p The output neuron weight vector of the updated water pollution neural network is given, where η(t) is the learning coefficient that decays over time, T(t) is the neighborhood influence coefficient, η0 is the initial value of the learning coefficient, and x... p w is the input neuron weight vector of the winning neuron. p Let d be the output neuron weight vector of the winning neuron. pq τ is the Euclidean distance between the winning neuron and its nearest neighbor neurons. η The learning coefficient is fixed, and σ(t) represents the influence range of the winning neuron.

[0100] For specific limitations regarding the device, please refer to the limitations on groundwater pollution source identification methods mentioned above, which will not be repeated here. Each module in the aforementioned device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0101] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for identifying groundwater pollution sources.

[0102] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0103] Obtain sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data;

[0104] Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector;

[0105] The winning neuron is determined by weighted calculation of the Euclidean distance and the clustering distance.

[0106] The output neuron weight vector of the water pollution neural network is updated based on the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0107] When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

[0108] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0109] Obtain sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data;

[0110] Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector;

[0111] The winning neuron is determined by weighted calculation of the Euclidean distance and the clustering distance.

[0112] The output neuron weight vector of the water pollution neural network is updated based on the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0113] When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

[0114] In one embodiment, a computer program product is provided, the computer program product comprising a computer program that is executed by a processor to perform the following steps:

[0115] Obtain sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data;

[0116] Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector;

[0117] The winning neuron is determined by weighted calculation of the Euclidean distance and the clustering distance.

[0118] The output neuron weight vector of the water pollution neural network is updated based on the input neuron weight vector and the output neuron weight vector of the winning neuron.

[0119] When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

[0120] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0121] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0122] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for identifying a source of groundwater pollution, characterized by, The method includes: Obtain sample data for groundwater pollution source detection, wherein the sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data; An input vector reflecting the spatial distribution characteristics of groundwater is constructed based on the latitude and longitude coordinates and surface water system data in the sample data. Calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector, wherein the clustering distance is used to characterize the hydraulic connection correlation between different monitoring points; The winning neuron is determined by weighted calculation of the Euclidean distance and the clustering distance. The output neuron weight vector of the water pollution neural network is updated based on the input neuron weight vector and the output neuron weight vector of the winning neuron. The step of updating the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and the output neuron weight vector of the winning neuron includes: Obtain the influence range of the winning neuron; The output neuron weight vector of the water pollution neural network is updated based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector. The process of obtaining the influence range of the winning neuron includes: The influence range of the winning neuron is calculated using the following formula: wherein, is the influence range of the winning neuron, is the initial influence range of the winning neuron, t is the length of time, is a fixed decay coefficient; The step of updating the output neuron weight vector of the water pollution neural network based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector includes: The output neuron weight vector of the water pollution neural network is updated using the following formula: , , ,in, To update the output neuron weight vector of the water pollution neural network, It is a learning coefficient that decays over time. It is the neighborhood influence coefficient. The initial values ​​for the learning coefficients, The input neuron weight vector of the winning neuron. The output neuron weight vector of the winning neuron. The Euclidean distance between the winning neuron and its nearest neighbor neurons. To use a fixed learning coefficient, The range of influence of the winning neuron; When the number of times the weight vector of the output neuron of the water pollution neural network is updated reaches a preset value, the last updated water pollution neural network determines the groundwater pollution source in the target area.

2. The method of claim 1, wherein, The calculation of the Euclidean distance between the sample data and the corresponding output neuron weight vector includes: The Euclidean distance between the sample data and the output neuron is calculated using the following formula: ,in, For the first The Euclidean distance between each sample data point and the output neuron , The number of sample data. To output the neuron weight vector, For the first In the sample data of the th , the ... The input neuron weight vector corresponding to the data dimension. The data dimension of the sample data.

3. The method of claim 1, wherein, The calculation of the clustering distance between the sample data and the corresponding output neuron weight vector includes: The clustering distance between the sample data and the output neuron is calculated using the following formula: ,in, The clustering distance between the sample data and the output neuron. , The number of sample data, For the first One sample data, For the first One output neuron.

4. The method according to any one of claims 1 to 3, characterized in that, The step of determining the winning neuron by weighting the Euclidean distance and the clustering distance includes: Find the minimum Euclidean distance and the minimum cluster distance; The winning neuron is determined by a weighted calculation of the minimum Euclidean distance and the minimum cluster distance.

5. An apparatus for identifying a source of groundwater pollution, for implementing the method for identifying a source of groundwater pollution according to any one of claims 1 to 4, characterized in that, The device includes: The acquisition module is used to acquire sample data for groundwater pollution source detection. The sample data includes at least water chemical index concentration data, pollutant concentration data, latitude and longitude coordinates, surface water system data, and enterprise type data. The calculation module constructs an input vector reflecting the spatial distribution characteristics of groundwater based on the latitude and longitude coordinates and surface water system data in the sample data. This vector is used to calculate the Euclidean distance and clustering distance between the sample data and the corresponding output neuron weight vector. The clustering distance is used to characterize the hydraulic connection correlation between different monitoring points. The calculation module is also used to perform a weighted calculation on the Euclidean distance and the clustering distance to determine the winning neuron; The update module is used to update the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and the output neuron weight vector of the winning neuron. The step of updating the output neuron weight vector of the water pollution neural network based on the input neuron weight vector and the output neuron weight vector of the winning neuron includes: Obtain the influence range of the winning neuron; The output neuron weight vector of the water pollution neural network is updated based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector. The process of obtaining the influence range of the winning neuron includes: The influence range of the winning neuron is calculated using the following formula: wherein, is the influence range of the winning neuron, is the initial influence range of the winning neuron, t is the length of time, is a fixed decay coefficient; The step of updating the output neuron weight vector of the water pollution neural network based on the influence range of the winning neuron, the input neuron weight vector, and the output neuron weight vector includes: The output neuron weight vector of the water pollution neural network is updated using the following formula: , , ,in, To update the output neuron weight vector of the water pollution neural network, It is a learning coefficient that decays over time. It is the neighborhood influence coefficient. The initial values ​​for the learning coefficients, The input neuron weight vector of the winning neuron. The output neuron weight vector of the winning neuron. The Euclidean distance between the winning neuron and its nearest neighbor neurons. To use a fixed learning coefficient, The range of influence of the winning neuron; The determination module is used to determine the groundwater pollution source in the target area when the number of times the output neuron weight vector of the water pollution neural network is updated reaches a preset value.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the groundwater pollution source identification method as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, wherein the computer program comprises the following steps of: receiving a request for a resource from a client; determining whether the client is authorized to access the resource; and if the client is authorized to access the resource, providing the resource to the client. When the computer program is executed by the processor, it implements the groundwater pollution source identification method as described in any one of claims 1 to 4.

Images