Pollution source position determination method and device, electronic equipment and program product

Abstract

The invention discloses a pollution source position determination method and device, electronic equipment and a program product, and relates to the technical field of artificial intelligence, and the determination method comprises the steps: collecting environment parameter data and spectral image data of a plurality of monitoring points, and constructing a pollutant spatial distribution diagram; clustering all heavy metal concentration data in the pollutant spatial distribution diagram to obtain a cluster set, and calculating the concentration gradient direction of each core point; based on all the concentration gradient directions, a pollution source candidate position set is determined by adopting a reverse particle tracking technology, all pollution source candidate positions are screened, and a target pollution source candidate position set is obtained; and scoring each pollution source candidate position in the target pollution source candidate position set by adopting a preset pollution source scoring model to obtain a target score, and determining the pollution source candidate position corresponding to the maximum target score as the pollution source position. According to the invention, the technical problem of low pollution source positioning accuracy in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a pollution source position determination method and device, electronic equipment and program product. BACKGROUND

[0002] Soil heavy metal pollution widely exists in industrial, agricultural and frequently-traffic areas, and has the characteristics of long-term, concealment and irreversibility. Pollution source identification is the premise of pollution control. Current tracing methods mainly rely on manual sampling and experimental analysis, such as collecting soil, water or dust samples on site, determining heavy metal concentration distribution through chemical detection means, and then judging the pollution diffusion direction and possible source point according to experience. However, the above method is tedious, long-period and high-cost, and has high requirements for spatial coverage and real-time performance. However, the current supervised learning method needs to know the pollution source label (i.e. some places are explicitly labeled as "pollution source" or "non-pollution source") as training data in advance, which is difficult to obtain in actual pollution scenarios, greatly limiting the generalization ability and emergency response efficiency of the model, thereby making it difficult to accurately determine the pollution source position.

[0003] At present, no effective solution has been proposed for the above problems. SUMMARY

[0004] The embodiments of the present application provide a pollution source position determination method and device, electronic equipment and program product, to at least solve the technical problem of low accuracy in positioning pollution sources in related technologies.

[0005] According to an aspect of an embodiment of the present application, a pollution source position determination method is provided, comprising: collecting environmental parameter data of a plurality of monitoring points in a target soil area and spectral image data of the target soil area, and based on all the environmental parameter data and the spectral image data, constructing a pollutant spatial distribution map, wherein the pollutant spatial distribution map at least includes heavy metal concentration data of a plurality of interpolation grid points; clustering all the heavy metal concentration data in the pollutant spatial distribution map to obtain a cluster set, and based on all the heavy metal concentration data, calculating the concentration gradient direction of the core point of each cluster in the cluster set; obtaining a plurality of environmental factor data of the target soil area, and based on all the environmental factor data and all the concentration gradient directions, using a reverse particle tracking technology to determine a set of pollution source candidate positions, and screening all the pollution source candidate positions in the set of pollution source candidate positions to obtain a set of target pollution source candidate positions; using a preset pollution source scoring model to score each pollution source candidate position in the set of target pollution source candidate positions to obtain a target score of each pollution source candidate position, and determining the pollution source candidate position corresponding to the maximum target score as the pollution source position.

[0006] Furthermore, the environmental parameter data includes at least: heavy metal concentration parameter data. Each environmental parameter data is associated with the spatial coordinate information of a monitoring point. The step of constructing a pollutant spatial distribution map based on all environmental parameter data and spectral image data includes: normalizing each environmental parameter data to obtain processed environmental parameter data, wherein the processed environmental parameter data includes at least: processed heavy metal concentration parameter data; dividing the target soil area into multiple interpolation grid points based on spectral image data, and determining the spatial location coordinates of each interpolation grid point; for each interpolation grid point, calculating the distance between the interpolation grid point and each monitoring point based on the spatial location coordinates and all spatial coordinate information, and determining a preset distance weight; calculating the heavy metal concentration data of each interpolation grid point based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data; and constructing a pollutant spatial distribution map based on all heavy metal concentration data.

[0007] Furthermore, the step of calculating the concentration gradient direction of the core point of each cluster in the cluster set based on all heavy metal concentration data includes: calculating the target heavy metal concentration data of each grid point to be interpolated using a preset interpolation method, wherein the core point is the target grid point to be interpolated for each cluster; determining the target heavy metal concentration data corresponding to each core point; calculating the concentration gradient vector of each core point based on the target heavy metal concentration data, and calculating the magnitude of the concentration gradient vector; and determining the concentration gradient direction of each core point based on the concentration gradient vector and the magnitude.

[0008] Furthermore, the environmental factor data includes at least: topographic data, wind field data, and water flow direction data. Based on all environmental factor data and all concentration gradient directions, the step of determining the set of candidate pollution source locations using reverse particle tracing technology includes: determining the migration velocity vector based on topographic data, wind field data, water flow direction data, and preset weights; initializing multiple particles within all clusters and determining the movement step size of each particle, where each particle corresponds to initial spatial position information; propagating multiple particles in the opposite direction of the concentration gradient direction to obtain the next spatial position information of each particle, where the next spatial position information is obtained through the particle's initial spatial position information. The target position information of each particle is determined by the intermediate position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. The intermediate position information is determined by the next spatial position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. All spatial position information of each particle is recorded, and a path density map is generated based on all spatial position information. The paths in the path density map are clustered to obtain multiple path intersection points, and each path intersection point is determined as a candidate pollution source location. All candidate pollution source locations are added to the pollution source candidate location set.

[0009] Further, the step of filtering all candidate pollution sources in the candidate pollution source location set to obtain the target candidate pollution source location set includes: acquiring an image of each candidate pollution source location in the candidate pollution source location set, and processing each image using a preset image classification model to obtain a recognition result, wherein the recognition result includes at least: image pollution feature coordinates and pollution classification labels; performing spatial transformation on each image pollution feature coordinate to obtain spatial coordinates, and constructing a spatial classification label map based on all spatial coordinates and all pollution classification labels; overlaying the spatial classification label map with the path density map, and determining the spatial relationship between the image pollution feature coordinates and the path intersection point; if the image pollution feature coordinates are located outside the preset area of ​​the path intersection point, deleting the candidate pollution source location corresponding to the path intersection point from the candidate pollution source location set to obtain the target candidate pollution source location set.

[0010] Furthermore, after removing the candidate pollution source locations corresponding to the path intersection points from the set of candidate pollution source locations to obtain the set of candidate target pollution source locations, the process also includes: determining multiple scoring indicators and assigning preset weight coefficients to each scoring indicator; and constructing a preset pollution source scoring model based on all scoring indicators and all preset weight coefficients.

[0011] Furthermore, the identification results include at least a probability score, and the scoring indicators include at least a concentration scoring indicator, a direction scoring indicator, a clustering scoring indicator, and a confidence scoring indicator. A preset pollution source scoring model is used to score each candidate pollution source location in the target pollution source candidate location set, obtaining the target score for each candidate pollution source location. This includes: for each candidate pollution source location, calculating the distance from the candidate pollution source location to the core point of the cluster; determining a preset attenuation coefficient; and based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, determining the score for the concentration scoring indicator. The process involves: calculating the vector difference between the candidate pollution source location and the core point, and calculating the magnitude of the vector difference; determining the unit vector between the candidate pollution source location and the core point based on the vector difference and the magnitude; calculating the cosine value of the direction angle based on the unit vector and the concentration gradient vector of the core point, and using the average value of the cosine value of the direction angle as the score value of the direction scoring index; counting the number of paths within a preset range of the candidate pollution source location, determining the number of paths as the score value of the aggregation degree scoring index, and using the probability score as the score value of the confidence degree scoring index; and inputting all the score values ​​into the preset pollution source scoring model to obtain the target score.

[0012] According to another aspect of the embodiments of this application, a device for determining the location of a pollution source is also provided, comprising: a construction unit, used to collect environmental parameter data and spectral image data of a target soil area from multiple monitoring points, and construct a spatial distribution map of pollutants based on all environmental parameter data and spectral image data, wherein the spatial distribution map of pollutants includes at least: heavy metal concentration data of multiple grid points to be interpolated; a calculation unit, used to cluster all heavy metal concentration data in the spatial distribution map of pollutants to obtain a set of clusters, and calculate the concentration gradient direction of the core point of each cluster in the set of clusters based on all heavy metal concentration data; a screening unit, used to acquire multiple environmental factor data of the target soil area, and determine a set of candidate pollution source locations based on all environmental factor data and all concentration gradient directions using reverse particle tracking technology, and screen all candidate pollution source locations in the set of candidate pollution source locations to obtain a set of target pollution source candidate locations; and a scoring unit, used to score each candidate pollution source location in the set of target pollution source candidate locations using a preset pollution source scoring model to obtain a target score for each candidate pollution source location, and determine the candidate pollution source location corresponding to the highest target score as the pollution source location.

[0013] Furthermore, the environmental parameter data includes at least: heavy metal concentration parameter data. Each environmental parameter data is associated with the spatial coordinate information of a monitoring point. The construction unit includes: a first processing module, used to normalize each environmental parameter data to obtain processed environmental parameter data, wherein the processed environmental parameter data includes at least: processed heavy metal concentration parameter data; a first determining module, used to divide the target soil area into multiple interpolation grid points based on spectral image data, and determine the spatial location coordinates of each interpolation grid point; a first calculation module, used to calculate the distance between each interpolation grid point and each monitoring point based on the spatial location coordinates and all spatial coordinate information, and determine a preset distance weight; a second calculation module, used to calculate the heavy metal concentration data of each interpolation grid point based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data; and a first construction module, used to construct a pollutant spatial distribution map based on all heavy metal concentration data.

[0014] Furthermore, the calculation unit includes: a third calculation module, used to calculate the target heavy metal concentration data of each grid point to be interpolated using a preset interpolation method, wherein the core point is the target grid point to be interpolated for each cluster; a second determination module, used to determine the target heavy metal concentration data corresponding to each core point; a fourth calculation module, used to calculate the concentration gradient vector of each core point based on the target heavy metal concentration data, and calculate the magnitude of the concentration gradient vector; and a third determination module, used to determine the concentration gradient direction of each core point based on the concentration gradient vector and the magnitude.

[0015] Furthermore, the environmental factor data includes at least: topographic data, wind field data, and water flow direction data. The filtering unit includes: a fourth determination module, used to determine the migration velocity vector based on topographic data, wind field data, water flow direction data, and preset weights; a fifth determination module, used to initialize multiple particles within all clusters and determine the movement step size of each particle, wherein the particles correspond to initial spatial position information; and a first propagation module, used to propagate multiple particles along the opposite direction of the concentration gradient to obtain the next spatial position information of each particle, wherein the next spatial position information is obtained through the particle's initial spatial position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. The quantity is determined; the sixth determination module is used to determine the target position information of each particle based on the intermediate position information, the movement step size, the diffusion coefficient, the preset items, and the migration velocity vector, wherein the intermediate position information is determined by the next spatial position information, the movement step size, the diffusion coefficient, the preset items, and the migration velocity vector; the first generation module is used to record all spatial position information of each particle and generate a path density map based on all spatial position information; the first clustering module is used to cluster the paths in the path density map to obtain multiple path intersection points, and determine each path intersection point as a candidate pollution source location, and add all candidate pollution source locations to the set of candidate pollution source locations.

[0016] Furthermore, the screening unit also includes: a second processing module, used to acquire an image of each pollution source candidate location in the pollution source candidate location set, and process each image using a preset image classification model to obtain a recognition result, wherein the recognition result includes at least: image pollution feature coordinates and pollution classification labels; a first transformation module, used to perform spatial transformation on each image pollution feature coordinate to obtain spatial coordinates, and construct a spatial classification label map based on all spatial coordinates and all pollution classification labels; a first judgment module, used to overlay the spatial classification label map with the path density map, and judge the spatial relationship between the image pollution feature coordinates and the path intersection point; and a first deletion module, used to delete the pollution source candidate location corresponding to the path intersection point from the pollution source candidate location set when the image pollution feature coordinates are located outside the preset area of ​​the path intersection point, to obtain the target pollution source candidate location set.

[0017] Furthermore, the pollution source location determination device also includes: a seventh determination module, used to determine multiple scoring indicators and assign a preset weight coefficient to each scoring indicator after deleting the pollution source candidate locations corresponding to the path intersection points from the pollution source candidate location set; and a second construction module, used to construct a preset pollution source scoring model based on all scoring indicators and all preset weight coefficients.

[0018] Furthermore, the identification results include at least a probability score, and the scoring indicators include at least a concentration scoring indicator, a direction scoring indicator, a clustering scoring indicator, and a confidence scoring indicator. The scoring unit includes: a fifth calculation module, used to calculate the distance from each pollution source candidate location to the core point of the cluster; an eighth determination module, used to determine a preset attenuation coefficient and, based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, determine the score value of the concentration scoring indicator; and a sixth calculation module, used to calculate the vector difference between the pollution source candidate location and the core point. The module determines the unit vector between the candidate pollution source location and the core point based on the vector difference and the module length; the seventh calculation module calculates the cosine value of the direction angle based on the unit vector and the concentration gradient vector of the core point, and uses the average value of the cosine value of the direction angle as the score value of the direction scoring index; the first statistics module counts the number of paths within a preset range of the candidate pollution source location, determines the number of paths as the score value of the aggregation scoring index, and uses the probability score as the score value of the confidence scoring index; the first input module inputs all the score values ​​into the preset pollution source scoring model to obtain the target score.

[0019] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any of the above-described methods for determining the location of pollution sources.

[0020] According to another aspect of the embodiments of this application, an electronic device is also provided, including one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors cause the one or more processors to implement any of the above-described methods for determining the location of pollution sources.

[0021] In this invention, environmental parameter data and spectral image data of multiple monitoring points in the target soil area are collected. Based on all environmental parameter data and spectral image data, a spatial distribution map of pollutants is constructed. All heavy metal concentration data in the spatial distribution map are clustered to obtain a set of clusters. Based on all heavy metal concentration data, the concentration gradient direction of the core point of each cluster in the set of clusters is calculated. Multiple environmental factor data of the target soil area are obtained. Based on all environmental factor data and all concentration gradient directions, inverse particle tracing technology is used to determine a set of candidate pollution source locations. All candidate pollution source locations in the set of candidate pollution source locations are screened to obtain a set of target pollution source candidate locations. A preset pollution source scoring model is used to score each candidate pollution source location in the set of target pollution source candidate locations to obtain a target score for each candidate pollution source location. The candidate pollution source location corresponding to the highest target score is determined as the pollution source location, thus solving the technical problem of low accuracy in locating pollution sources in related technologies.

[0022] In this invention, environmental parameter data is first collected using sensors deployed at multiple monitoring points within the target soil area. This data is then combined with spectral image data acquired by a drone equipped with a multispectral camera. Interpolation techniques are used to extend the discrete sensor data to continuous grid points, thereby constructing a three-dimensional spatial distribution map of pollutants. Each grid point in the pollutant spatial distribution map is associated with heavy metal concentration data. Then, based on the pollutant spatial distribution map, a clustering algorithm is applied to process all heavy metal concentration data, automatically identifying hotspots with concentration anomalies and forming cluster sets. Furthermore, based on all heavy metal concentration data, the concentration gradient direction of the core point of each cluster can be calculated. Simultaneously, multiple environmental factor data for the target soil area are acquired, and... Using inverse particle tracing technology with all environmental factor data and all concentration gradient directions, the concentration gradient direction is used as the starting point to simulate the propagation path of pollutants in reverse. This allows for the determination of a set of candidate pollution source locations. All candidate pollution source locations in the set can be filtered to obtain a target set of candidate pollution source locations. Then, a pre-set pollution source scoring model is used to score each candidate pollution source location in the target set, obtaining a target score for each candidate pollution source location. The candidate pollution source location with the highest target score is determined as the pollution source location. This achieves automated and high-precision identification of soil heavy metal pollution sources even when the pollution source location is unknown and there is a lack of prior pollution source labels. Attached Figure Description

[0023] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0024] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for determining the location of a pollution source is shown.

[0025] Figure 2 This is a flowchart of the method for determining the location of a pollution source according to Embodiment 1 of this application;

[0026] Figure 3 This is an optional scoring flowchart based on multiple scoring indicators according to an embodiment of this application;

[0027] Figure 4 This is a flowchart of an optional method for determining the location of a pollution source based on multiple scoring indicators, according to an embodiment of this application.

[0028] Figure 5 This is a schematic diagram of an optional pollution source location determination device according to an embodiment of this application;

[0029] Figure 6 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] It should be noted that all related information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) collected and involved in this invention are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data comply with the relevant laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and it does not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse. For example, this system has an interface with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0033] In this invention, by fusing ground-based multi-sensor data and high-resolution remote sensing image data from UAVs, pollution source labeling data is not required. In complex environments where the source is unknown and labeling data is missing, it can automatically identify abnormal pollution areas, infer the migration direction of pollutants, and simulate their reverse propagation path. Ultimately, it achieves accurate location and credibility scoring of pollution sources, realizing unsupervised pollution source tracing based on multi-source perception. At the same time, by utilizing image classification technology, it can not only pinpoint the exact geographical location of pollution sources but also identify their type, such as industrial sewage outlets or open material storage yards, thus broadening the dimensions of pollution source identification. In situations where source information is unknown, it achieves high-precision and high-efficiency pollution source identification.

[0034] The present invention will now be described in detail with reference to various embodiments.

[0035] Example 1

[0036] According to an embodiment of this application, an embodiment of a method for determining the location of a pollution source is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0037] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing a method to determine the location of a pollution source is shown. Figure 1 As shown, computer terminal 10 (or mobile device) may include one or more ( Figure 1The processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions may also be included. In addition, it may include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera, wherein the network interface can be connected to wired and / or wireless networks. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0038] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0039] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the pollution source location determination method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the pollution source location determination method described above. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0040] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0041] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0042] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for determining the location of pollution sources is shown. Figure 2 This is a flowchart of the method for determining the location of a pollution source according to Embodiment 1 of this application, as follows: Figure 2 As shown, the method includes the following steps:

[0043] Step S201: Collect environmental parameter data and spectral image data of multiple monitoring points in the target soil area, and construct a spatial distribution map of pollutants based on all environmental parameter data and spectral image data. The spatial distribution map of pollutants includes at least: heavy metal concentration data of multiple grid points to be interpolated.

[0044] In this embodiment of the invention, environmental parameter data (which can be collected by ground sensors deployed in the soil area, including but not limited to heavy metal concentration data, pH (an indicator of acidity and alkalinity), conductivity, etc., for monitoring the chemical environmental state of the location) and spectral image data of the target soil area (which can be acquired by a multispectral camera mounted on a UAV, containing reflectance information of different bands) are collected from multiple monitoring points in the target soil area (i.e., the geographical area where pollution needs to be monitored and analyzed). Based on all environmental parameter data and spectral image data, the discrete heavy metal concentration data can be extended to each grid point (i.e., the grid point to be interpolated, which can be divided based on spectral image data) of the study area (i.e., the target soil area) through spatial interpolation methods to construct a spatial distribution map of pollutants.

[0045] Optionally, it is not limited to collecting heavy metal concentration data; it can also collect various types of concentration data (such as organic pollutant concentration data, inorganic pollutant concentration data, gaseous pollutant concentration data, etc.) as well as other environmental parameters.

[0046] Step S202: Cluster all heavy metal concentration data in the spatial distribution map of pollutants to obtain a set of clusters, and calculate the concentration gradient direction of the core point of each cluster in the set of clusters based on all heavy metal concentration data.

[0047] In this embodiment of the invention, by performing cluster analysis on the heavy metal concentration data at grid points in the pollutant spatial distribution map (e.g., using the DBSCAN clustering algorithm (Density-Based Spatial Clustering of Applications with Noise), hotspot regions with similar concentration values ​​can be identified. Each hotspot region corresponds to a cluster, thus obtaining a set of clusters. The core point of each cluster is... , Where K represents K core points, i.e. K clusters, and based on all heavy metal concentration data, the concentration gradient direction of the core points of each cluster can be calculated, which may indicate the direction of pollutant propagation.

[0048] Step S203: Obtain multiple environmental factor data for the target soil area, and based on all environmental factor data and all concentration gradient directions, use reverse particle tracking technology to determine the set of candidate pollution source locations, and filter all candidate pollution source locations in the set to obtain the target pollution source candidate location set.

[0049] In this embodiment of the invention, environmental factor data may include, but is not limited to, topographic elevation data, wind field data, water flow direction data, etc. Based on all environmental factor data and all concentration gradient directions, reverse particle tracing technology is used (by tracing the possible migration paths of pollutants in reverse, simulating the reverse propagation trajectory of pollutants from hotspot areas to possible source points, to determine the location of potential pollution sources) to obtain a set of candidate pollution source locations. All candidate pollution source locations in the set of candidate pollution source locations are then screened to obtain a set of target pollution source candidate locations.

[0050] Step S204: Using a preset pollution source scoring model, score each pollution source candidate location in the target pollution source candidate location set to obtain the target score for each pollution source candidate location, and determine the pollution source candidate location corresponding to the highest target score as the pollution source location.

[0051] In this embodiment of the invention, a preset pollution source scoring model (pre-built, combining multiple factors such as concentration anomaly, concentration gradient consistency, path intersection density, and image recognition credibility) is used to score each pollution source candidate location in the target pollution source candidate location set, thereby obtaining a target score for each pollution source candidate location. This target score reflects the comprehensive credibility of the candidate location as a pollution source, and the pollution source candidate location corresponding to the highest target score can be determined as the pollution source location.

[0052] In summary, data from multiple monitoring points in the target soil area, including heavy metal concentration, pH value, and conductivity, were first collected. A spatial distribution map of pollutants was then constructed using continuous spectral images captured by drones. Next, a clustering algorithm was used to identify concentration anomalies and calculate the concentration gradient direction at the core points. Then, incorporating environmental factors such as topography, wind field, and water flow, reverse particle tracking technology was introduced to automatically simulate pollutant migration paths and preliminarily determine multiple candidate pollution source locations. Finally, a pre-defined pollution source scoring model based on multi-factor fusion was used to quantify the credibility of each candidate pollution source location. This approach achieves accurate pollution source identification without relying on detailed pollution source labeling data, thus solving the technical problem of low accuracy in locating pollution sources in related technologies.

[0053] The environmental parameter data includes at least: heavy metal concentration parameter data. Each environmental parameter data is associated with the spatial coordinate information of the monitoring point. In order to accurately construct a pollutant spatial distribution map, in the method for determining the location of pollution sources provided in Embodiment 1 of this application, each environmental parameter data is normalized to obtain processed environmental parameter data. The processed environmental parameter data includes at least: processed heavy metal concentration parameter data; based on spectral image data, the target soil area is divided into multiple interpolation grid points, and the spatial location coordinates of each interpolation grid point are determined; for each interpolation grid point, based on the spatial location coordinates and all spatial coordinate information, the distance between the interpolation grid point and each monitoring point is calculated, and a preset distance weight is determined; based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data, the heavy metal concentration data of each interpolation grid point is calculated; based on all heavy metal concentration data, a pollutant spatial distribution map is constructed.

[0054] Optionally, a ground-based sensor array is deployed in the target soil area to collect environmental parameter data from different soil samples, including heavy metal concentrations, denoted as... Soil pH value, denoted as Soil electrical conductivity, denoted as Each environmental parameter data is associated with the spatial coordinate information of the monitoring point. This allows for the construction of data sets. , where i represents the i-th monitoring point, i=1,…,n.

[0055] In this embodiment of the invention, each environmental parameter data is normalized to obtain processed environmental parameter data. The set of normalized environmental parameter data can be represented as follows: ,in, This represents the heavy metal concentration parameters after processing. This indicates the pH value of the treated soil. This indicates the processed soil electrical conductivity. Then, the ground sensor data and UAV image data are fused, first based on spatial coordinate information. Normalized sensor data is superimposed onto the UAV image coordinate system. Since sensor data consists of discrete points and UAV data (spectral image data) consists of continuous images, spatial interpolation methods are required to extend the discrete sensor data to a spatially continuous form in order to achieve fusion.

[0056] The target soil region is divided into regular grids according to spatial resolution, and a set of grid points is constructed. Where j represents the number of grid points, j=1,…,M (i.e., dividing the target soil area into multiple interpolation grid points and determining the spatial coordinates of each grid point), the grid points are used as the interpolation area, and each grid point is the heavy metal pollution concentration point to be predicted. Since the target pollutant for pollution source tracing is the soil heavy metal concentration, the main pollution parameter during interpolation is the normalized heavy metal concentration data, which is used as the core variable for interpolation calculation. For each grid point to be interpolated, based on the spatial coordinates… and all spatial coordinate information Calculate the distance between the grid points to be interpolated and each monitoring point. ,in, This represents the spatial coordinate information of the i-th monitoring point, and a preset distance weight p is determined. Based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data, the heavy metal concentration data of each grid point to be interpolated is calculated (i.e., It can also construct a spatial distribution map of pollutants based on all heavy metal concentration data.

[0057] To accurately determine the concentration gradient direction of each core point, the pollution source location determination method provided in Embodiment 1 of this application employs a preset interpolation method to calculate the target heavy metal concentration data for each grid point to be interpolated, wherein the core point is the target grid point to be interpolated for each cluster; the target heavy metal concentration data corresponding to each core point is determined; based on the target heavy metal concentration data, the concentration gradient vector of each core point is calculated, and the magnitude of the concentration gradient vector is calculated; based on the concentration gradient vector and the magnitude, the concentration gradient direction of each core point is determined.

[0058] In this embodiment of the invention, based on the heavy metal concentration data of the monitoring points and the spatial coordinate information of all monitoring points, a preset interpolation method (such as Kriging interpolation, a statistical interpolation method) can be used to calculate the target heavy metal concentration data for each grid point to be interpolated. This allows the construction of a continuous pollutant concentration field (including the target heavy metal concentration data for all grid points to be interpolated). The core point of each cluster obtained through clustering is a specific grid point (i.e., the target grid point) among all grid points to be interpolated. Based on the calculated target heavy metal concentration data for all core points, the target heavy metal concentration data corresponding to each core point can be determined. For each core point's target heavy metal concentration data (i.e., the target heavy metal concentration data for each core point...), the target heavy metal concentration data can be... Taking the partial derivative (where k represents the k-th core point), we can calculate the concentration gradient vector of each core point. ), and can calculate the magnitude of the concentration gradient vector (i.e. Based on the concentration gradient vector and magnitude, the concentration gradient direction of each core point is determined (i.e., By calculating the concentration gradient directions of multiple hotspots, the possible directions or ranges of pollution sources (i.e., a set of candidate pollution source locations) can be preliminarily determined. Furthermore, near the concentration gradient directions in hotspot areas (i.e., the core points of clusters), UAV imagery can be used to determine the presence of signs of pollution sources (such as factories, storage areas, and discharge outlets) to verify the rationality of the gradient directions and correct any potential directional deviations. This achieves automated identification of pollutant propagation paths and potential source locations without the need for manual labeling of pollution sources.

[0059] Environmental factor data includes at least: topographic data, wind field data, and water flow direction data. To accurately obtain the set of candidate pollution source locations, the pollution source location determination method provided in Embodiment 1 of this application determines the migration velocity vector based on topographic data, wind field data, water flow direction data, and preset weights; multiple particles are initialized within all clusters, and the movement step size of each particle is determined, wherein each particle corresponds to initial spatial location information; multiple particles are propagated along the opposite direction of the concentration gradient to obtain the next spatial location information of each particle, wherein the next spatial location information is obtained by using the particle's initial spatial location information, movement step size, and expansion... The diffusion coefficient, preset terms, and migration velocity vector are used to determine the target position information of each particle. The intermediate position information is determined by the next spatial position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. All spatial position information of each particle is recorded, and a path density map is generated based on all spatial position information. The paths in the path density map are clustered to obtain multiple path intersection points, and each path intersection point is determined as a candidate pollution source location. All candidate pollution source locations are added to the pollution source candidate location set.

[0060] Optionally, the environmental factor data shall include at least: topographic data. Wind field data and water flow direction data Preset weights can be assigned to each environmental factor data (e.g., , , ).

[0061] In this embodiment of the invention, the migration velocity vector (i.e., based on terrain data, wind field data, water flow direction data, and preset weights) can be determined. ,in, g is the acceleration due to gravity. (This is the gradient vector of the terrain data), and multiple particles can be initialized within all clusters, with the step size of each particle determined. In this process, each particle has initial spatial position information. Multiple particles propagate in the opposite direction of the concentration gradient, and the next spatial position information of each particle is obtained. This next spatial position information is derived from the particle's initial spatial position information, movement step size, diffusion coefficient (D, representing the ability of a pollutant to diffuse in a medium), and preset terms (…). The target position of each particle is determined by the following parameters: (following a normal distribution, simulating diffusion randomness) and the migration velocity vector. Then, based on intermediate position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector, the target position information of each particle can be determined. The intermediate position information is determined by the next spatial position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. Essentially, the above calculation process is a discrete-time particle trajectory backtracking equation. Particles migrate backward under the influence of environmental factors such as terrain slope, wind, and water flow, continuously iterating and updating their positions.

[0062] For example, the initial time is t, and the initial spatial position information is the position of the particle at time t. Next spatial location information ,in, Let t be the migration velocity vector. Repeat the above steps to record the particle's propagation trajectory until a stable position (target position information) is reached.

[0063] The propagation trajectory of each particle (i.e., all spatial location information) is recorded, and a path density map is generated based on all spatial location information. The DBSCAN algorithm can be used to cluster the paths in the path density map to obtain multiple path intersection points, and each path intersection point is identified as a candidate pollution source location. All candidate pollution source locations are added to the pollution source candidate location set. Furthermore, existing environmental data such as satellite and UAV imagery can be combined for manual confirmation of the pollution source candidate locations, ultimately resulting in a complete set of pollution source candidate points. .

[0064] To accurately obtain the set of candidate pollution source locations, the pollution source location determination method provided in Embodiment 1 of this application acquires an image of each candidate pollution source location in the set of candidate pollution source locations, and processes each image using a preset image classification model to obtain a recognition result. The recognition result includes at least: image pollution feature coordinates and pollution classification labels. Spatial transformation is performed on each image pollution feature coordinate to obtain spatial coordinates, and a spatial classification label map is constructed based on all spatial coordinates and all pollution classification labels. The spatial classification label map is superimposed on the path density map, and the spatial relationship between the image pollution feature coordinates and the path intersection point is determined. If the image pollution feature coordinates are located outside the preset area of ​​the path intersection point, the pollution source candidate location corresponding to the path intersection point is deleted from the set of candidate pollution source locations to obtain the set of candidate pollution source locations.

[0065] Optionally, drones can be used to take aerial photos of candidate pollution source locations. The obtained image data is then registered with coordinates and stitched together to form a remote sensing image dataset that is consistent with the coordinates of the geographic information system. This image data is then segmented into an input format acceptable to the model and input into a pre-trained image classification model (i.e., a preset image classification model) for processing. This model can identify images related to pollution activities, such as industrial plants, sewage pipes or outlets, exposed ground, large piles of soil or waste dumps.

[0066] In this embodiment of the invention, an image of each pollution source candidate location in the pollution source candidate location set is acquired, and each image is processed using a preset image classification model to obtain a recognition result. For each image, a corresponding label (such as a sewage outlet, an exposed stockpile, etc.) and a probability score are output. All image recognition results can be remapped back to the original spatial coordinates (i.e., spatial transformation is performed on the pollution feature coordinates of each image to obtain spatial coordinates). Based on all spatial coordinates and all pollution classification labels, a spatial classification label map is constructed. The spatial classification label map can be superimposed with the path density map, and the spatial relationship between the image pollution feature coordinates and the path intersection point is determined. If the image pollution feature coordinates are located outside the preset area of ​​the path intersection point, the pollution source candidate location corresponding to the path intersection point can be deleted from the pollution source candidate location set to obtain the target pollution source candidate location set.

[0067] In order to accurately construct a preset pollution source scoring model, in the method for determining the location of pollution sources provided in Embodiment 1 of this application, multiple scoring indicators are determined, and preset weight coefficients are assigned to each scoring indicator; based on all scoring indicators and all preset weight coefficients, a preset pollution source scoring model is constructed.

[0068] Optionally, multiple scoring indicators may include: concentration anomaly indicators. Gradient consistency index Path intersection density index Image recognition credibility index The information for each indicator is shown in Table 1.

[0069] Table 1

[0070]

[0071] In this embodiment of the invention, a preset weight coefficient can be assigned to each scoring indicator (e.g., ...). , , as well as Based on all scoring indicators and all preset weight coefficients, a preset pollution source scoring model can be constructed (i.e. ).

[0072] Figure 3 This is an optional scoring flowchart based on multiple scoring indicators according to an embodiment of this application, such as... Figure 3 As shown, firstly, several scoring indicators are determined, which may include: concentration anomaly. Gradient consistency Path intersection density Image recognition reliability Then, for each candidate pollution source location, all the above scoring indicators can be weighted and scored to output the pollution source score result.

[0073] The identification results include at least a probability score, and the scoring indicators include at least a concentration scoring indicator, a direction scoring indicator, a clustering scoring indicator, and a confidence scoring indicator. To accurately obtain the target score, in the pollution source location determination method provided in Embodiment 1 of this application, for each pollution source candidate location, the distance from the candidate location to the core point of the cluster is calculated; a preset attenuation coefficient is determined, and based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, the score value of the concentration scoring indicator is determined; the vector difference between the pollution source candidate location and the core point is calculated, and the magnitude of the vector difference is calculated; based on the vector difference and the magnitude, the unit vector between the pollution source candidate location and the core point is determined; based on the unit vector and the concentration gradient vector of the core point, the cosine value of the direction angle is calculated, and the average value of the cosine value of the direction angle is used as the score value of the direction scoring indicator; the number of paths within a preset range of the pollution source candidate location is counted, and the number of paths is determined as the score value of the clustering scoring indicator, and the probability score is used as the score value of the confidence scoring indicator; all score values ​​are input into a preset pollution source scoring model to obtain the target score.

[0074] In embodiments of the present invention, the identification result includes at least a probability score, and the scoring index includes at least a concentration scoring index (i.e., ), directional scoring indicators (i.e. ), clustering score index (i.e. ) and confidence score indicators (i.e. For each candidate pollution source location, taking that location as the assumed pollution source, the coverage and consistency of existing concentration hotspots are scored, and the distance from the candidate pollution source location to the core point of the cluster is calculated. And determine the factors affecting the attenuation coefficient (i.e., the preset attenuation coefficient). (This can be based on the target heavy metal concentration data corresponding to the core point) The distance and preset attenuation coefficient are used to determine the score value of the concentration scoring index. ,in, , This involves using candidate pollution source locations as anti-diffusion centers and, under a propagation model (such as a Gaussian diffusion kernel), predicting their impact on the concentration of each core point in the monitored grid.

[0075] Calculate candidate locations of pollution sources With core points Calculate the vector difference between them and the magnitude of the vector difference. Based on the vector difference and magnitude, the unit vector between the candidate pollution source location and the core point is determined (i.e., (and can be based on unit vectors and the concentration gradient vector at the core point), Calculate the cosine value of the direction angle (i.e.) ), and the average value of the cosine of the direction angle (i.e. Where K is the number of core points) is used as the score value of the direction scoring index (i.e. If the direction of the concentration gradient vector of multiple core points is exactly towards a candidate pollution source location, it indicates that the pollution trend may originate from that point, and the score is high. If the gradient direction is unrelated to or deviates from that point, the score is low.

[0076] Count the number of paths within a preset range (e.g., within a radius r, where r can be set manually) of candidate pollution source locations. And it can be obtained by normalization. (i.e., the score of the aggregation score index). The closer the value is to 1, the more obvious the path intersection is, and the higher the credibility of the pollution source.

[0077] If a candidate pollution source location is identified as having clear pollution characteristics (e.g., a sewage outlet or factory building), the probability score is used as the score value for the confidence level indicator. If it is a non-polluting characteristic, then it can be set as follows: The initial value is 0 or a low value. Then, all score values ​​are input into a preset pollution source scoring model, and a score set is output. ,in, This represents the candidate location of the m-th pollution source. The target score represents the m-th pollution source candidate location. The pollution source candidate location corresponding to the highest target score can be determined as the pollution source location. Finally, the output can be the pollution source location + source type (i.e. pollution behavior type, such as "sewage outlet", "storage yard", "exposed ground" etc.) + reliability score + score index value.

[0078] For example, the output could be: pollution source number. Spatial location coordinates [Unit: meter], Source type is sewage outlet (model confidence score: ),score Concentration anomaly Gradient consistency Path intersection density Image recognition reliability .

[0079] Figure 4 This is an optional flowchart for determining the location of a pollution source based on multiple scoring indicators according to an embodiment of this application, such as... Figure 4 As shown, ground sensor data and UAV multispectral images are first collected. Based on the ground sensor data and UAV multispectral images, a clustering algorithm is used to identify hotspot (i.e., the core points of clusters) areas and construct a spatial distribution map of pollutants. Then, a fine concentration field is constructed by interpolation, and concentration gradient analysis is performed to infer the direction of pollution sources. Through anti-diffusion path simulation, path intersection areas are marked (i.e., the DBSCAN algorithm is used to cluster paths in the path density map to obtain multiple path intersection points). Each path intersection point can be determined as a candidate location of pollution sources. Then, the candidate source areas (i.e., candidate pollution source locations) are scanned at high resolution by UAVs to classify and identify pollution characteristics. Through multi-factor fusion, a weighted scoring model is constructed, which can output pollution source location + source type + confidence score + factor score value (i.e., the value of the scoring index).

[0080] The pollution source location determination method provided in this application can automatically identify the core area of ​​pollution anomalies by using interpolation methods and clustering algorithms. Then, concentration gradient analysis is performed on the core points. Without relying on complete pollution source label data, the preliminary direction of the pollution source in an unsupervised environment is located. Combined with simulation of environmental factors (topography, wind field, water flow direction), the reverse diffusion path of the pollution source is determined. This not only improves the accuracy of pollution source location, but also automatically generates a set of candidate pollution source locations through particle tracking and spatial intersection point analysis, realizing automated and efficient source tracing path simulation. Subsequently, the feature coordinates and classification labels of the pollution source area are obtained by combining image recognition technology. The candidate locations are scored by a multi-factor fusion scoring model (concentration score, direction score, clustering score, confidence score), and the precise location, type, confidence, and score data of the pollution source are output. This achieves accurate identification of soil heavy metal pollution sources based on multi-sensor data and UAV images.

[0081] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0082] Example 2

[0083] This application also provides a device for determining the location of a pollution source. It should be noted that this device can be used to execute the method for determining the location of a pollution source provided in this application. The following describes the device for determining the location of a pollution source provided in this application.

[0084] According to an embodiment of this application, an apparatus for implementing the above-described method for determining the location of a pollution source is also provided. Figure 5 This is a schematic diagram of an optional pollution source location determination device according to an embodiment of this application, such as... Figure 5 As shown, the device for determining the location of the pollution source may include: a construction unit 50, a calculation unit 51, a screening unit 52, and a scoring unit 53.

[0085] The construction unit 50 is used to collect environmental parameter data and spectral image data of multiple monitoring points in the target soil area, and to construct a spatial distribution map of pollutants based on all environmental parameter data and spectral image data. The spatial distribution map of pollutants includes at least: heavy metal concentration data of multiple grid points to be interpolated.

[0086] The calculation unit 51 is used to cluster all heavy metal concentration data in the spatial distribution map of pollutants to obtain a set of clusters, and to calculate the concentration gradient direction of the core point of each cluster in the set of clusters based on all heavy metal concentration data.

[0087] The screening unit 52 is used to acquire multiple environmental factor data of the target soil area, and based on all environmental factor data and all concentration gradient directions, it uses reverse particle tracking technology to determine the set of candidate pollution source locations, and filters all the candidate pollution source locations in the set of candidate pollution source locations to obtain the target pollution source candidate location set.

[0088] The scoring unit 53 is used to score each pollution source candidate location in the target pollution source candidate location set using a preset pollution source scoring model, to obtain the target score of each pollution source candidate location, and to determine the pollution source candidate location corresponding to the highest target score as the pollution source location.

[0089] The pollution source location determination device provided in this application embodiment can collect environmental parameter data and spectral image data of multiple monitoring points in the target soil area through the construction unit 50, and construct a pollutant spatial distribution map based on all environmental parameter data and spectral image data. The calculation unit 51 can cluster all heavy metal concentration data in the pollutant spatial distribution map to obtain a set of clusters, and calculate the concentration gradient direction of the core point of each cluster in the set of clusters based on all heavy metal concentration data. The filtering unit 52 can obtain multiple environmental factor data of the target soil area, and determine a set of candidate pollution source locations using reverse particle tracking technology based on all environmental factor data and all concentration gradient directions. The filtering unit 53 can use a preset pollution source scoring model to score each candidate pollution source location in the set of candidate pollution source locations to obtain a target score for each candidate pollution source location, and determine the candidate pollution source location corresponding to the highest target score as the pollution source location.

[0090] Optionally, the environmental parameter data includes at least: heavy metal concentration parameter data, each environmental parameter data being associated with the spatial coordinate information of a monitoring point. The construction unit includes: a first processing module, used to normalize each environmental parameter data to obtain processed environmental parameter data, wherein the processed environmental parameter data includes at least: processed heavy metal concentration parameter data; a first determining module, used to divide the target soil area into multiple interpolation grid points based on spectral image data, and determine the spatial location coordinates of each interpolation grid point; a first calculation module, used to calculate the distance between each interpolation grid point and each monitoring point based on the spatial location coordinates and all spatial coordinate information, and determine a preset distance weight; a second calculation module, used to calculate the heavy metal concentration data of each interpolation grid point based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data; and a first construction module, used to construct a pollutant spatial distribution map based on all heavy metal concentration data.

[0091] Optionally, the calculation unit includes: a third calculation module, used to calculate the target heavy metal concentration data of each grid point to be interpolated using a preset interpolation method, wherein the core point is the target grid point to be interpolated for each cluster; a second determination module, used to determine the target heavy metal concentration data corresponding to each core point; a fourth calculation module, used to calculate the concentration gradient vector of each core point based on the target heavy metal concentration data, and calculate the magnitude of the concentration gradient vector; and a third determination module, used to determine the concentration gradient direction of each core point based on the concentration gradient vector and the magnitude.

[0092] Optionally, the environmental factor data includes at least: topographic data, wind field data, and water flow direction data. The filtering unit includes: a fourth determination module, used to determine the migration velocity vector based on topographic data, wind field data, water flow direction data, and preset weights; a fifth determination module, used to initialize multiple particles within all clusters and determine the movement step size of each particle, wherein the particles correspond to initial spatial position information; and a first propagation module, used to propagate multiple particles along the opposite direction of the concentration gradient to obtain the next spatial position information of each particle, wherein the next spatial position information is obtained through the particle's initial spatial position information, movement step size, diffusion coefficient, preset terms, and migration velocity vector. The quantity is determined; the sixth determination module is used to determine the target position information of each particle based on the intermediate position information, the movement step size, the diffusion coefficient, the preset items, and the migration velocity vector, wherein the intermediate position information is determined by the next spatial position information, the movement step size, the diffusion coefficient, the preset items, and the migration velocity vector; the first generation module is used to record all spatial position information of each particle and generate a path density map based on all spatial position information; the first clustering module is used to cluster the paths in the path density map to obtain multiple path intersection points, and determine each path intersection point as a candidate pollution source location, and add all candidate pollution source locations to the set of candidate pollution source locations.

[0093] Optionally, the filtering unit further includes: a second processing module, used to acquire an image of each pollution source candidate location in the pollution source candidate location set, and process each image using a preset image classification model to obtain a recognition result, wherein the recognition result includes at least: image pollution feature coordinates and pollution classification labels; a first transformation module, used to perform spatial transformation on each image pollution feature coordinate to obtain spatial coordinates, and construct a spatial classification label map based on all spatial coordinates and all pollution classification labels; a first judgment module, used to overlay the spatial classification label map with the path density map, and judge the spatial relationship between the image pollution feature coordinates and the path intersection point; and a first deletion module, used to delete the pollution source candidate location corresponding to the path intersection point from the pollution source candidate location set when the image pollution feature coordinates are located outside the preset area of ​​the path intersection point, to obtain the target pollution source candidate location set.

[0094] Optionally, the pollution source location determination device further includes: a seventh determination module, used to determine multiple scoring indicators and assign a preset weight coefficient to each scoring indicator after deleting the pollution source candidate locations corresponding to the path intersection points from the pollution source candidate location set; and a second construction module, used to construct a preset pollution source scoring model based on all scoring indicators and all preset weight coefficients.

[0095] Optionally, the identification result includes at least a probability score, and the scoring indicators include at least a concentration scoring indicator, a direction scoring indicator, a clustering scoring indicator, and a confidence scoring indicator. The scoring unit includes: a fifth calculation module, used to calculate the distance from each pollution source candidate location to the core point of the cluster; an eighth determination module, used to determine a preset attenuation coefficient, and based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, determine the score value of the concentration scoring indicator; and a sixth calculation module, used to calculate the vector difference between the pollution source candidate location and the core point, and calculate the vector difference. The module determines the unit vector between the candidate pollution source location and the core point based on the vector difference and the module length; the seventh calculation module calculates the cosine value of the direction angle based on the unit vector and the concentration gradient vector of the core point, and uses the average value of the cosine value of the direction angle as the score value of the direction scoring index; the first statistics module counts the number of paths within a preset range of the candidate pollution source location, determines the number of paths as the score value of the aggregation scoring index, and uses the probability score as the score value of the confidence scoring index; the first input module inputs all the score values ​​into the preset pollution source scoring model to obtain the target score.

[0096] The aforementioned pollution source location determination device may also include a processor and a memory. The aforementioned construction unit 50, calculation unit 51, screening unit 52, scoring unit 53, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0097] The aforementioned processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured. By adjusting kernel parameters, a preset pollution source scoring model is used to score each candidate pollution source location in the target pollution source candidate location set, obtaining a target score for each candidate location. The candidate pollution source location corresponding to the highest target score is then determined as the pollution source location.

[0098] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0099] It should be noted that the aforementioned construction unit 50, calculation unit 51, filtering unit 52, and scoring unit 53 correspond to steps S201 to S204 in Embodiment 1. The instances and application scenarios implemented by the aforementioned units and their corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1. It should be noted that the aforementioned units may be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The aforementioned units may also be part of a device and can run in the computer terminal 10 provided in Embodiment 1.

[0100] Example 3

[0101] Embodiments of this application may provide a computer terminal, which may be any computer terminal device in a group of computer terminals. Optionally, in this embodiment, the aforementioned computer terminal may also be replaced with a mobile terminal or an electronic device, etc.

[0102] Optionally, in this embodiment, the computer terminal may be located in at least one of a plurality of network devices in a computer network.

[0103] In this embodiment, the aforementioned computer terminal can execute the program code for the following steps in the method for determining the location of pollution sources: collecting environmental parameter data and spectral image data of multiple monitoring points in the target soil area, and constructing a spatial distribution map of pollutants based on all environmental parameter data and spectral image data, wherein the spatial distribution map of pollutants includes at least: heavy metal concentration data of multiple grid points to be interpolated; clustering all heavy metal concentration data in the spatial distribution map of pollutants to obtain a set of clusters, and calculating the concentration gradient direction of the core point of each cluster in the set of clusters based on all heavy metal concentration data; acquiring multiple environmental factor data of the target soil area, and determining a set of candidate pollution source locations using reverse particle tracing technology based on all environmental factor data and all concentration gradient directions, and filtering all candidate pollution source locations in the set of candidate pollution source locations to obtain a set of target pollution source candidate locations; using a preset pollution source scoring model to score each candidate pollution source location in the set of target pollution source candidate locations to obtain a target score for each candidate pollution source location, and determining the candidate pollution source location corresponding to the highest target score as the pollution source location.

[0104] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the location of pollution sources: normalizing each environmental parameter data to obtain processed environmental parameter data, wherein the processed environmental parameter data includes at least: processed heavy metal concentration parameter data; dividing the target soil area into multiple interpolation grid points based on spectral image data, and determining the spatial coordinates of each interpolation grid point; for each interpolation grid point, calculating the distance between the interpolation grid point and each monitoring point based on the spatial coordinates and all spatial coordinate information, and determining a preset distance weight; calculating the heavy metal concentration data of each interpolation grid point based on the preset distance weight, all distances, and all processed heavy metal concentration parameter data; and constructing a spatial distribution map of pollutants based on all heavy metal concentration data.

[0105] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the location of pollution sources: using a preset interpolation method to calculate the target heavy metal concentration data for each grid point to be interpolated, wherein the core point is the target grid point to be interpolated for each cluster; determining the target heavy metal concentration data corresponding to each core point; calculating the concentration gradient vector for each core point based on the target heavy metal concentration data, and calculating the magnitude of the concentration gradient vector; determining the concentration gradient direction for each core point based on the concentration gradient vector and the magnitude.

[0106] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the location of pollution sources: determining a migration velocity vector based on terrain data, wind field data, water flow direction data, and preset weights; initializing multiple particles within all clusters and determining the movement step size of each particle, wherein each particle corresponds to initial spatial location information; propagating multiple particles along the opposite direction of the concentration gradient to obtain the next spatial location information of each particle, wherein the next spatial location information is determined by the particle's initial spatial location information, movement step size, diffusion coefficient, preset terms, and migration velocity vector; determining the target location information of each particle based on intermediate location information, movement step size, diffusion coefficient, preset terms, and migration velocity vector, wherein the intermediate location information is determined by the next spatial location information, movement step size, diffusion coefficient, preset terms, and migration velocity vector; recording all spatial location information of each particle and generating a path density map based on all spatial location information; clustering the paths in the path density map to obtain multiple path intersection points, determining each path intersection point as a candidate pollution source location, and adding all candidate pollution source locations to the set of candidate pollution source locations.

[0107] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the location of pollution sources: acquiring an image of each candidate pollution source location in the set of candidate pollution source locations, and processing each image using a preset image classification model to obtain a recognition result, wherein the recognition result includes at least: image pollution feature coordinates and pollution classification labels; performing spatial transformation on each image pollution feature coordinate to obtain spatial coordinates, and constructing a spatial classification label map based on all spatial coordinates and all pollution classification labels; overlaying the spatial classification label map with a path density map, and determining the spatial relationship between the image pollution feature coordinates and the path intersection point; if the image pollution feature coordinates are located outside a preset area of ​​the path intersection point, deleting the pollution source candidate location corresponding to the path intersection point from the set of candidate pollution source locations to obtain a target set of candidate pollution source locations.

[0108] Optionally, the computer terminal described above can execute program code for the following steps in the method for determining the location of pollution sources: determining multiple scoring indicators and assigning preset weight coefficients to each scoring indicator; and constructing a preset pollution source scoring model based on all scoring indicators and all preset weight coefficients.

[0109] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the method for determining the location of pollution sources: For each candidate pollution source location, calculate the distance from the candidate pollution source location to the core point of the cluster; determine a preset attenuation coefficient, and based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, determine the score value of the concentration scoring index; calculate the vector difference between the candidate pollution source location and the core point, and calculate the magnitude of the vector difference, and based on the vector difference and the magnitude, determine the unit vector between the candidate pollution source location and the core point; based on the unit vector and the concentration gradient vector of the core point, calculate the cosine value of the direction angle, and use the average value of the cosine value of the direction angle as the score value of the direction scoring index; count the number of paths within a preset range of the candidate pollution source location, determine the number of paths as the score value of the clustering score index, and use the probability score as the score value of the confidence score index; input all score values ​​into the preset pollution source scoring model to obtain the target score.

[0110] Optionally, Figure 6 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 6 As shown, the electronic device may include: one or more ( Figure 6 (Only one is shown) Processor 602, memory 604, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0111] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the pollution source location determination method and apparatus in this application embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned pollution source location determination method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0112] The processor can call the information and application program stored in the memory through the transmission device to execute the steps described above in the method for determining the location of the pollution source.

[0113] The embodiments of this application provide a scheme for determining the location of pollution sources. By using concentration gradient modeling, path inversion and image recognition, the joint identification of candidate pollution source locations and types is achieved. By using a scoring model that integrates multiple factors, a target score for each candidate pollution source location can be output, and the candidate pollution source location with the highest target score is determined as the pollution source location. This realizes a source tracing process without pollution source labels, thereby solving the technical problem of low accuracy in locating pollution sources in related technologies.

[0114] Those skilled in the art will understand that Figure 6 The structure shown is for illustrative purposes only. Electronic devices can also be terminal devices such as smartphones, tablets, PDAs, and mobile internet devices (MIDs). Figure 6 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 6 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 6 The different configurations shown.

[0115] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0116] Example 4

[0117] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method for determining the location of the pollution source provided in Embodiment 1.

[0118] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0119] This application also provides a computer program product, which, when executed on a data processing device, is adapted to perform the steps of a method for determining the location of a pollution source.

[0120] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0121] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0122] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0123] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0124] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0125] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0126] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method of determining the location of a pollution source, characterized in that, The method comprises the following steps: Collecting environmental parameter data of a plurality of monitoring points in a target soil area and spectral image data of the target soil area, and constructing a pollutant spatial distribution map based on all the environmental parameter data and the spectral image data, wherein the pollutant spatial distribution map at least comprises heavy metal concentration data of a plurality of grid points to be interpolated; Clustering all the heavy metal concentration data in the pollutant spatial distribution map to obtain a cluster set, and calculating the concentration gradient direction of the core point of each cluster in the cluster set based on all the heavy metal concentration data; Obtaining a plurality of environmental factor data of the target soil area, and determining a set of candidate positions of pollution sources based on all the environmental factor data and all the concentration gradient directions by using a reverse particle tracking technology, and screening all the candidate positions of pollution sources in the set of candidate positions of pollution sources to obtain a set of target candidate positions of pollution sources; Scoring each of the candidate positions of pollution sources in the set of target candidate positions of pollution sources by using a preset pollution source scoring model to obtain a target score of each of the candidate positions of pollution sources, and determining the candidate position of pollution sources corresponding to the maximum target score as the position of the pollution source.

2. The method of determining the position of a pollution source according to claim 1, wherein, The environmental parameter data at least comprises heavy metal concentration parameter data, and each of the environmental parameter data is associated with spatial coordinate information of the monitoring point, and the step of constructing a pollutant spatial distribution map based on all the environmental parameter data and the spectral image data comprises: Normalizing each of the environmental parameter data to obtain processed environmental parameter data, wherein the processed environmental parameter data at least comprises processed heavy metal concentration parameter data; Dividing the target soil area into a plurality of grid points to be interpolated based on the spectral image data, and determining the spatial position coordinates of each of the grid points to be interpolated; For each of the grid points to be interpolated, calculating the distance between the grid point to be interpolated and each of the monitoring points based on the spatial position coordinates and all the spatial coordinate information, and determining a preset distance weight; Calculating the heavy metal concentration data of each of the grid points to be interpolated based on the preset distance weight, all the distances, and all the processed heavy metal concentration parameter data; Constructing the pollutant spatial distribution map based on all the heavy metal concentration data.

3. The method of claim 1, wherein The step of calculating the concentration gradient direction of the core point of each cluster in the cluster set based on all the heavy metal concentration data comprises: Calculating target heavy metal concentration data of each of the grid points to be interpolated by using a preset interpolation method, wherein the core point is the target grid point to be interpolated of each of the clusters; Determining the target heavy metal concentration data corresponding to each of the core points; Calculating the concentration gradient vector of each of the core points based on the target heavy metal concentration data, and calculating the module length of the concentration gradient vector; Determining the concentration gradient direction of each of the core points based on the concentration gradient vector and the module length.

4. The method of claim 1, wherein The environmental factor data at least includes terrain data, wind field data and water flow direction data, and based on all the environmental factor data and all the concentration gradient directions, a reverse particle tracking technology is used to determine a set of pollution source candidate positions, including: Based on the terrain data, the wind field data, the water flow direction data and a preset weight, a migration velocity vector is determined; A plurality of particles are initialized in all the clustering clusters, and the moving step of each particle is determined, wherein the particle corresponds to initial spatial position information; A plurality of particles are propagated along the opposite direction of the concentration gradient direction to obtain next spatial position information of each particle, wherein the next spatial position information is determined by the initial spatial position information of the particle, the moving step, a diffusion coefficient, a preset term and the migration velocity vector; Based on intermediate position information, the moving step, the diffusion coefficient, the preset term and the migration velocity vector, target position information of each particle is determined, wherein the intermediate position information is determined by the next spatial position information, the moving step, the diffusion coefficient, the preset term and the migration velocity vector; All spatial position information of each particle is recorded, and a path density map is generated based on all the spatial position information; The paths in the path density map are clustered to obtain a plurality of path intersection points, each path intersection point is determined as a pollution source candidate position, and all the pollution source candidate positions are added to the set of pollution source candidate positions.

5. The method of claim 1, wherein The steps of screening all pollution source candidate positions in the set of pollution source candidate positions to obtain a set of target pollution source candidate positions, including: An image of each pollution source candidate position in the set of pollution source candidate positions is obtained, and each image is processed using a preset image classification model to obtain an identification result, wherein the identification result at least includes image pollution feature coordinates and pollution classification labels; Each image pollution feature coordinate is spatially converted to obtain a spatial coordinate, and a spatial classification label map is constructed based on all the spatial coordinates and all the pollution classification labels; The spatial classification label map is superimposed with the path density map, and the spatial relationship between the image pollution feature coordinates and the path intersection points is judged; In the case that the image pollution feature coordinates are located outside the preset area of the path intersection points, the pollution source candidate position corresponding to the path intersection point is deleted from the set of pollution source candidate positions to obtain the set of target pollution source candidate positions.

6. The method of determining the location of a pollution source according to claim 5, wherein, After the pollution source candidate position corresponding to the path intersection point is deleted from the set of pollution source candidate positions to obtain the set of target pollution source candidate positions, further including: A plurality of scoring indicators are determined, and each scoring indicator is respectively assigned a preset weight coefficient; Based on all the scoring indicators and all the preset weight coefficients, the preset pollution source scoring model is constructed.

7. The method of claim 1, wherein The identification result at least includes a probability score, and the scoring indicator at least includes a concentration scoring indicator, a direction scoring indicator, a clustering degree scoring indicator, and a confidence scoring indicator. A preset pollution source scoring model is used to score each pollution source candidate position in the target pollution source candidate position set to obtain a target score of each pollution source candidate position. For each pollution source candidate position, the distance from the pollution source candidate position to the core point of the cluster is calculated. A preset attenuation coefficient is determined, and based on the target heavy metal concentration data corresponding to the core point, the distance, and the preset attenuation coefficient, the scoring value of the concentration scoring indicator is determined. The vector difference between the pollution source candidate position and the core point is calculated, and the module of the vector difference is calculated. Based on the vector difference and the module, the unit vector between the pollution source candidate position and the core point is determined. Based on the unit vector and the concentration gradient vector of the core point, the direction angle cosine value is calculated, and the average value of the direction angle cosine value is taken as the scoring value of the direction scoring indicator. The number of paths within the preset range of the pollution source candidate position is counted, and the number of paths is taken as the scoring value of the clustering degree scoring indicator, and the probability score is taken as the scoring value of the confidence scoring indicator. All the scoring values are input into the preset pollution source scoring model to obtain the target score.

8. An apparatus for determining the location of a pollution source, characterized in that The method comprises the following steps: A construction unit is configured to collect environmental parameter data of a plurality of monitoring points in a target soil region and spectral image data of the target soil region, and construct a pollutant spatial distribution map based on all the environmental parameter data and the spectral image data. The pollutant spatial distribution map at least includes heavy metal concentration data of a plurality of grid points to be interpolated. A calculation unit is configured to cluster all the heavy metal concentration data in the pollutant spatial distribution map to obtain a cluster set, and calculate the concentration gradient direction of a core point of each cluster in the cluster set based on all the heavy metal concentration data. A screening unit is configured to obtain a plurality of environmental factor data of the target soil region, and based on all the environmental factor data and all the concentration gradient directions, determine a pollution source candidate position set by using a reverse particle tracking technology, and screen all pollution source candidate positions in the pollution source candidate position set to obtain a target pollution source candidate position set. A scoring unit is configured to score each pollution source candidate position in the target pollution source candidate position set by using a preset pollution source scoring model to obtain a target score of each pollution source candidate position, and determine the pollution source candidate position corresponding to the maximum target score as a pollution source position.

9. A computer program product, characterised in that, The non-volatile computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method for determining the pollution source position in any one of claims 1 to 7. The non-volatile computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method for determining the pollution source position in any one of claims 1 to 7.

10. An electronic device, comprising: The pollution source position determination method according to any one of claims 1 to 7 is implemented by one or more processors and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the pollution source position determination method.

Images