Pollutant prediction method and device, storage medium and electronic device

Abstract

The invention discloses a pollutant prediction method and device, a storage medium and an electronic device. The method comprises the following steps: respectively acquiring a plurality of depths of a pollutant to be predicted at a plurality of point locations, and determining a first target depth from the plurality of depths; determining a correlation coefficient between any two point locations in the plurality of point locations at the first target depth; based on the correlation coefficient and the plurality of point locations, constructing a knowledge graph corresponding to the to-be-predicted pollutant at the first target depth; and predicting the pollutant content of the to-be-predicted pollutant at the second target depth based on the knowledge graph and the pollutant content of the to-be-predicted pollutant at the first target depth. The technical problem that the efficiency of predicting the pollutant content is low is solved.

Description

Technical Field

[0001] The present disclosure relates to the field of cross-integration of artificial intelligence technology and environmental engineering, and specifically, to a pollutant prediction method, device, storage medium and electronic device. Background Art

[0002] At present, it is necessary to predict the pollutant content in order to formulate countermeasures to solve the pollution problem. In the existing technology, big data and artificial intelligence technology are usually used to monitor and prevent pollutant content. For example, by analyzing air quality data and calling machine learning models to predict air pollution levels. However, this method has data dimension constraints, which limits the effectiveness of artificial intelligence models, resulting in low efficiency in predicting pollutant content.

[0003] Currently, no effective solution has been proposed to the above-mentioned problem of low efficiency in predicting pollutant content. Summary of the invention

[0004] At least some embodiments of the present disclosure provide a method, device, storage medium and electronic device for predicting pollutants to at least solve the technical problem of low efficiency in predicting pollutant content.

[0005] According to one embodiment of the present disclosure, a method for predicting pollutants is provided. The method may include: obtaining multiple depths of the pollutants to be predicted at multiple points respectively, and determining a first target depth from the multiple depths; determining the correlation coefficient between any two points among the multiple points at the first target depth, wherein the correlation coefficient is used to indicate the degree of correlation between any two points; constructing a knowledge graph corresponding to the pollutants to be predicted at the first target depth based on the correlation coefficient and the multiple points, wherein the knowledge graph is used to indicate the degree of correlation between the multiple points, and the importance of the points to the distribution of the pollutants to be predicted at the first target depth; predicting the pollutant content of the pollutants to be predicted at the second target depth based on the knowledge graph and the pollutant content of the pollutants to be predicted at the first target depth, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths.

[0006] According to one embodiment of the present disclosure, a pollutant prediction device is also provided. The device may include: a first determination unit, used to respectively obtain multiple depths of the pollutants to be predicted at multiple points, and determine the first target depth from the multiple depths; a second determination unit, used to determine the correlation coefficient between any two points among the multiple points located at the first target depth, wherein the correlation coefficient is used to indicate the degree of correlation between any two points; a construction unit, used to construct a knowledge graph corresponding to the pollutants to be predicted at the first target depth based on the correlation coefficient and the multiple points, wherein the knowledge graph is used to indicate the degree of correlation between the multiple points, and the importance of the points to the distribution of the pollutants to be predicted at the first target depth; a prediction unit, used to predict the pollutant content of the pollutants to be predicted at the second target depth based on the knowledge graph and the pollutant content of the pollutants to be predicted at the first target depth, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths.

[0007] According to one embodiment of the present disclosure, a computer-readable storage medium is further provided, wherein a computer program is stored in the computer-readable storage medium, wherein the computer program is configured to execute any of the above-mentioned pollutant prediction methods when running.

[0008] According to one embodiment of the present disclosure, an electronic device is also provided, which may include a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any of the above pollutant prediction methods.

[0009] According to one embodiment of the present disclosure, a computer program product is also provided, which may include a computer program, wherein the computer program is configured to execute any of the above pollutant prediction methods when running.

[0010] According to one embodiment of the present disclosure, a computer program product is also provided. The computer program product may include a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and the computer program is configured to execute any of the above pollutant prediction methods when running.

[0011] According to one embodiment of the present disclosure, a computer program is also provided, which is configured to execute any of the above pollutant prediction methods when running.

[0012] In at least some embodiments of the present disclosure, multiple depths of the pollutant to be predicted at multiple points are obtained respectively, and the first target depth is determined from the multiple depths; the correlation coefficient between any two points among the multiple points located at the first target depth is determined, wherein the correlation coefficient is used to represent the degree of correlation between any two points; based on the correlation coefficient and the multiple points, a knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed, wherein the knowledge graph is used to represent the degree of correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths. That is, the embodiment of the present disclosure constructs a knowledge graph corresponding to the pollutant to be predicted at the first target depth, and predicts the pollutant content of the pollutant to be predicted at the second target depth based on the constructed knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, thereby achieving the technical effect of improving the efficiency of predicting the underground pollutant content and solving the technical problem of low efficiency in predicting the underground pollutant content. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The drawings described herein are used to provide a further understanding of the present disclosure and constitute a part of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure. In the drawings:

[0014] Figure 1 is a hardware structure block diagram of a mobile terminal of a pollutant prediction method according to an embodiment of the present disclosure;

[0015] Figure 2 is a flow chart of a method for predicting pollutants according to one embodiment of the present disclosure;

[0016] Figure 3 is a flowchart of constructing a knowledge graph according to one embodiment of the present disclosure;

[0017] Figure 4 is a schematic diagram of node degrees corresponding to pollution points according to one embodiment of the present disclosure;

[0018] Figure 5 is a schematic diagram of degree centrality of nodes corresponding to a contaminated point according to one embodiment of the present disclosure;

[0019] Figure 6 is a schematic diagram of the closed connection centrality of nodes corresponding to a contaminated point according to one embodiment of the present disclosure;

[0020] Figure 7is a schematic diagram of a clustering coefficient of nodes corresponding to a pollution point according to one embodiment of the present disclosure;

[0021] Figure 8 is a flow chart of determining association indicators based on a knowledge graph according to one embodiment of the present disclosure;

[0022] Fig. 9 is a schematic diagram of a first-layer pollutant copper concentration distribution map according to one embodiment of the present disclosure;

[0023] Fig.10 is a schematic diagram of a first-layer pollutant zinc concentration distribution map according to one embodiment of the present disclosure;

[0024] Fig.11 is a schematic diagram of the point distribution of cadmium concentration of soil pollutants in the first layer according to one embodiment of the present disclosure;

[0025] Fig.12 is a schematic diagram of a first layer soil petroleum hydrocarbon concentration network according to one embodiment of the present disclosure;

[0026] Fig.13 is a schematic diagram of the distribution of copper contamination nodes according to one embodiment of the present disclosure;

[0027] Fig.14 is a schematic diagram of the degree centrality distribution of copper network nodes according to one embodiment of the present disclosure;

[0028] Fig.15 is a schematic diagram of the closed connection centrality distribution of copper network nodes according to one embodiment of the present disclosure;

[0029] Fig.16 It is a schematic diagram of the comparison between the actual value and the predicted value of the copper concentration predicted by LSTM according to one embodiment of the present disclosure;

[0030] Fig.17 is a schematic diagram of a comparison between a true value and a predicted value of petroleum hydrocarbon concentration predicted by SVM according to one embodiment of the present disclosure;

[0031] Fig.18 is a structural block diagram of a pollutant prediction device according to one embodiment of the present disclosure;

[0032] Fig.19 is a schematic diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the disclosed solution, the technical solution in the disclosed embodiment will be clearly and completely described below in conjunction with the drawings in the disclosed embodiment. Obviously, the described embodiment is only a part of the disclosed embodiment, not all of the embodiments. Based on the embodiments in the disclosed embodiment, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the disclosed embodiment.

[0034] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0035] According to one embodiment of the present disclosure, an embodiment of a method for predicting pollutants is provided. It should be noted that the steps shown in the flowchart of 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 can be executed in an order different from that shown here.

[0036] The method embodiment can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, the mobile terminal can be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, referred to as MID), a tablet computer (PAD) and other terminal devices. Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for predicting pollutants according to an embodiment of the present disclosure. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1Only one is shown in the figure) processor 102 (the processor 102 may include but is not limited to a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP) chip, a microcontroller unit (MCU), a programmable logic device (Field Programmable Gate Array, FPGA), a neural network processor (NPU), a tensor processor (TPU), an artificial intelligence (AI) type processor, etc.) and a memory 104 for storing data. Optionally, the mobile terminal may further include a transmission device 106, an input and output device 108, and a display device 110 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0037] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the pollutant prediction method in the embodiment of the present disclosure. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above-mentioned pollutant prediction method is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0038] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0039] The input in the input / output device 108 may come from multiple human interface devices (HID), such as keyboard and mouse. Some HIDs may provide output functions in addition to input functions, such as audio output of a controller.

[0040] The display device 110 may be, for example, a Heads Up Display (HUD), a touch screen liquid crystal display (LCD), and a touch display (also referred to as a "touch screen" or "touch display screen"). The LCD display enables the user to interact with the user interface of the mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), and the user can interact with the GUI by finger contacts and / or gestures on the touch-sensitive surface. The human-computer interaction functions here may optionally include the following interactions: creating web pages, drawing, word processing, making electronic documents, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital videos, playing digital music and / or web browsing, etc. The executable instructions for executing the above human-computer interaction functions are configured / stored in a computer program product or readable storage medium executable by one or more processors.

[0041] Those skilled in the art will understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0042] In a possible implementation, the present disclosure provides a method for predicting pollutants, which may be a method for predicting pollutant content. Figure 2 is a flow chart of a method for predicting pollutants according to one embodiment of the present disclosure, such as Figure 2As shown, the method may include the following steps:

[0043] Step S202 , respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths.

[0044] In the technical solution provided in the above step S202 of the present disclosure, multiple depths of the pollutants to be predicted at multiple points can be obtained respectively, and the first target depth can be determined from the multiple depths. Among them, the pollutants to be predicted can be pollutants to be predicted, and the pollutants can be excessive pollutants, and the excessive pollutants can at least include organic pollutants and inorganic pollutants. Organic pollutants can at least include but are not limited to pentachlorophenol, polycyclic aromatic hydrocarbons, petroleum hydrocarbons, benzanthracene, benzopyrene, benzofluoranthene, dibenzanthracene, indenopyrene and naphthalene, etc., and inorganic pollutants can at least include but are not limited to copper, zinc, cadmium, mercury, hexavalent chromium, arsenic, nickel, lead and cyanide, etc. This is only for illustration, and no specific restrictions are made on the types of pollutants to be predicted.

[0045] In this embodiment, the above-mentioned points can be sampling points of the polluted site, and the polluted site can be a site where pollutants exist, or a site where pollutants may exist. The depth can be the exceeding depth of the pollutant to be predicted, for example, the exceeding depth can be 0 to 25 meters (m). The first target depth can be a depth determined from the above-mentioned multiple depths, for example, the multiple depths are 5m, 10m and 15m, etc., and 10m can be selected from multiple depths such as 5m, 10m and 15m, and the 10m is determined as the first target depth. This is only for example, and no specific restrictions are imposed on the determination method and value of the first target depth.

[0046] Optionally, pollutant monitoring data collected by scientific research institutions and data providers can be obtained, and pollutant monitoring data can also be obtained in real time using tools such as sensor networks and satellite remote sensing. After obtaining the pollutant monitoring data, multiple depths of the pollutant to be predicted at multiple points can be obtained from the obtained pollutant monitoring data. For example, if the pollutant is predicted to be copper, multiple depths of copper at the first point and the second point can be obtained. Further, the first target depth can be determined from the multiple depths obtained.

[0047] Step S204: determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points.

[0048] In the technical solution provided in the above step S204 of the present disclosure, after obtaining multiple depths of the pollutants to be predicted at multiple points respectively and determining the first target depth from the multiple depths, the correlation coefficient between any two points in the multiple points located at the first target depth can be determined. The correlation coefficient can be used to represent the degree of correlation between any two points, can be used to measure the degree of difference in the pollutant content between any two points, can be called correlation, and can be represented by V.

[0049] In this embodiment, after the first target depth is determined from the multiple depths, the correlation coefficient between any two points in the multiple points located at the first target depth can be determined by the following formula:

[0050]

[0051] Among them, V can be used to represent the correlation coefficient, a can be used to represent the pollutant content at point A, which is also called the pollutant concentration, and b can be used to represent the pollutant concentration at point B, where point A and point B have the same pollutant. The above formula can be used to measure the difference in pollutant concentration between point A and point B, thereby determining the correlation between point A and point B.

[0052] For example, suppose that point A has a copper concentration of a, and point B has a copper concentration of b, and the concentration unit is milligrams per kilogram (mg / kg). For example, the copper concentrations of points A and B in the first layer are 50 mg / kg and 27 mg / kg respectively, then the correlation coefficient V between points A and B can be expressed by the following formula:

[0053]

[0054] Step S206, based on the correlation coefficient and multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the degree of correlation between multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth.

[0055] In the technical solution provided in the above step S206 of the present disclosure, after determining the correlation coefficient between any two points among the multiple points located at the first target depth, a knowledge graph corresponding to the pollutant to be predicted at the first target depth can be constructed based on the determined correlation coefficient and the multiple points obtained. The knowledge graph can be used to represent the degree of correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth.

[0056] In this embodiment, the obtained multiple points can be respectively determined as multiple nodes of the knowledge graph to be constructed, and the two are in a one-to-one correspondence. Further, based on the correlation coefficient between any two points, it can be determined whether the two nodes corresponding to the any two points are connected by an edge, so as to construct the knowledge graph corresponding to the pollutant to be predicted at the first target depth based on the nodes and edges.

[0057] Step S208, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, wherein the second target depth is the next depth layer of the first target depth among multiple depths.

[0058] In the technical solution provided in the above step S208 of the present disclosure, after constructing the knowledge graph corresponding to the pollutant to be predicted at the first target depth based on the correlation coefficient and multiple points, the pollutant content of the pollutant to be predicted at the second target depth can be predicted based on the constructed knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth. The second target depth can be the next depth of the first target depth among multiple depths. For example, if the first target depth is 10m, the second target depth can be 15m. This is only for example, and no specific limitation is imposed on the value of the second target depth.

[0059] In this embodiment, after constructing the knowledge graph corresponding to the pollutant to be predicted at the first target depth, multiple associated indicators of the pollutant to be predicted can be determined based on the constructed knowledge graph. After determining the multiple associated indicators of the pollutant to be predicted, the initial prediction model can be trained using the associated indicators to obtain a prediction model. The prediction model is further called to predict the pollutant content of the pollutant to be predicted at the second target depth.

[0060] Through the above steps S202 to S208, multiple depths of the pollutant to be predicted at multiple points are obtained respectively, and the first target depth is determined from the multiple depths; the correlation coefficient between any two points among the multiple points located at the first target depth is determined, wherein the correlation coefficient is used to indicate the degree of correlation between any two points; based on the correlation coefficient and the multiple points, a knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed, wherein the knowledge graph is used to indicate the degree of correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths. That is to say, the embodiment of the present disclosure constructs a knowledge graph corresponding to the pollutant to be predicted at the first target depth, and predicts the pollutant content of the pollutant to be predicted at the second target depth based on the constructed knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, thereby achieving the technical effect of improving the efficiency of predicting the underground pollutant content and solving the technical problem of low efficiency in predicting the underground pollutant content.

[0061] The above method of the embodiment of the present disclosure is further introduced below.

[0062] This embodiment first introduces the method for constructing a knowledge graph as follows.

[0063] As an optional embodiment, step S206, based on the correlation coefficient and multiple points, constructs a knowledge graph corresponding to the pollutants to be predicted at the first target depth, including: respectively determining the multiple points as nodes of the knowledge graph to be constructed to obtain multiple nodes; from the multiple nodes, determining two nodes corresponding to any two points, and based on the correlation coefficient between any two points, determining the correlation result between the two nodes, wherein the correlation result is used to indicate that the correlation between the two nodes is allowed, or the correlation between the two nodes is prohibited; constructing a knowledge graph based on the multiple nodes and the correlation result between the two nodes.

[0064] In this embodiment, multiple points can be determined as nodes of the knowledge graph to be constructed, and multiple nodes can be obtained. After determining the correlation coefficient between any two points among the multiple points located at the first target depth, two nodes corresponding to the arbitrary two points can be determined from the multiple nodes, and based on the determined correlation coefficient between the arbitrary two points, the correlation result between the two nodes can be determined. After obtaining multiple nodes and determining the correlation result between the two nodes, a knowledge graph can be constructed based on the correlation results between the multiple nodes and the two nodes. Among them, the correlation result can be used to indicate that the association between the two nodes is allowed, or the association between the two nodes is prohibited.

[0065] Optionally, after obtaining multiple points of pollutants to be predicted, the multiple points can be respectively determined as nodes of the knowledge graph to be constructed to obtain multiple nodes. Further, after determining the correlation coefficient between any two points among the multiple points, two nodes corresponding to the above-mentioned any two points can be determined from the multiple nodes, and based on the correlation coefficient between the any two points, the correlation result between the two nodes is determined. Since the determined correlation result can be used to indicate that the association between the two nodes is allowed, or the association between the two nodes is prohibited, that is, the above-mentioned association result is used to indicate that the edge connection between the two nodes is allowed, or the edge connection between the two nodes is prohibited, a knowledge graph can be constructed based on the connection relationship between the obtained multiple nodes and the edges between the nodes.

[0066] This embodiment constructs a knowledge graph and uses the structured representation of the knowledge graph to analyze the connection relationship between the nodes in the knowledge graph. Since the nodes in the knowledge graph correspond to the points of the contaminated site, the association relationship between the points can be determined by analyzing the connection relationship between the nodes. For example, the degree of influence between the points can be determined to achieve the purpose of measuring underground pollutants at points with a higher degree of influence, thereby achieving the technical effect of improving the efficiency of predicting the content of underground pollutants.

[0067] The method for determining the correlation result between two nodes based on the correlation coefficient between any two points is further introduced below.

[0068] As an optional embodiment, based on the correlation coefficient between any two points, the correlation result between two nodes is determined, including: in response to the correlation coefficient being less than the correlation coefficient threshold, determining that the correlation between the two nodes is allowed; in response to the correlation coefficient being greater than or equal to the correlation coefficient threshold, determining that the correlation between the two nodes is prohibited.

[0069] In this embodiment, after determining the correlation coefficient between any two points among the multiple points, when the correlation coefficient between any two points is less than the correlation coefficient threshold, in response to the correlation coefficient being less than the correlation coefficient threshold, it can be determined that the two nodes are allowed to be associated. When the correlation coefficient between any two points is greater than or equal to the correlation coefficient threshold, in response to the correlation coefficient being greater than or equal to the correlation coefficient threshold, it can be determined that the two nodes are prohibited from being associated. Among them, the correlation coefficient threshold can be a threshold set according to actual conditions, which can be represented by Vα. The correlation coefficient threshold has a certain elastic space. For example, the value interval of the correlation coefficient threshold can be (0.05, 0.12). This is only for example description, and no specific restriction is made on the value of the correlation coefficient threshold.

[0070] It should be noted that the correlation coefficient threshold can be an important parameter for screening the correlation between two points. The setting of the correlation coefficient threshold can be used to screen out pairs of points with relatively high correlation, so that the edges of the constructed knowledge graph will not be too sparse or too dense. If the correlation coefficient threshold is set too large, there will be fewer edges, while if the correlation coefficient threshold is set too small, there will be more edges.

[0071] Optionally, after calculating the correlation coefficient V between the above-mentioned point A and point B, when the correlation coefficient V is less than the set correlation coefficient threshold Vα, it is considered that there is a correlation between point A and point B, and the two nodes corresponding to point A and point B can be connected. When the correlation coefficient V is greater than or equal to the set correlation coefficient threshold Vα, it is considered that there is no correlation between point A and point B, and the connection of the two nodes corresponding to point A and point B is prohibited.

[0072] It should be noted that the larger the correlation coefficient V, the weaker the correlation between the two points in terms of pollutants, and the smaller the correlation coefficient V, the stronger the correlation between the two points in terms of pollutants.

[0073] For example, assume that the correlation coefficient threshold Vα = 3, the copper concentrations of the first-layer point A and point B are 50 mg / kg and 27 mg / kg respectively. After calculating that the correlation coefficient V between point A and point B is V = 2.621, since V < Vα, it indicates that the correlation of the copper content between point A and point B exceeds the expectation, and it is determined that the connection of the edges between the two nodes corresponding to point A and point B is allowed.

[0074] Next, the method for constructing a knowledge graph based on the association results between multiple nodes and two nodes will be further introduced.

[0075] As an optional embodiment, constructing a knowledge graph based on the association results between multiple nodes and two nodes includes: traversing the association results between any two nodes among the multiple nodes, and adding any two nodes with an allowable association result to the edge matrix; connecting any two nodes according to the edge matrix to obtain the knowledge graph.

[0076] In this embodiment, after determining the association result between two nodes based on the association coefficient between any two points, the association result between any two nodes in the multiple nodes can be traversed, and any two nodes whose association result is to allow association can be added to the edge matrix. Further, any two nodes can be connected according to the edge matrix to obtain a knowledge graph. Among them, the edge matrix can be a matrix composed of multiple starting node and ending node pairs. For example, the edge matrix can be [[1, 3], [2, 5], [4, 8]], which indicates that the starting node 1 is connected to the ending node 3, the starting node 2 is connected to the ending node 5, and the starting node 4 is connected to the ending node 8. This is only for example, and no specific limitation is made on the representation of the edge matrix.

[0077] Optionally, after determining the association result between two nodes based on the association coefficient between any two points, a two-layer loop can be used to traverse all combinations of points, and the correlation between any two points can be calculated according to the association coefficient formula. If the association coefficient is less than the association coefficient threshold, the corresponding two nodes are added to the edge matrix. Further, the two nodes can be connected according to the edge matrix to obtain a knowledge graph.

[0078] This embodiment introduces a method for predicting the pollutant content of the pollutant to be predicted at the second target depth as follows.

[0079] As an optional embodiment, step S208, based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, predicts the pollutant content of the pollutant to be predicted at the second target depth, including: determining the association index of the pollutant to be predicted based on the knowledge graph, wherein the association index is used to represent the structural characteristics of the pollutant to be predicted; based on the association index and the pollutant content of the pollutant to be predicted at the first target depth, predicting the pollutant content of the pollutant to be predicted at the second target depth.

[0080] In this embodiment, after constructing the knowledge graph corresponding to the pollutant to be predicted at the first target depth based on the correlation coefficient and multiple points, the correlation index of the pollutant to be predicted can be determined based on the constructed knowledge graph. Further, based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth can be predicted. Among them, the correlation index can be used to represent the structural characteristics of the pollutant to be predicted, and the correlation index can at least include node degree, degree centrality, clustering coefficient, closure centrality and information centrality.

[0081] Optionally, after constructing the knowledge graph, an association index can be determined based on the number of edges connecting nodes in the knowledge graph and the degree of influence of each node on other nodes. The association index refers to the structural characteristics of pollutants at different depths of each node, rather than the structural characteristics of the network as a whole. Further, based on the association index and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth can be predicted.

[0082] The following is a further introduction to the method for predicting the pollutant content of the pollutant to be predicted at the second target depth based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth.

[0083] As an optional embodiment, based on the associated indicators and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted, including: based on the associated indicators and the pollutant content of the pollutant to be predicted at the first target depth, a prediction model is constructed, wherein the prediction model is used to predict the pollutant content; calling the prediction model to analyze the associated indicators and the pollutant content of the pollutant to be predicted at the first target depth, to obtain the pollutant content of the pollutant to be predicted at the second target depth.

[0084] In this embodiment, after determining the associated index of the pollutant to be predicted based on the constructed knowledge graph, a prediction model can be constructed based on the determined associated index and the pollutant content of the pollutant to be predicted at the first target depth. The constructed prediction model can be further called to analyze the associated index and the pollutant content of the pollutant to be predicted at the first target depth to obtain the pollutant content of the pollutant to be predicted at the second target depth.

[0085] Optionally, after determining the associated index of the pollutant to be predicted based on the constructed knowledge graph, a prediction model can be constructed based on the determined associated index and the pollutant content of the pollutant to be predicted at the first target depth. For example, the determined associated index and the pollutant content of the pollutant to be predicted at the first target depth can be used as input to construct the prediction model. After the prediction model is constructed, the constructed prediction model can be called to analyze the associated index and the pollutant content of the pollutant to be predicted at the first target depth to obtain the pollutant content of the pollutant to be predicted at the second target depth.

[0086] The following is a further introduction to the method for constructing a prediction model based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth.

[0087] As an optional embodiment, a prediction model is constructed based on associated indicators and the pollutant content of the pollutant to be predicted at the first target depth, including: merging the associated indicators with the pollutant content of the pollutant to be predicted at the first target depth to obtain a comprehensive indicator of the pollutant to be predicted; and using the comprehensive indicator to train the initial prediction model to obtain a prediction model.

[0088] In this embodiment, after determining the associated index of the pollutant to be predicted based on the constructed knowledge graph, the associated index can be combined with the pollutant content of the pollutant to be predicted at the first target depth to obtain a comprehensive index of the pollutant to be predicted. The comprehensive index can be further used to train the initial prediction model to obtain a prediction model.

[0089] Optionally, after merging the associated indicators with the pollutant content of the pollutant to be predicted at the first target depth to obtain a comprehensive indicator of the pollutant to be predicted, a supervised machine learning method such as an artificial neural network, a support vector machine (SVM), a long short-term memory network (LSTM), etc. is selected, with the comprehensive indicator as input data and the next layer of content as a label to build a prediction model.

[0090] Optionally, this embodiment introduces the following method for determining the node degree, degree centrality, clustering coefficient, closure centrality and information centrality of the pollutants to be predicted based on the knowledge graph.

[0091] As an optional embodiment, the association index includes node degree, and the node degree is used to represent the number of connections between nodes in the knowledge graph. Based on the knowledge graph, the association index of the pollutant to be predicted is determined, including: determining the number of edges connected to the node in the knowledge graph as the node degree corresponding to the node.

[0092] In this embodiment, the association index may include node degree, which can be used to represent the number of connections between nodes in the knowledge graph. After constructing the knowledge graph, the number of edges connected to the nodes in the knowledge graph can be determined as the node degree corresponding to the node. The degree of node i can be expressed as k. i To express.

[0093] Alternatively, the node degree refers to the number of edges directly connected to the node, which can be expressed by the following formula:

[0094]

[0095] Among them, k ican be used to represent the degree of node i, and n can be used to represent the total number of nodes in the knowledge graph. For example, the knowledge graph includes node 49, which is connected to six edges, and the degree of node 49 is 6.

[0096] As an optional embodiment, the association index includes degree centrality, which is used to represent the connection strength of the node. Based on the knowledge graph, the association index of the pollutant to be predicted is determined, including: determining the quotient between the node degree and the target number of nodes as the degree centrality corresponding to the node, wherein the target number is determined by the total number of nodes.

[0097] In this embodiment, the association index may include degree centrality, which can be used to represent the connection strength of a node. After constructing the knowledge graph, the quotient between the node degree and the target number of nodes can be determined as the degree centrality corresponding to the node. The degree centrality of node i can be expressed as C D (i) The total number of nodes in the knowledge graph can be represented by n. The target number can be determined by the total number of nodes, for example, the target number can be n-1.

[0098] Optionally, degree centrality refers to the ratio of the node degree to the number of other nodes in the knowledge graph, which can be expressed by the following formula:

[0099]

[0100] Among them, C D (i) can be used to represent the degree centrality of node i, k can be used to represent the node degree, and n can be used to represent the total number of nodes in the network. For example, if the degree of node 42 in the knowledge graph is 7, and the total number of other nodes in the knowledge graph is 5, then the degree centrality of node 42 is 7 / 5.

[0101] As an optional embodiment, the association index includes a clustering coefficient, which is used to indicate the degree of clustering of nodes. Based on the knowledge graph, the association index of the pollutant to be predicted is determined, including: using the node degree to adjust the number of edges connected to the node's neighbor nodes to obtain the clustering coefficient corresponding to the node.

[0102] In this embodiment, the association index may include a clustering coefficient, which may be used to indicate the degree of clustering of nodes. After constructing the knowledge graph, the node degree may be used to adjust the number of edges connected to the neighboring nodes of the node to obtain the clustering coefficient corresponding to the node. The clustering coefficient of node i may be represented by C(i).

[0103] Optionally, the clustering coefficient can be used to measure the degree to which the neighboring nodes of a node are connected. The clustering coefficient can be expressed by the following formula:

[0104]

[0105] Among them, C(i) can be used to represent the clustering coefficient of node i, e i It can be used to represent the number of connected edges between neighbor nodes of node i, k i It can be used to represent the degree of node i. For example, the degree of node 53 in the knowledge graph is 5, and the neighbor nodes of node 53 are node 52 and node 66, where the number of edges connected to node 52 is 4 and the number of edges connected to node 66 is 7, then the clustering coefficient of node 53 is 22 / 20=1.1.

[0106] As an optional embodiment, the association index includes closed connection centrality, which is used to represent the distance between any two nodes in the knowledge graph. Based on the knowledge graph, the association index of the pollutant to be predicted is determined, including: determining the shortest path length between any two nodes to obtain multiple shortest path lengths; determining the average value of the sum of multiple shortest path lengths; and determining the inverse of the average value as the closed connection centrality corresponding to the node.

[0107] In this embodiment, the association index may include closure centrality, which can be used to represent the distance between any two nodes in the knowledge graph. After constructing the knowledge graph, the shortest path length between any two nodes can be determined to obtain multiple shortest path lengths. Further, the average value of the sum of the multiple shortest path lengths can be determined, and the reciprocal of the average value is determined as the closure centrality corresponding to the node. The closure centrality of node i can be expressed as C C The shortest path length from node i to node j can be represented by d(i,j).

[0108] Alternatively, closure centrality can be used to measure the inverse of the average distance from a node to other nodes. Closure centrality can be expressed as follows:

[0109]

[0110] Among them, C C (i) can be used to represent the closed centrality of node i, and d(i,j) can be used to represent the shortest path length from node i to node j.

[0111] As an optional embodiment, the association index includes information centrality, which is used to indicate the importance of a node in a knowledge graph in a shortest path. Based on the knowledge graph, the association index of the pollutant to be predicted is determined, including: determining the number of first shortest paths between any two nodes in the knowledge graph, and determining the number of second shortest paths passing through the nodes from the number of first shortest paths; and determining the quotient between the number of second shortest paths and the number of first shortest paths as the information centrality corresponding to the node.

[0112] In this embodiment, the association index may include information centrality, which may be used to indicate the importance of a node in the knowledge graph in the shortest path. After constructing the knowledge graph, the number of first shortest paths between any two nodes in the knowledge graph may be determined, and the number of second shortest paths passing through the nodes may be determined from the number of first shortest paths. The quotient between the number of second shortest paths and the number of first shortest paths may be further determined as the information centrality corresponding to the node. The information centrality of node i may be expressed as C B (i) is used to represent the first shortest path number. The first shortest path number can be the shortest path number from node s to node t, which can be represented by σ st The second shortest path number can be the number of shortest paths from node s to node t that pass through node i, which can be expressed by σ st (i) To express.

[0113] Alternatively, information centrality can be used to measure the importance of a node in all shortest paths in the knowledge graph, which can be expressed by the following formula:

[0114]

[0115] Among them, C B (i) can be used to represent the information centrality of node i, σ st It can be used to represent the number of shortest paths from node s to node t, σ st (i) can be used to represent the number of shortest paths passing through node i.

[0116] This embodiment can determine the key nodes in the knowledge graph by analyzing the degree distribution, degree centrality, and closed connection centrality of the nodes in the knowledge graph. For example, nodes with high degree distribution may be important monitoring sites, and nodes with high degree centrality and closed connection centrality may be key pollution transmission points. By predicting the underground pollutant content at the points corresponding to the key nodes, the technical effect of improving the efficiency of predicting the underground pollutant content is achieved.

[0117] As an optional embodiment, the method also includes: obtaining monitoring data of the pollutant to be predicted, wherein the monitoring data is used to indicate the distribution state of the pollutant content of the pollutant to be predicted at multiple depths; and preprocessing the monitoring data to obtain the pollutant content of the pollutant to be predicted at a first target depth.

[0118] In this embodiment, monitoring data of the pollutant to be predicted can be obtained. After obtaining the monitoring data of the pollutant to be predicted, the monitoring data can be preprocessed to obtain the pollutant content of the pollutant to be predicted at the first target depth. The monitoring data can be used to indicate the distribution state of the pollutant content of the pollutant to be predicted at multiple depths.

[0119] Optionally, after obtaining the monitoring data of the pollutants to be predicted, the acquired monitoring data usually has various problems, such as missing data, inconsistent format and low accuracy. Therefore, after obtaining the monitoring data, it is necessary to clean and preprocess the monitoring data and convert the monitoring data into a unified format and structure. Secondly, use data integration technology to integrate monitoring data from different sources into a unified database for subsequent analysis and query.

[0120] In the embodiment of the present disclosure, multiple depths of the pollutant to be predicted at multiple points are obtained respectively, and the first target depth is determined from the multiple depths; the correlation coefficient between any two points among the multiple points located at the first target depth is determined, wherein the correlation coefficient is used to represent the degree of correlation between any two points; based on the correlation coefficient and the multiple points, a knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed, wherein the knowledge graph is used to represent the degree of correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths. That is, the embodiment of the present disclosure constructs a knowledge graph corresponding to the pollutant to be predicted at the first target depth, and predicts the pollutant content of the pollutant to be predicted at the second target depth based on the constructed knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, thereby achieving the technical effect of improving the efficiency of predicting the underground pollutant content and solving the technical problem of low efficiency in predicting the underground pollutant content.

[0121] The technical solution of the embodiment of the present disclosure is further introduced below with reference to preferred implementation modes.

[0122] At present, with the rapid development of industrialization and urbanization, groundwater and soil pollution problems are becoming increasingly serious, and the emission and spread of pollutants have become one of the focuses of attention. Groundwater and soil pollution are closely related to factors such as urbanization, industrialization and population density. Therefore, effective means are needed to quickly and accurately discover the pollution source and its diffusion process in order to scientifically formulate countermeasures.

[0123] In the field of environmental protection, the application of data technology has become increasingly important. With the intensification of global environmental problems, such as climate change, air and water pollution, and biodiversity loss, the use of data technology for environmental monitoring, analysis, and governance has become a key way to achieve sustainable development. On the one hand, remote sensing technology and geographic information system (GIS) are widely used in environmental monitoring and resource management. Through satellite imagery and spatial data analysis, environmental changes such as deforestation, land use changes, and urban expansion can be tracked. These technologies provide scientific basis and data support for the formulation of environmental protection policies and management measures. On the other hand, big data and artificial intelligence technologies are gradually being applied to the monitoring and prevention of environmental pollution. For example, by analyzing the data from air quality monitoring stations, calling machine learning models can predict air pollution levels, so that timely measures can be taken to reduce pollution and health risks. In addition, by combining social media data, IoT devices, and online monitoring systems, pollution sources can be tracked and analyzed in real time, thereby more effectively guiding pollution prevention and environmental governance strategies.

[0124] In addition, data technology also plays an important role in promoting environmental science research and environmental education. By establishing and utilizing environmental databases and knowledge bases, environmental issues can be studied more systematically and new environmental protection technologies and methods can be developed. At the same time, through data visualization and environmental simulation tools, environmental issues and protection measures can be understood more intuitively, and environmental protection awareness and participation can be enhanced.

[0125] With the advancement of technology, the application of predictive models in artificial intelligence in the field of environmental protection is becoming more and more extensive. By integrating environmental monitoring data, economic and social statistics and model predictions, it can provide multi-faceted information and decision-making support, help evaluate the effectiveness and cost-effectiveness of different environmental policies and management measures, and make more reasonable and effective decisions.

[0126] However, although the application of artificial intelligence technology in environmental protection has brought many positive effects, there are also problems such as data dimension constraints that limit the effectiveness of artificial intelligence models. Therefore, how to effectively utilize the data resources provided by existing monitoring methods and effectively improve the prediction effect by expanding dimensions and other methods has become an urgent problem to be solved.

[0127] In order to solve the above problems, this paper proposes a pollutant content prediction method based on underground pollution knowledge graph, constructs pollutant knowledge graph according to underground environmental pollution monitoring results, extracts characteristic indicators from the knowledge graph, and integrates the extracted characteristic indicators with monitoring indicators to obtain a new indicator set. Further based on the new indicator set, a deeper underground pollution prediction model is established.

[0128] This embodiment combines the knowledge graph with pollution events and proposes a new hybrid knowledge representation, which not only includes static entities and attributes, but also integrates events and event-related dynamic information, providing a richer and more comprehensive analysis dimension for environmental monitoring. Through the above combination, we can better understand the occurrence and evolution of environmental pollution events, so as to formulate monitoring and governance strategies in a more targeted manner. By using the pollutant knowledge graph as a decision support tool, new perspectives and methods are provided for the monitoring and governance of environmental pollution. Through the structured representation of the knowledge graph, we can have a more comprehensive understanding of the underground environmental pollution in a certain area. By expanding the application of knowledge graphs in underground environmental monitoring, new technical means and decision support tools are provided for pollution control, which helps to improve the efficiency and effectiveness of environmental monitoring and pollution control.

[0129] All sampling points are sorted into n points and divided into m depths. A knowledge graph is established for each pollutant at each depth. For example, if there are 17 types of pollutants, 85 knowledge graphs can be established. Each layer of the knowledge graph represents a set of sample points at a specific depth. After cleaning, segmenting, and sorting the relevant data that collects pollution patterns, a multi-layer knowledge graph is established, and each layer represents a data view at a different depth. The knowledge graph can provide structured information for further analysis and provide a basis for later indicator expansion.

[0130] A monitoring point of each pollutant at a certain depth is used as a node in the underground pollutant knowledge graph. The point and the node are in a one-to-one correspondence, and repeated correspondence is not allowed. The specific process of node extraction in the knowledge graph can be to select a pollutant, determine the point, determine the depth, and extract the point as a node of the knowledge graph. Further correlation calculation of the extracted points can be expressed by the following correlation coefficient formula:

[0131]

[0132] Among them, V can be used to represent the correlation coefficient, and a and b can be used to represent the concentrations of the same pollutant at point A and point B, respectively. The above formula can be used to measure the difference in pollutant concentrations between two points, thereby determining the correlation between the two points. After determining the correlation between two points, a threshold needs to be set. The threshold can be an important parameter for screening correlations. The setting of the threshold can be used to screen out pairs of points with higher correlations, so that the edges in the network diagram are not too sparse or too dense. Setting the threshold too large will result in fewer edges, while setting the threshold too small will result in more edges.

[0133] For example, assume that point A and point B have copper concentrations of a and b, respectively, and the concentration unit is milligrams per kilogram (mg / kg). After calculating the correlation coefficient V between the above two points, when the correlation coefficient V is less than the set threshold Vα, it is considered that there is a correlation between point A and B, and the two points A and B can be connected in the network diagram. For example, the first-layer point A has a copper concentration of 50 mg / kg, and point B has a copper concentration of 27 mg / kg. Then the correlation coefficient V between point A and point B can be expressed by the following formula:

[0134]

[0135] The larger the correlation coefficient V, the weaker the correlation between the two points in terms of the pollutant copper. The smaller the correlation coefficient V, the stronger the correlation between the two points in terms of the pollutant copper. If the threshold is set to Vα = 3, since V < Vα, it indicates that the correlation of the copper content between point A and point B exceeds the expectation, and the edge between the two is added to the edge matrix. Two nested loops are used in the program to traverse all combinations of points, calculate the correlation according to the correlation coefficient formula, and add these points to the edge matrix if they meet the conditions, thereby constructing the network diagram.

[0136] Figure 3 is a flowchart of constructing a knowledge graph according to an embodiment of the present disclosure, as Figure 3 shown. The process of constructing the knowledge graph may at least include the following steps:

[0137] Step S301, obtain the concentration value of the pollutant at point i.

[0138] In the above step S301, the monitoring data of the pollutant can be obtained. After obtaining the monitoring data of the pollutant, the concentration value of the pollutant at point i can be read from the obtained monitoring data, and this concentration value can be represented by ρ i for representation.

[0139] Step S302, obtain the point j with the serial number equal to zero.

[0140] In the above step S302, the point number j=0 may be read from a plurality of points.

[0141] Step S303, determine whether i is equal to j.

[0142] In the above step S303, when it is determined that i is equal to j, the process proceeds to step S304, otherwise, the process proceeds to step S305.

[0143] Step S304, accumulating j.

[0144] In the above step S304, when i is equal to j, let j=j+1.

[0145] Step S305, obtaining the concentration value of the pollutant at point j.

[0146] In the above step S305, when i is not equal to j, or after j is accumulated, the concentration value of the pollutant at point j can be read, and the concentration value can be expressed as ρ j To express.

[0147] Step S306, determining whether the correlation coefficient between point i and point j is greater than a correlation coefficient threshold.

[0148] In the above step S306, the concentration value ρ of the pollutant at the reading point i is i , and the concentration value of the pollutant at point j ρ j After that, the concentration value ρ can be calculated by the following formula i With concentration value ρ j To calculate the correlation:

[0149]

[0150] Among them, ε can be used to represent a smaller value, which can be used to prevent the abnormal situation that the denominator in the above formula appears to be zero. θ can be used to represent the correlation coefficient threshold. i With concentration value ρ j After the calculated correlation coefficient is less than the correlation coefficient threshold θ, the process proceeds to step S307; otherwise, the process proceeds to step S304.

[0151] Step S307, connect point i and point j with an edge.

[0152] In the above step S307, when the calculated correlation coefficient between point i and point j is less than the correlation coefficient threshold, point i and point j are connected. When the calculated correlation coefficient between point i and point j is greater than or equal to the correlation coefficient threshold, j is further accumulated.

[0153] The input data of the above process is the concentration value and the correlation coefficient threshold at each pollutant point, and the output data is the edge between the pollutant point i and other nodes. After executing the above process, the knowledge graph of the pollutant can be obtained.

[0154] After constructing the knowledge graph, association indicators can be extracted based on the knowledge graph. The indicators are the structural characteristics of pollutants at different depths of each node, rather than the structural characteristics of the network as a whole. The extracted association indicators of pollutants can at least include: node degree, degree centrality, closure centrality, information centrality and clustering coefficient.

[0155] The node degree refers to the number of edges directly connected to the node, which can be expressed by the following formula:

[0156]

[0157] Among them, k i It can be used to represent the degree of node i, and n can be used to represent the total number of nodes in the network. Figure 4 is a schematic diagram of node degrees corresponding to pollution points according to one embodiment of the present disclosure, such as Figure 4 As shown, node 39, node 49 and node 59 represent sampling points of the contaminated site, and node 49 is connected to six edges, so the degree of node 49 is 6, while node 39 and node 59 are not connected by edges, so the degrees of node 39 and node 59 are both 0.

[0158] Degree centrality refers to the ratio of a node's degree to the number of other nodes in the network. Degree centrality can be expressed by the following formula:

[0159]

[0160] Among them, C D (i) can be used to represent degree centrality, k can be used to represent node degree, and n can be used to represent the total number of nodes in the network. Figure 5 is a schematic diagram of the degree centrality of nodes corresponding to a contaminated point according to one embodiment of the present disclosure, such as Figure 5 As shown, node 42, node 45, node 58, node 60, node 52 and node 40 represent sampling points of the contaminated site, the degree of node 42 is 7, and the total number of other nodes in the network is 5, then the degree centrality of node 42 is 7 / 5.

[0161] Closeness Centrality measures the inverse of the average distance from a node to other nodes. It can be expressed by the following formula:

[0162]

[0163] Among them, C C (i) can be used to represent the closed centrality of node i, and d(i,j) can be used to represent the shortest path length from node i to node j. Figure 6 is a schematic diagram of the closed connection centrality of nodes corresponding to a contaminated point according to one embodiment of the present disclosure, such as Figure 6 As shown, node 53 links node 71, node 50, node 49 and node 66. Through these four distances, the closed centrality can be calculated as 1 / (d_71+d_50+d_49+d_66).

[0164] Information centrality measures the importance of a node in all the shortest paths in the network, that is, the proportion of the shortest paths through the node. Information centrality can be expressed by the following formula:

[0165]

[0166] Among them, C B (i) can be used to represent the information centrality of node i, σ st It can be used to represent the number of shortest paths from node s to node t, σ st (i) can be used to represent the number of shortest paths passing through node i.

[0167] The clustering coefficient measures the degree to which the neighbors of a node are connected. The clustering coefficient can be expressed by the following formula:

[0168]

[0169] Among them, C(i) can be used to represent the clustering coefficient of node i, e i It can be used to represent the number of connected edges between neighbor nodes of node i, k i Can be used to represent the degree of node i. Figure 7 is a schematic diagram of a clustering coefficient of nodes corresponding to a pollution point according to one embodiment of the present disclosure, Figure 7 The clustering coefficient of pollutant nodes is depicted, from Figure 6 It can be seen that the neighbor nodes of node 53 are node 52 and node 66, among which the number of edges connected to node 52 is 4, the number of edges connected to node 66 is 7, and the degree of node 53 is 5, then the clustering coefficient of node 53 is 22 / 20=1.1.

[0170] Figure 8 is a flowchart of determining association indicators based on a knowledge graph according to one embodiment of the present disclosure, such as Figure 8As shown, the process of determining the association index based on the knowledge graph may include the following steps:

[0171] Step S801, obtaining a knowledge graph of pollutants.

[0172] In the above step S801, after constructing the knowledge graph of pollutant w, the knowledge graph of pollutant w can be read.

[0173] Step S802, counting the number of nodes in the knowledge graph.

[0174] In the above step S802, in obtaining the knowledge graph of pollutants, the number of nodes in the knowledge graph can be counted to obtain a total number of nodes of n.

[0175] Step S803, obtaining the node i with a sequence number equal to zero.

[0176] In the above step S803, after counting the number of nodes in the knowledge graph, the node number i=0 can be read from multiple nodes.

[0177] Step S804, determine whether i is equal to n-1.

[0178] In the above step S804, when i is equal to n-1, the whole process ends, otherwise, it goes to step S805.

[0179] Step S805, determining the correlation index.

[0180] In the above step S805, the association indicators may include node degree, degree centrality, closure centrality, information centrality and clustering coefficient. Based on the formulas of the above node degree, degree centrality, closure centrality, information centrality and clustering coefficient, each association indicator may be determined.

[0181] Step S806, accumulating i.

[0182] In the above step S806, when determining the association index corresponding to the node i, i=i+1 may be set, and the above determination steps may be further repeated.

[0183] The input data of the above process is the knowledge graph of pollutant w, and the output data is the association index of the pollutant. Through the above process, relevant association indexes can be extracted based on the constructed knowledge graph.

[0184] The knowledge graph can be used to correlate different indicators in environmental pollution data to form an intuitive view, thereby revealing the complex relationships hidden behind the data. By analyzing the connection patterns in the graph, potential connections between pollutants can be discovered and the common behavior of pollutants in the environment can be understood.

[0185] By combining statistical analysis and graphical methods, environmental pollution data is analyzed and interpreted from multiple perspectives, emphasizing the importance of structured knowledge representation and its role in interpreting environmental data. The knowledge graph connects various data points in environmental monitoring through the structure of entities, attributes and relationships to form a comprehensive and queryable knowledge network. This structure not only promotes the integration and sharing of information, but also provides the possibility for in-depth analysis of pollution sources and transmission pathways.

[0186] By studying the degree distribution, degree centrality and closed connection centrality of nodes in the network, we can find the key nodes in the network, which may play a decisive role in environmental pollution monitoring and response measures. For example, nodes with high degree distribution may be important monitoring sites, and nodes with high degree centrality and closed connection centrality may be key pollution transmission points, which can help identify pollution patterns, provide a scientific basis for environmental management, and provide support for the implementation of effective pollution control strategies.

[0187] Fig. 9 is a schematic diagram of a first-layer pollutant copper concentration distribution map according to one embodiment of the present disclosure, Fig. 9 The distribution of copper concentration at the first depth of the sampling point is shown. Each node represents a sampling point. The lines between nodes indicate that the copper concentration exceeds the set threshold, indicating that there may be pollution diffusion or similar pollution sources. Through this distribution map, we can intuitively observe the correlation between the copper concentration of pollutants in different locations. Fig.10 is a schematic diagram of a first-layer pollutant zinc concentration distribution map according to one embodiment of the present disclosure, Fig.10 The concentration distribution of the pollutant zinc at the first depth is shown. The sampling points represented by the nodes show the zinc concentrations at different locations. The connection between the nodes indicates that the zinc concentration exceeds the set threshold, indicating that there may be diffusion of pollutants or similar pollution sources. The distribution map can intuitively observe the correlation between the zinc concentrations of pollutants at different locations.

[0188] Fig.11 is a schematic diagram of the distribution of cadmium concentration points of the first layer of soil pollutants according to one embodiment of the present disclosure, such as Fig.11 As shown in the figure, the distribution of cadmium pollution is reflected, which not only helps to understand the concentration distribution of cadmium in different locations, but also provides an important reference for environmental protection decision-making. Fig.11 Potential pollution sources can be identified and appropriate measures can be taken to reduce or eliminate the impact of cadmium pollution on the environment and public health. Fig.12 is a schematic diagram of a first layer soil petroleum hydrocarbon concentration network according to one embodiment of the present disclosure, such as Fig.12As shown in the figure, a network diagram depicting the concentration of petroleum hydrocarbons in the first layer of soil samples is shown. Each circular node represents a soil sample point, and the number on it corresponds to the specific sampling location, while the lines between the nodes symbolize the relationship between the concentrations of petroleum hydrocarbons. The number and density of the lines reflect the intensity of the interaction between the concentrations of petroleum hydrocarbons at different points, which can help visualize the distribution of pollution and its potential connections. This network diagram provides an intuitive perspective for understanding the spatial distribution characteristics of pollutants, which in turn helps environmental scientists identify pollution clusters and evaluate and develop more effective soil pollution control strategies.

[0189] Knowledge graph is a structured knowledge organization and representation method. It presents knowledge in a graphical structure through three basic components: entities, attributes, and relationships. It is particularly critical for intelligent applications. In environmental pollution data analysis, knowledge graphs can associate different indicators, reveal the complex connections behind them, and provide scientific support for the formulation of environmental protection strategies.

[0190] Fig.13 is a schematic diagram of the distribution of copper contamination nodes according to one embodiment of the present disclosure, such as Fig.13 As shown, the horizontal axis represents the node and the vertical axis represents the node degree, that is, the figure shows the degree distribution. Fig.13 The degree distribution of each node in a network graph consisting of nodes and edges is revealed. In this network, each node represents a unique entity, such as a sampling point in environmental monitoring, and the degree of the node (i.e., the number of nodes connected) reflects the direct connection between the node and other nodes. The degree distribution describes the structural characteristics of the importance of nodes in the network, such as which nodes are highly connected central nodes, and the overall distribution pattern of network connections. In the context of environmental pollution monitoring, the height of a node may mean that a sampling point shares similar pollution levels with many other sampling points, indicating a potential source of pollution or a pollution transmission pathway. By analyzing the degree distribution, key monitoring points and pollution control points can be better located.

[0191] Fig.14 is a schematic diagram of the degree centrality distribution of copper network nodes according to one embodiment of the present disclosure, such as Fig.14 As shown, the horizontal axis represents the node and the vertical axis represents the degree centrality. Fig.14 The degree centrality distribution of each node in the network is presented in detail, reflecting the importance and influence of each node in the entire network. Degree centrality is a criterion for measuring the importance of a node. Degree centrality is based on an assumption that the importance of a node is proportional to the number of nodes it connects to. Fig.14As can be seen in the figure, some nodes have relatively high degree centrality, indicating that these nodes may be key connection points in the network. In fields such as environmental monitoring or social network analysis, nodes with high centrality may have a significant impact on the structure and information flow of the overall network. By analyzing these nodes, key structural features in the network can be identified, such as the most important pollution sources or the most active social interactors, thus providing a basis for further analysis and decision-making.

[0192] Fig.15 is a schematic diagram of the closed connection centrality distribution of copper network nodes according to one embodiment of the present disclosure, such as Fig.15 As shown, the horizontal axis represents the nodes and the vertical axis represents the closed centrality. Fig.15 The closed connection centrality of each node in the network is presented in detail. Closed connection centrality is an indicator to measure the central position of a node in the network. Closed connection centrality reflects the average distance between a node and all other nodes. The higher the closed connection centrality of a node, the shorter the average distance between the node and other nodes in the network, which means that the node is more likely to contact other nodes in the network and has a higher information flow efficiency. In the context of environmental monitoring, a sampling point with high closed connection centrality may indicate that the sampling point is associated with multiple pollution points and is a potential key monitoring and intervention location. By analyzing closed connection centrality, important nodes in the network can be identified, which is of great significance for understanding and optimizing the network structure, as well as formulating corresponding strategies.

[0193] Through on-site collection and other methods, various components contained in various pollution sources at different depths and points can be obtained. Table 1 is the original data set collected. As shown in Table 1, the sample point number (ID), point, location, sampling depth interval, and concentration values ​​of various pollutants are listed, including inorganic pollutants and organic pollutants. Among them, inorganic substances can at least include: copper (Cu), zinc (Zn), cadmium (Cd), mercury (Hg), hexavalent chromium (Cr6+), arsenic, nickel, lead, cyanide, etc. Organic substances can at least include: total hydrocarbons (CH), petroleum hydrocarbons (C10-C41), pentachlorophenol, polycyclic aromatic hydrocarbons, benzanthracene, benzopyrene, benzofluoranthene, dibenzanthracene, indenopyrene, naphthalene, etc. If some columns in the table show ND, it means that the pollutant was not detected or was below the detection limit at the corresponding depth.

[0194] Table 1 Original dataset

[0195]

[0196] Table 2 is the association index extracted based on the knowledge graph. As shown in Table 2, the point location, the pollutant is Cu, the number of layers of the knowledge graph, the content of Cu in this layer, degree, degree centrality, closure centrality, information centrality, clustering coefficient and the content of Cu in the next layer are listed, and the content of Cu in the next layer is used as a label.

[0197] Table 2 Correlation indicators

[0198]

[0199] After obtaining the original data set and the associated indicators, the original data set and the associated indicators can be merged to obtain a comprehensive indicator system, which can at least include point location, pollutant, number of layers, content of this layer, degree, degree centrality, closure centrality, information centrality, clustering coefficient, Cu, Zn, Cd, Hg, Cr6+, CH, (C10-C41) and the content (label) of Cu in the next layer.

[0200] Furthermore, the performance of the prediction model before and after adding the correlation index can be verified. For the prediction of copper concentration, Table 3 compares the mean square error (MSE), mean absolute error (MAE) and root mean square error (RMSE) of the SVM model with and without correlation indicators. As shown in Table 3, it shows the influence of considering the correlation index on the prediction performance when predicting copper concentration using the support vector machine SVM model. It can be clearly seen from Table 3 that after the introduction of the correlation index in the model, the prediction accuracy is significantly improved, which can be confirmed by the fact that the various error indicators with the correlation index are lower than the corresponding values ​​of the uncorrelated index. MSE, MAE and RMSE decreased from 0.457569, 0.560627 and 0.676438 before the introduction of correlation indicators to 0.164616, 0.325301 and 0.405729 after the introduction of correlation indicators, respectively, indicating that the addition of correlation indicators helps the SVM model to capture and predict the changes in copper concentration more accurately.

[0201] Table 3 Comparison of the performance of SVM models with and without associated indicators in predicting copper

[0202] Evaluation Metrics Related indicators No correlation indicator MSE 0.164616 0.457569 MAE 0.325301 0.560627 RMSE 0.405729 0.676438

[0203] For the prediction of copper concentration, Table 4 compares the mean square error MSE, mean absolute error MAE and root mean square error RMSE of the LSTM model with and without associated indicators. As shown in Table 4, the performance indicators of the LSTM model with and without associated indicators are compared. In the case of associated indicators, the mean square error MSE of the LSTM model is 0.193485, the mean absolute error MAE is 0.34891610951096697, and the root mean square error RMSE is 0.4398694002767101. In the case of unassociated indicators, the above indicators are 0.4966482661, 0.5843104745 and 0.7047327622 respectively. It can be seen that the LSTM model considering the associated indicators performs better under all indicators.

[0204] Table 4 Comparison of the performance of LSTM model with and without related indicators in predicting copper

[0205]

[0206]

[0207] Fig.16 is a schematic diagram of the comparison between the actual value and the predicted value of the copper concentration predicted by LSTM according to one embodiment of the present disclosure, such as Fig.16 As shown in the figure, LSTM is used to predict the actual value and predicted value of copper concentration, and the predicted value includes the cases where there are no related indicators in the model and the cases where there are related indicators. Fig.16 It can be observed that after adding the correlation index, the prediction curve is closer to the actual value curve, which shows that the correlation index is beneficial to improve the prediction accuracy. The prediction value without the correlation index is far from the true value in some periods, especially in the first few training cycles. In contrast, the prediction with the addition of the correlation index shows a more stable trend throughout the training process and more accurately captures the fluctuation of the true value in many cycles, indicating that considering the correlation index when building the neural network model is a key step to improve the accuracy of pollutant concentration prediction.

[0208] Compared with the SVM model, it can be seen that the LSTM model (especially the version with correlation indicators) performs more smoothly when processing sequence prediction tasks, reflecting the potential of LSTM in capturing long-term dependencies, revealing the performance differences of different models when processing similar prediction tasks, and the potential impact of input indicator selection on prediction accuracy.

[0209] For the prediction of petroleum hydrocarbon concentration, Table 5 compares the mean square error MSE, mean absolute error MAE and root mean square error RMSE of the SVM model with and without associated indicators. As shown in Table 5, the mean square error MSE, mean absolute error MAE and root mean square error RMSE of the SVM model with associated indicators are 0.1646, 0.3253 and 0.4057 respectively. In the case of no associated indicators, the above indicators are 0.4576, 0.5606 and 0.6764 respectively. Therefore, the SVM model performs better with associated indicators.

[0210] Table 5 Comparison of the performance of SVM models with and without associated indicators in predicting petroleum hydrocarbons

[0211] Evaluation Metrics Related indicators No correlation indicator MSE 0.1646 0.4576 MAE 0.3253 0.5606 RMSE 0.4057 0.6764

[0212] Fig.17 FIG. 4 is a schematic diagram showing a comparison between the actual value and the predicted value of the petroleum hydrocarbon concentration predicted by SVM according to one embodiment of the present disclosure. Fig.17 As shown in Figure 1, the actual value and predicted value of petroleum hydrocarbon concentration are predicted using SVM, and the predicted value includes the cases where there are no related indicators and the cases where there are related indicators in the model. Fig.17 It can be observed that the SVM model taking into account the correlation index is closer to the true value when predicting petroleum hydrocarbons, and the deviation of the prediction results is smaller.

[0213] For the prediction of petroleum hydrocarbon concentration, Table 6 compares the mean square error (MSE), mean absolute error (MAE), and root mean square error (RMSE) of the LSTM model with and without related indicators. As shown in Table 6, the mean square error (MSE), mean absolute error (MAE), and root mean square error (RMSE) of the LSTM model with related indicators are 0.3575, 0.2888, and 0.5979, respectively. In the case of no related indicators, the above indicators are 0.5783, 0.4268, and 0.7605, respectively. Therefore, the LSTM model performs better with related indicators.

[0214] Table 6 Comparison of the performance of LSTM model with and without related indicators in predicting petroleum hydrocarbons

[0215]

[0216]

[0217] In summary, both the SVM model and the LSTM model achieved better prediction performance when the correlation index was considered. This shows that the introduction of the correlation index can improve the performance and prediction accuracy of the model.

[0218] In this embodiment, a site that has been determined to be contaminated or may be contaminated is usually selected. Exceeding pollutants include organic and inorganic substances, among which organic substances may include at least pentachlorophenol, polycyclic aromatic hydrocarbons, petroleum hydrocarbons, benzanthracene, benzopyrene, benzofluoranthene, dibenzanthracene, indenopyrene and naphthalene. Inorganic substances may include at least copper, zinc, cadmium, mercury, hexavalent chromium, arsenic, nickel, lead and cyanide. Pollutant monitoring data from different sources can be collected. For example, various organizations such as scientific research institutions and data providers can be cooperated to obtain data resources. At the same time, tools such as sensor networks and satellite remote sensing can also be used to obtain pollution source data in real time. The exceeding depth of pollutants is 0 to 25m, slowly migrating to the downstream and groundwater in the southeast direction of the site, but there is no obvious distribution pattern. Each point in the soil pollution point represents a sampling point, and the same point is sampled once every 5 meters. The measurement units of different pollutants will vary, generally mg / kg, micrograms / cubic meter (μg / m 3 ), mg / m3 (mg / m 3 ), etc., that is, the mass of pollutants contained in the unit sampled soil.

[0219] Since the collected data usually have various problems, such as missing data, inconsistent formats and low accuracy. Therefore, after collecting the pollutant monitoring data, it is necessary to clean and preprocess the pollutant monitoring data and convert the data into a unified format and structure. Secondly, use data integration technology to integrate data from different sources into a unified database for subsequent analysis and query. Construct a pollutant knowledge graph based on the preprocessed data and the above-mentioned method of constructing a knowledge graph. Extract relationship indicators based on the pollutant knowledge graph. The data after data preprocessing forms an initial indicator set, which is merged with the relationship indicator set to form a comprehensive indicator system. Select supervised machine learning methods such as artificial neural networks, SVM, LSTM, etc., and use comprehensive indicators as input and the next layer of content as labels to build a model.

[0220] In the embodiment of the present disclosure, multiple depths of the pollutant to be predicted at multiple points are obtained respectively, and the first target depth is determined from the multiple depths; the correlation coefficient between any two points among the multiple points located at the first target depth is determined, wherein the correlation coefficient is used to represent the degree of correlation between any two points; based on the correlation coefficient and the multiple points, a knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed, wherein the knowledge graph is used to represent the degree of correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted, wherein the second target depth is the next layer of depth of the first target depth among the multiple depths. That is, the embodiment of the present disclosure constructs a knowledge graph corresponding to the pollutant to be predicted at the first target depth, and predicts the pollutant content of the pollutant to be predicted at the second target depth based on the constructed knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, thereby achieving the technical effect of improving the efficiency of predicting the underground pollutant content and solving the technical problem of low efficiency in predicting the underground pollutant content.

[0221] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present disclosure, or the part that contributes to the relevant technology, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of each embodiment of the present disclosure.

[0222] In this embodiment, a pollutant prediction device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "unit" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0223] Fig.18 is a structural block diagram of a pollutant prediction device according to one embodiment of the present disclosure, such as Fig.18 As shown, the pollutant prediction device 1800 may include: a first determination unit 1802 , a second determination unit 1804 , a construction unit 1806 and a prediction unit 1808 .

[0224] The first determination unit 1802 is used to respectively obtain multiple depths of the pollutants to be predicted at multiple points, and determine a first target depth from the multiple depths.

[0225] The second determining unit 1804 is configured to determine a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points.

[0226] Construction unit 1806 is used to construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth based on the correlation coefficient and multiple points, wherein the knowledge graph is used to represent the degree of correlation between multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth.

[0227] The prediction unit 1808 is used to predict the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is the next depth layer of the first target depth among multiple depths.

[0228] Optionally, the construction unit 1806 includes: a first determination module, used to respectively determine multiple points as nodes of the knowledge graph to be constructed, and obtain multiple nodes; a second determination module, used to determine two nodes corresponding to any two points from the multiple nodes, and based on the correlation coefficient between any two points, determine the correlation result between the two nodes, wherein the correlation result is used to indicate that the correlation between the two nodes is allowed, or that the correlation between the two nodes is prohibited; a construction module, used to construct a knowledge graph based on the correlation results between the multiple nodes and the two nodes.

[0229] Optionally, the second determination module is used to determine the association result between two nodes based on the association coefficient between any two points, including: a first determination submodule, used to determine that the association between the two nodes is allowed in response to the association coefficient being less than the association coefficient threshold; and a second determination submodule, used to determine that the association between the two nodes is prohibited in response to the association coefficient being greater than or equal to the association coefficient threshold.

[0230] Optionally, the construction module includes: a traversal sub-module, used to traverse the association results between any two nodes among multiple nodes, and add any two nodes whose association results allow association to the edge matrix; a connection sub-module, used to connect any two nodes according to the edge matrix to obtain a knowledge graph.

[0231] Optionally, the prediction unit 1808 includes: a third determination module, used to determine the association index of the pollutant to be predicted based on the knowledge graph, wherein the association index is used to represent the structural characteristics of the pollutant to be predicted; a prediction module, used to predict the pollutant content of the pollutant to be predicted at a second target depth based on the association index and the pollutant content of the pollutant to be predicted at the first target depth.

[0232] Optionally, the prediction includes: a construction submodule, used to construct a prediction model based on associated indicators and the pollutant content of the pollutant to be predicted at a first target depth, wherein the prediction model is used to predict the pollutant content; a prediction submodule, used to call the prediction model, analyze the associated indicators and the pollutant content of the pollutant to be predicted at the first target depth, and obtain the pollutant content of the pollutant to be predicted at a second target depth.

[0233] Optionally, the construction submodule is also used to: combine the associated index with the pollutant content of the pollutant to be predicted at the first target depth to obtain a comprehensive index of the pollutant to be predicted; and use the comprehensive index to train the initial prediction model to obtain a prediction model.

[0234] Optionally, the association indicator includes node degree, which is used to represent the number of connections between nodes in the knowledge graph. The third determination module includes: a third determination submodule, used to determine the number of edges connected to the node in the knowledge graph as the node degree corresponding to the node.

[0235] Optionally, the association index includes degree centrality, which is used to indicate the connection strength of a node. The third determination module includes: a fourth determination submodule, used to determine the quotient between the node degree and the target number of nodes as the degree centrality corresponding to the node, wherein the target number is determined by the total number of nodes.

[0236] Optionally, the association index includes a clustering coefficient, which is used to indicate the degree of clustering of nodes. The third determination module includes: an adjustment submodule, which is used to adjust the number of edges connected to the node's neighbor nodes using the node degree to obtain the clustering coefficient corresponding to the node.

[0237] Optionally, the association index includes closed connection centrality, which is used to represent the distance between any two nodes in the knowledge graph. The third determination module includes: a fifth determination submodule, used to determine the shortest path length between any two nodes to obtain multiple shortest path lengths; a sixth determination submodule, used to determine the average value of the sum of multiple shortest path lengths; and a seventh determination submodule, used to determine the inverse of the average value as the closed connection centrality corresponding to the node.

[0238] Optionally, the association index includes information centrality, which is used to indicate the importance of a node in the knowledge graph in the shortest path. The third determination module includes: an eighth determination submodule, used to determine the number of first shortest paths between any two nodes in the knowledge graph, and determine the number of second shortest paths passing through the nodes from the number of first shortest paths; and a ninth determination submodule, used to determine the quotient between the number of second shortest paths and the number of first shortest paths as the information centrality corresponding to the node.

[0239] Optionally, the device also includes: an acquisition unit, used to acquire monitoring data of the pollutant to be predicted, wherein the monitoring data is used to indicate the distribution state of the pollutant content of the pollutant to be predicted at multiple depths; and a processing unit, used to pre-process the monitoring data to obtain the pollutant content of the pollutant to be predicted at a first target depth.

[0240] In this embodiment, a plurality of depths of the pollutant to be predicted at a plurality of points are obtained respectively by the first determination unit 1802, and the first target depth is determined from the plurality of depths. The correlation coefficient between any two points among the plurality of points at the first target depth is determined by the second determination unit 1804, wherein the correlation coefficient is used to indicate the degree of correlation between any two points. The knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed by the construction unit 1806 based on the correlation coefficient and the plurality of points, wherein the knowledge graph is used to indicate the degree of correlation between the plurality of points, and the importance of the points to the distribution of the pollutant to be predicted at the first target depth. The pollutant content of the pollutant to be predicted at the second target depth is predicted by the prediction unit 1808 based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, wherein the second target depth is the next layer of depth of the first target depth among the plurality of depths. That is to say, the embodiment of the present disclosure constructs a knowledge graph corresponding to the pollutants to be predicted at the first target depth, and predicts the pollutant content of the pollutants to be predicted at the second target depth based on the constructed knowledge graph and the pollutant content of the pollutants to be predicted at the first target depth, thereby achieving the technical effect of improving the efficiency of predicting underground pollutant content and solving the technical problem of low efficiency in predicting underground pollutant content.

[0241] It should be noted that the above-mentioned units can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to this: the above-mentioned units are all located in the same processor; or the above-mentioned units are located in different processors in any combination.

[0242] An embodiment of the present disclosure further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any one of the above method embodiments when running.

[0243] Optionally, in this embodiment, the computer-readable storage medium may be configured to store a computer program for performing the following steps:

[0244] S1, respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths;

[0245] S2, determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points;

[0246] S3, based on the correlation coefficient and the multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth;

[0247] S4, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is a next layer of depth of the first target depth among multiple depths.

[0248] Optionally, in this embodiment, the above-mentioned computer-readable storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.

[0249] An embodiment of the present disclosure further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0250] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0251] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0252] S1, respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths;

[0253] S2, determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points;

[0254] S3, based on the correlation coefficient and the multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth;

[0255] S4, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is a next layer of depth of the first target depth among multiple depths.

[0256] An embodiment of the present disclosure further provides a computer program product, which may include a computer program, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0257] Optionally, in this embodiment, the computer program product may be configured to include a computer program for performing the following steps:

[0258] S1, respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths;

[0259] S2, determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points;

[0260] S3, based on the correlation coefficient and the multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth;

[0261] S4, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is a next layer of depth of the first target depth among multiple depths.

[0262] An embodiment of the present disclosure further provides a computer program product, which may include a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and the computer program is configured to execute the steps of any of the above method embodiments when running.

[0263] Optionally, in this embodiment, the computer program product may be configured to include a non-volatile computer-readable storage medium, and the non-volatile computer-readable storage medium may be configured to store a computer program for performing the following steps:

[0264] S1, respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths;

[0265] S2, determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points;

[0266] S3, based on the correlation coefficient and the multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth;

[0267] S4, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is a next layer of depth of the first target depth among multiple depths.

[0268] An embodiment of the present disclosure further provides a computer program, which is configured to execute the steps of any of the above method embodiments when running.

[0269] Optionally, in this embodiment, the computer program is configured to perform the following steps:

[0270] S1, respectively obtaining multiple depths of the pollutant to be predicted at multiple points, and determining a first target depth from the multiple depths;

[0271] S2, determining a correlation coefficient between any two points among a plurality of points located at the first target depth, wherein the correlation coefficient is used to indicate a degree of correlation between any two points;

[0272] S3, based on the correlation coefficient and the multiple points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the correlation between the multiple points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth;

[0273] S4, predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at a first target depth, wherein the second target depth is a next layer of depth of the first target depth among multiple depths.

[0274] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0275] Fig.19 is a schematic diagram of an electronic device according to an embodiment of the present disclosure. Fig.19 As shown, the electronic device 1900 is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0276] like Fig.19 As shown, the electronic device 1900 is in the form of a general computing device. The components of the electronic device 1900 may include but are not limited to: at least one processor 1910, at least one memory 1920, a bus 1930 connecting different system components (including the memory 1920 and the processor 1910), and a display 1940.

[0277] The memory 1920 stores program codes, which can be executed by the processor 1910, so that the processor 1910 executes the steps described in the method part of the embodiment of the present disclosure according to various exemplary embodiments of the present disclosure.

[0278] The memory 1920 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 19201 and / or a cache memory unit 19202, and may further include a read-only memory unit (ROM) 19203, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.

[0279] In some examples, the memory 1920 may also include a program / utility 19204 having a set (at least one) of program modules 19205, such program modules 19205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination thereof may include the implementation of a network environment. The memory 1920 may further include a memory remotely disposed relative to the processor 1910, and these remote memories may be connected to the electronic device 1900 via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0280] Bus 1930 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a local bus of processor 1910, or a bus using any of a variety of bus architectures.

[0281] The display 1940 may be, for example, a touch screen type liquid crystal display (LCD) that enables a user to interact with a user interface of the electronic device 1900 .

[0282] Optionally, the electronic device 1900 may also communicate with one or more external devices 2000 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device 1900, and / or any device that enables the electronic device 1900 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed through an input / output (I / O) interface 1950. Furthermore, the electronic device 1900 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 1960. Fig.19 As shown, the network adapter 1960 communicates with other modules of the electronic device 1900 via the bus 1930. It should be understood that although Fig.19 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 1900, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, redundant arrays of independent disks (RAID) systems, tape drives, and data backup storage systems.

[0283] The electronic device 1900 may further include: a keyboard, a cursor control device (such as a mouse), an input / output interface (I / O interface), a network interface, a power supply and / or a camera.

[0284] It can be understood by those skilled in the art that Fig.19 The structure shown is only for illustration and does not limit the structure of the above electronic device. Fig.19 More or fewer components as shown, or with Figure 1 The memory 1920 may be used to store computer programs and corresponding data, such as the computer programs and corresponding data corresponding to the pollutant prediction method in the embodiment of the present disclosure. The processor 1910 executes various functional applications and data processing by running the computer programs stored in the memory 1920, that is, implementing the above-mentioned pollutant prediction method.

[0285] The serial numbers of the above-mentioned embodiments of the present disclosure are only for description and do not represent the advantages or disadvantages of the embodiments.

[0286] In the above embodiments of the present disclosure, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0287] In the several embodiments provided in the present disclosure, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0288] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0289] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0290] If the integrated unit is implemented in the form of 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 the present disclosure, or the part that contributes to the relevant technology or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the various embodiments of the present disclosure. The aforementioned storage medium includes: U disk, read-only memory (ROM), random access memory (RAM), mobile hard disk, disk or optical disk, etc., which can store program code.

[0291] The above are only preferred embodiments of the present disclosure. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present disclosure. These improvements and modifications should also be regarded as the protection scope of the present disclosure.

Claims

1. A method for predicting pollutants, characterized in that: include: Respectively obtaining a plurality of depths of the pollutant to be predicted at a plurality of points, and determining a first target depth from the plurality of depths; Determine a correlation coefficient between any two points among the plurality of points located at the first target depth, wherein the correlation coefficient is used to represent a degree of correlation between the any two points; Based on the correlation coefficient and the plurality of points, construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth, wherein the knowledge graph is used to represent the degree of correlation between the plurality of points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; Based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at a second target depth is predicted, wherein the second target depth is a next layer of depth of the first target depth among the multiple depths.

2. The method according to claim 1, characterized in that Based on the correlation coefficient and the plurality of points, a knowledge graph corresponding to the pollutant to be predicted at the first target depth is constructed, including: Respectively determining the plurality of points as nodes of the knowledge graph to be constructed, to obtain a plurality of nodes; Determine two nodes corresponding to the two arbitrary points from the plurality of nodes, and determine an association result between the two nodes based on the association coefficient between the two arbitrary points, wherein the association result is used to indicate that the association between the two nodes is allowed, or that the association between the two nodes is prohibited; The knowledge graph is constructed based on the association results between the multiple nodes and the two nodes.

3. The method according to claim 2, characterized in that Determining the association result between the two nodes based on the association coefficient between the two arbitrary points includes: In response to the association coefficient being less than an association coefficient threshold, determining that association is allowed between the two nodes; In response to the association coefficient being greater than or equal to the association coefficient threshold, it is determined that association between the two nodes is prohibited.

4. The method according to claim 3, characterized in that Based on the plurality of nodes and the association results between the two nodes, constructing the knowledge graph includes: Traversing the association results between any two of the multiple nodes, and adding any two nodes whose association results indicate that association is allowed to be performed to the edge matrix; According to the edge matrix, any two nodes are connected to obtain the knowledge graph.

5. The method according to claim 1, characterized in that Predicting the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth includes: Based on the knowledge graph, determining the association index of the pollutant to be predicted, wherein the association index is used to represent the structural characteristics of the pollutant to be predicted; Based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth, the pollutant content of the pollutant to be predicted at the second target depth is predicted.

6. The method according to claim 5, characterized in that Predicting the pollutant content of the to-be-predicted pollutant at the second target depth based on the correlation index and the pollutant content of the to-be-predicted pollutant at the first target depth includes: Based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth, constructing a prediction model, wherein the prediction model is used to predict the pollutant content; The prediction model is called to analyze the associated index and the pollutant content of the pollutant to be predicted at the first target depth to obtain the pollutant content of the pollutant to be predicted at the second target depth.

7. The method according to claim 6, characterized in that Based on the correlation index and the pollutant content of the pollutant to be predicted at the first target depth, a prediction model is constructed, including: Combining the correlation index with the pollutant content of the pollutant to be predicted at the first target depth to obtain a comprehensive index of the pollutant to be predicted; The initial prediction model is trained using the comprehensive index to obtain the prediction model.

8. A pollutant prediction device, characterized in that: include: A first determination unit is used to respectively obtain a plurality of depths of the pollutant to be predicted at a plurality of points, and determine a first target depth from the plurality of depths; A second determining unit is used to determine a correlation coefficient between any two points among the plurality of points located at the first target depth, wherein the correlation coefficient is used to represent a degree of correlation between the any two points; A construction unit, configured to construct a knowledge graph corresponding to the pollutant to be predicted at the first target depth based on the correlation coefficient and the plurality of points, wherein the knowledge graph is used to represent the correlation between the plurality of points and the importance of the points to the distribution of the pollutant to be predicted at the first target depth; A prediction unit is used to predict the pollutant content of the pollutant to be predicted at a second target depth based on the knowledge graph and the pollutant content of the pollutant to be predicted at the first target depth, wherein the second target depth is a next layer of depth of the first target depth among the multiple depths.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 7 when executed by a processor.

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 7.

Images