Intelligent underground water quality supervision method and system

Abstract

The invention discloses an intelligent underground water quality supervision method and system. The method comprises heterogeneous sensing acquisition, water quality benchmark modeling, sewage traceability enhancement, water quality prediction and early warning and underground water quality supervision. The invention relates to the technical field of water quality supervision, in particular to an intelligent underground water quality supervision method and system, and adopts an underground water quality comprehensive supervision method combining water quality benchmark modeling, pollution traceability enhancement and dynamic threshold water quality prediction and early warning. The perception capability, the identification capability and the response capability of the underground water pollution event are effectively improved, and the supervision conversion from passive response to active pre-judgment is realized; carrying out water quality benchmark modeling by adopting an improved space-time diagram convolutional network method combined with a space-time dynamic baseline algorithm; pollution traceability enhancement is carried out by adopting an improved adversarial generative network method of virtual and real traceability modeling.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality supervision, and in particular to an intelligent groundwater quality supervision method and system. Background Art

[0002] The intelligent groundwater quality supervision method is a modern environmental monitoring technology system with intelligent analysis. It collects multi-dimensional water quality data in real time by deploying intelligent sensor networks, builds an adaptive assessment benchmark using spatiotemporal dynamic modeling and machine learning algorithms, and combines digital twins and pollution source tracing technology to achieve accurate positioning and diffusion prediction of pollution incidents, ultimately forming a full-process closed-loop management of "monitoring-early warning-decision-making-disposal", thereby providing a scientific basis for water resource protection, pollution control and emergency decision-making, and is the core technical means to achieve sustainable management of groundwater resources and smart environmental protection.

[0003] However, in the existing groundwater quality supervision methods, there are basic problems such as scattered monitoring data, difficulty in timely tracking of pollution changes, and lagging early warning mechanisms, which lead to a lack of overall grasp of pollution evolution in the supervision process and difficulty in taking effective response measures in a timely manner. When faced with the actual situation of unclear pollution sources, complex groundwater flow paths and diverse pollutants, traditional methods often rely only on fixed-point monitoring, which cannot achieve comprehensive tracking of groundwater pollution and trend prediction. In the existing water quality benchmark modeling methods, there is a technical problem that existing methods often use static standards or historical averages as the basis for judgment, ignoring the complex influence of natural volatility and human interference in groundwater quality in different regions and at different time scales. This "average" modeling method is prone to misjudgment, either false reporting of pollution or missing reporting of abnormalities. In the existing sewage source tracing enhancement methods, there is a technical problem that traditional source tracing methods mostly rely on the concentration inversion of pollutants and the inference of hydrogeological maps, which are easily limited by the problems of sparse sample points and incomplete data, and are difficult to deal with complex scenarios of multi-source superposition or hidden pollution, resulting in unclear tracing results and inaccurate path judgment. Summary of the invention

[0004] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent groundwater quality supervision method and system. In view of the basic problems of scattered monitoring data, difficulty in timely tracking of pollution changes, and lagging early warning mechanisms in the existing groundwater quality supervision methods, the supervision process lacks an overall grasp of the evolution of pollution and it is difficult to take effective countermeasures in a timely manner. In the face of the actual situation of unclear pollution sources, complex groundwater flow paths and diverse pollutants, traditional methods often rely only on fixed-point monitoring and cannot achieve comprehensive tracking of groundwater pollution and trend prediction. Technical problems, this solution creatively adopts a comprehensive groundwater quality supervision method that combines water quality benchmark modeling, pollution source tracing enhancement and dynamic threshold water quality prediction and early warning, effectively improving the perception, recognition and response capabilities of groundwater pollution events, and realizing the transformation of supervision from "passive response" to "active prediction"; in view of the existing water quality benchmark modeling methods, the existing methods often use static standards or historical averages as the basis for judgment, ignoring the natural characteristics of groundwater quality in different regions and time scales. The complex influence of volatility and human interference, this "average" modeling method is prone to misjudgment, either false reporting of pollution or missing reporting of abnormal technical problems. This solution creatively adopts an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm to conduct water quality benchmark modeling, which more scientifically reflects which water quality changes belong to normal fluctuations and which belong to abnormal pollution, thereby providing a more accurate basis for judgment for subsequent source tracing and early warning; in the existing sewage source tracing enhancement methods, there are traditional source tracing methods that rely more on the concentration inversion of pollutants and the inference of hydrogeological maps, which are easily limited by the problems of sparse sample points and incomplete data, and are difficult to cope with complex scenarios of multi-source superposition or hidden pollution, resulting in unclear tracing results and inaccurate path judgment. Technical problems, this solution creatively adopts an improved adversarial generative network method of virtual-real source tracing modeling to enhance pollution source tracing, which not only improves the accuracy of pollution source positioning, but also reveals the diffusion trend of pollutants in groundwater, enabling regulatory units to discover hidden pollution earlier and accurately lock in the source of pollution, significantly enhancing the foresight and pertinence of pollution prevention and control.

[0005] The technical solution adopted by the present invention is as follows: The present invention provides an intelligent groundwater quality monitoring method, which includes the following steps:

[0006] Step S1: heterogeneous sensing acquisition;

[0007] Step S2: water quality benchmark modeling;

[0008] Step S3: sewage source tracing enhancement;

[0009] Step S4: water quality prediction and early warning;

[0010] Step S5: Groundwater quality supervision.

[0011] Further, in step S1, the heterogeneous sensing acquisition is used to collect the original data set required for groundwater quality supervision, specifically to obtain the original data set of groundwater quality through sensor array data acquisition, and to obtain groundwater sensing optimization data by optimizing and preprocessing the original data set of groundwater quality;

[0012] The groundwater quality original data set includes physical and chemical parameter data, ion concentration data and sensor metadata;

[0013] The optimization preprocessing steps include data cleaning, data alignment, noise filtering, data standardization, data normalization, feature optimization and data labeling processing;

[0014] The groundwater sensing optimization data includes optimized physical and chemical parameter data, optimized ion concentration data and optimized sensing metadata.

[0015] Further, in step S2, the water quality benchmark modeling is used to construct a dynamic water quality assessment benchmark model, specifically based on the groundwater sensor optimization data, using an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm to perform water quality benchmark modeling to obtain multidimensional reference data of water quality health, including the following steps:

[0016] Step S21: multi-scale feature extraction, specifically, extracting multi-scale feature data from the groundwater sensor optimization data by constructing a one-dimensional convolutional long-short term neural network as a hybrid encoder; the multi-scale feature data includes local features and global features; the local features are used to represent hour-level sensor time data, and the global features are used to represent seasonal-level sensor time data;

[0017] The one-dimensional convolutional long-term short-term neural network is improved by introducing a dilated convolution to replace the standard one-dimensional convolution layer to enhance the temporal perception;

[0018] Step S22: spatiotemporal dynamic baseline modeling, specifically, performing spatiotemporal dynamic baseline modeling through spatial correlation modeling and temporal dynamic modeling to obtain temporal geological characteristic data;

[0019] The spatial correlation modeling specifically constructs a groundwater spatial relationship graph structure by modeling the hydrogeological conductivity coefficient and the pollutant diffusion, and calculates the spatial correlation feature data through a standard graph convolution network;

[0020] The temporal dynamics modeling is specifically performed by using an improved spatiotemporal graph convolutional network to extract spatiotemporal attention. The improved spatiotemporal graph convolutional network specifically introduces a gated spatiotemporal attention mechanism on the basis of a standard spatiotemporal graph convolutional network to perform spatiotemporal feature processing to obtain temporal dynamics modeling feature data;

[0021] The temporal geological characteristic data is specifically obtained by combining the spatial correlation characteristic data and the temporal dynamic modeling characteristic data to output as a spatiotemporal dynamic baseline, and used as a reference benchmark for water quality health;

[0022] Step S23: water quality health calculation, specifically, obtaining deviation data by calculating the Mahalanobis distance between the groundwater sensor optimization data and the temporal geological characteristic data, and calculating the water quality health by constructing a multi-dimensional health fusion mechanism to obtain a water quality health output;

[0023] Step S24: water quality benchmark modeling training, specifically, performing water quality benchmark modeling training through the multi-scale feature extraction, the spatiotemporal dynamic benchmark modeling and the water quality health calculation to obtain a water quality benchmark model Model WHI ;

[0024] Step S25: water quality benchmark modeling, specifically, using the water quality benchmark model Model based on the groundwater sensor optimization data WHI , conduct water quality benchmark modeling and obtain multi-dimensional reference data on water quality health;

[0025] The multi-dimensional reference data of water quality health, including multi-dimensional quantitative output of water quality health and comprehensive health rating of water quality health;

[0026] The multi-dimensional quantitative output of water quality health includes chemical dimension, ecological dimension and toxicity dimension;

[0027] The water quality health level is a comprehensive health level rating, with specific ratings including excellent, good, fair and poor.

[0028] Further, in step S3, the sewage source tracing enhancement is used to locate the pollution source and restore the pollution diffusion path. Specifically, based on the multi-dimensional reference data of water quality health and the groundwater sensor optimization data, an improved adversarial generative network method of virtual-real source tracing modeling is used to enhance pollution source tracing, and obtain probabilistic groundwater pollution source tracing analysis data, including the following steps:

[0029] Step S31: virtual and real data reconstruction, specifically, based on the water quality health multidimensional reference data, using the groundwater sensor optimization data corresponding to the water quality health multidimensional reference data including medium rating and poor rating in the water quality health comprehensive health rating as the actual data input for the virtual and real data reconstruction, and constructing a groundwater hydrodynamic model, setting boundary conditions and initial pollution sources, and constructing a virtual pollution diffusion data map based on the actual data input to obtain virtual pollution diffusion data;

[0030] The steps of constructing the virtual pollution diffusion data map include calculating the pollution intensity index, generating the pollution heat map interpolation, and generating the pollution mask binarization; the virtual pollution diffusion data specifically refers to the binary pollution high-risk mask map data;

[0031] Step S32: constructing a pollution path generation network, specifically by collecting the real observed pollution diffusion data corresponding to the virtual pollution diffusion data, constructing a generator, obtaining a pollution path generation network, and generating pollution path generation data, the pollution path generation data includes generating predicted pollution path data and generating pollution source diffusion heat map data;

[0032] The generator is constructed by using a dual-branch ResNet to process the data generation of the predicted pollution path data and the pollution source diffusion heat map data respectively, and constructing a feature fusion embedding layer to perform conditional feature embedding to obtain the pollution path generation data;

[0033] Step S33: constructing a dual-path discriminator network, specifically, respectively constructing a coarse-grained discriminator and a fine-grained discriminator, performing dual-path discriminator network construction, and obtaining a discriminator network;

[0034] The coarse-grained discriminator is used to evaluate the overall trend of the pollution path and the regional pollution probability; the fine-grained discriminator is used to analyze the path change characteristics of local pollution diffusion;

[0035] Step S34: loss improvement design, specifically, performing loss improvement design for generative adversarial training by constructing adversarial loss, path consistency loss and pollution graph guided loss, and obtaining a comprehensive loss function by weighted combination of the adversarial loss, path consistency loss and pollution graph guided loss;

[0036] Step S35: Sewage source tracing model training, specifically, through the virtual and real data reconstruction, the construction of the pollution path generation network, the construction of the dual path discriminator network and the loss improvement design, the generation adversarial method is used to train the sewage source tracing model to obtain the sewage source tracing model Model VRG ;

[0037] Step S36: Sewage source tracing enhancement, specifically, using the sewage source tracing model Model based on the water quality health multi-dimensional reference data and the groundwater sensor optimization data VRG , conduct sewage source tracing enhancement and obtain probabilistic groundwater pollution source tracing analysis data;

[0038] The probabilistic groundwater pollution source tracing analysis data includes pollution source heat map data, pollution path map data and groundwater pollution source tracing credibility score reference data.

[0039] Furthermore, in step S4, the water quality prediction and warning is used to realize predictive groundwater quality warning, specifically, based on the probabilistic groundwater pollution source tracing analysis data and by setting intelligent periodic warning thresholds, water quality prediction and warning are performed to obtain groundwater quality prediction and warning reference data.

[0040] Furthermore, in step S5, the groundwater quality supervision is used to conduct comprehensive groundwater quality supervision and management in combination with the water quality health, pollution source tracing analysis results and water quality prediction and early warning results. Specifically, based on the multi-dimensional reference data of water quality health, multi-dimensional groundwater health modeling is performed, and basic management of groundwater quality is assisted through visual analysis. Comprehensive groundwater quality supervision is performed based on the probabilistic groundwater pollution source tracing analysis data and the groundwater quality prediction and early warning reference data to obtain a groundwater quality supervision reference plan.

[0041] The present invention provides an intelligent groundwater quality monitoring system, comprising a heterogeneous sensor acquisition module, a water quality benchmark modeling module, a sewage source tracing enhancement module, a water quality prediction and early warning module and a groundwater quality monitoring module;

[0042] The heterogeneous sensing acquisition module is used for heterogeneous sensing acquisition, obtains groundwater sensing optimization data through heterogeneous sensing acquisition, and sends the groundwater sensing optimization data to the water quality benchmark modeling module and the sewage tracing enhancement module;

[0043] The water quality benchmark modeling module is used for water quality benchmark modeling, and obtains multi-dimensional reference data of water quality health through water quality benchmark modeling, and sends the multi-dimensional reference data of water quality health to the sewage tracing enhancement module and the groundwater quality supervision module;

[0044] The sewage source tracing enhancement module is used for sewage source tracing enhancement, and obtains probabilistic groundwater pollution source tracing analysis data through sewage source tracing enhancement, and sends the probabilistic groundwater pollution source tracing analysis data to the water quality prediction and early warning module;

[0045] The water quality prediction and warning module is used for water quality prediction and warning, obtains groundwater quality prediction and warning reference data through water quality prediction and warning, and sends the groundwater quality prediction and warning reference data to the groundwater quality supervision module;

[0046] The groundwater quality supervision module is used for groundwater quality supervision, and a groundwater quality supervision reference plan is obtained through groundwater quality supervision.

[0047] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0048] (1) In view of the basic problems of scattered monitoring data, difficulty in timely tracking of pollution changes, and lagging early warning mechanisms in existing groundwater quality supervision methods, the supervision process lacks an overall grasp of pollution evolution and is difficult to take effective response measures in a timely manner. In the face of the actual situation where the pollution source is unclear, the groundwater flow path is complex, and the pollutants are diverse, the traditional methods often rely only on fixed-point monitoring and cannot achieve comprehensive tracking of groundwater pollution and trend prediction. This plan creatively adopts a comprehensive groundwater quality supervision method that combines water quality benchmark modeling, pollution source tracing enhancement, and dynamic threshold water quality prediction and early warning, which effectively improves the perception, identification and response capabilities of groundwater pollution events, and realizes the transformation of supervision from "passive response" to "active prediction";

[0049] (2) In view of the fact that existing water quality benchmark modeling methods often use static standards or historical averages as the basis for judgment, ignoring the natural volatility of groundwater quality in different regions and time scales and the complex influence of human interference, this "average" modeling method is prone to misjudgment, either falsely reporting pollution or missing abnormal technical problems. This solution creatively uses an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm to perform water quality benchmark modeling, which more scientifically reflects which water quality changes belong to normal fluctuations and which belong to abnormal pollution, thereby providing a more accurate basis for subsequent source tracing and early warning;

[0050] (3) In the existing sewage source tracing enhancement methods, traditional source tracing methods mostly rely on the concentration inversion of pollutants and the inference of hydrogeological maps. They are easily limited by the problems of sparse sample points and incomplete data. They are also difficult to deal with complex scenarios of multi-source superposition or hidden pollution, resulting in unclear tracing results and inaccurate path judgment. This solution creatively adopts the improved adversarial generative network method of virtual-real source tracing modeling to enhance pollution source tracing. It not only improves the accuracy of pollution source positioning, but also reveals the diffusion trend of pollutants in groundwater, enabling regulatory units to discover hidden pollution earlier and accurately identify the source of pollution, significantly enhancing the foresight and pertinence of pollution prevention and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a process flow of an intelligent groundwater quality monitoring method provided by the present invention;

[0052] Figure 2 A schematic diagram of an intelligent groundwater quality monitoring system provided by the present invention;

[0053] Figure 3 This is a schematic diagram of the process of optimizing the preprocessing in step S1;

[0054] Figure 4 A schematic diagram of the process of water quality benchmark modeling in step S2;

[0055] Figure 5 This is a schematic diagram of the enhanced process of sewage source tracing in step S3.

[0056] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] In the description of the present invention, it should be understood that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0059] Example 1, see Figure 1 The present invention provides an intelligent groundwater quality monitoring method, which comprises the following steps:

[0060] Step S1: heterogeneous sensing acquisition;

[0061] Step S2: water quality benchmark modeling;

[0062] Step S3: sewage source tracing enhancement;

[0063] Step S4: water quality prediction and early warning;

[0064] Step S5: Groundwater quality supervision.

[0065] By executing the above operations, the basic problems of existing groundwater quality supervision methods, such as scattered monitoring data, difficulty in timely tracking of pollution changes, and lagging early warning mechanisms, have been solved, resulting in a lack of overall grasp of pollution evolution in the supervision process and difficulty in taking effective response measures in a timely manner. When faced with the actual situation of unclear pollution sources, complex groundwater flow paths and diverse pollutants, traditional methods often rely only on fixed-point monitoring and are unable to achieve comprehensive tracking of groundwater pollution and technical problems such as trend prediction. This solution creatively adopts a comprehensive groundwater quality supervision method that combines water quality benchmark modeling, pollution source tracing enhancement and dynamic threshold water quality prediction and early warning, which effectively improves the perception, identification and response capabilities of groundwater pollution incidents, and realizes the transformation of supervision from "passive response" to "active prediction".

[0066] Example 2, see Figure 1 , Figure 2 and Figure 3 In step S1, the heterogeneous sensor acquisition is used to collect the original data set required for groundwater quality supervision, specifically, to obtain the original groundwater quality data set through sensor array data acquisition, and to obtain groundwater sensor optimization data by optimizing and preprocessing the original groundwater quality data set;

[0067] The original groundwater quality data set includes physical and chemical parameter data, ion concentration data and sensor metadata; the physical and chemical parameter data include pH value, conductivity, dissolved oxygen, redox potential, temperature, turbidity, total dissolved solids and chemical oxygen demand data; the ion concentration data include sodium ion, potassium ion, calcium ion, magnesium ion, chloride ion, sulfate, lead (Pb), arsenic (As), mercury (Hg), cadmium (Cd), chromium (Cr) and nutrient ion concentration data; the sensor metadata includes sampling timestamp, sampling location information, sensor number information, sensor type information, sampling depth and external air pressure information;

[0068] The optimization preprocessing steps include data cleaning, data alignment, noise filtering, data standardization, data normalization, feature optimization and data labeling processing;

[0069] The feature optimization includes water quality index derived feature construction, ion ratio feature construction and abnormal fluctuation index construction; the data labeling is specifically labeling processing through artificial pollution concentration judgment, and through data labeling, background water samples and polluted water samples are distinguished to obtain groundwater sensor optimization data;

[0070] The groundwater sensing optimization data includes optimized physical and chemical parameter data, optimized ion concentration data and optimized sensing metadata.

[0071] Example 3, see Figure 1 , Figure 2 and Figure 4 This embodiment is based on the above embodiment. In step S2, the water quality benchmark modeling is used to construct a dynamic water quality assessment benchmark model. Specifically, based on the groundwater sensor optimization data, an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm is used to perform water quality benchmark modeling to obtain multidimensional reference data of water quality health, including the following steps:

[0072] Step S21: multi-scale feature extraction, specifically, extracting multi-scale feature data from the groundwater sensor optimization data by constructing a one-dimensional convolutional long-short term neural network as a hybrid encoder; the multi-scale feature data includes local features and global features; the local features are used to represent hour-level sensor time data, and the global features are used to represent seasonal-level sensor time data;

[0073] The one-dimensional convolutional long-term short-term neural network introduces a dilated convolution to replace the standard one-dimensional convolution layer, performs time-series perception enhancement and improvement, and obtains multi-scale feature data. The calculation formula is:

[0074] ;

[0075] Where H is the multi-scale feature data, LSTM(·) is the long-short term neural network operation function, CNN(·) is the dilated convolution operation function, and X is the groundwater sensor optimization data;

[0076] Step S22: spatiotemporal dynamic baseline modeling, specifically, performing spatiotemporal dynamic baseline modeling through spatial correlation modeling and temporal dynamic modeling to obtain temporal geological characteristic data;

[0077] The spatial correlation modeling is specifically to construct the groundwater spatial relationship graph structure by modeling the hydrogeological conductivity coefficient and the pollutant diffusion, and to calculate the spatial correlation characteristic data through the standard graph convolution network. The calculation formula is:

[0078] ;

[0079] In the formula, w ij is the edge weight of the spatial correlation feature data, exp(·) is the natural base function, K ij is the geological hydrological conductivity coefficient, i is the index of the first monitoring point, j is the index of the second monitoring point, is the average value of the geohydrological conductivity coefficient, is the variance of the geohydrological conductivity coefficient, sim(·) is the pollutant diffusion similarity calculation function, h i is the multi-scale feature data corresponding to the first monitoring point, h jis the multi-scale feature data corresponding to the second monitoring point;

[0080] The temporal dynamics modeling is specifically performed by using an improved spatiotemporal graph convolutional network to extract spatiotemporal attention. The improved spatiotemporal graph convolutional network specifically introduces a gated spatiotemporal attention mechanism on the basis of a standard spatiotemporal graph convolutional network to perform spatiotemporal feature processing to obtain temporal dynamics modeling feature data. The calculation formula is:

[0081] ;

[0082] In the formula, is the temporal dynamics modeling feature data of the l+1th improved spatiotemporal graph convolutional network layer, SIG(·) is the S-type activation function, A is the spatial adjacency matrix, The lth layer improves the temporal dynamics modeling feature data of the spatiotemporal graph convolutional network layer, where l is the level index of the improved spatiotemporal graph convolutional network layer. are the weights of a standard spatiotemporal graph convolutional network, is the gated attention weight, which is calculated by the temporal attention generated by the one-dimensional convolutional long short-term neural network;

[0083] The temporal geological characteristic data is specifically obtained by combining the spatial correlation characteristic data and the temporal dynamic modeling characteristic data to output as a spatiotemporal dynamic baseline, and used as a reference benchmark for water quality health;

[0084] Step S23: water quality health calculation, specifically, obtaining deviation data by calculating the Mahalanobis distance between the groundwater sensor optimization data and the temporal geological characteristic data, and calculating the water quality health by constructing a multi-dimensional health fusion mechanism to obtain a water quality health output;

[0085] The calculation formula of the multi-dimensional health fusion mechanism is:

[0086] ;

[0087] Where WHI i,t is the water quality health output of the first monitoring point i at time t, k is the slope parameter of the S-type activation function, is the deviation data of the first monitoring point i at time t, It is the deviation threshold parameter, and its specific value is 90% of the historical deviation data;

[0088] By constructing a water quality health grading standard, the water quality health is quantified and rated according to the water quality health output;

[0089] Preferably, Table 1 is a reference table of the water quality health grading standard, such as Table 1, and the water quality health is quantified and rated based on the water quality health output;

[0090] Table 1 Water quality health classification standard reference table

[0091]

[0092] Step S24: water quality benchmark modeling training, specifically, performing water quality benchmark modeling training through the multi-scale feature extraction, the spatiotemporal dynamic benchmark modeling and the water quality health calculation to obtain a water quality benchmark model Model WHI ;

[0093] Step S25: water quality benchmark modeling, specifically, using the water quality benchmark model Model based on the groundwater sensor optimization data WHI , conduct water quality benchmark modeling and obtain multi-dimensional reference data on water quality health;

[0094] The multi-dimensional reference data of water quality health, including multi-dimensional quantitative output of water quality health and comprehensive health rating of water quality health;

[0095] The multi-dimensional quantitative output of water quality health includes chemical dimension, ecological dimension and toxicity dimension;

[0096] The water quality health level is a comprehensive health level rating, with specific ratings including excellent, good, fair and poor.

[0097] By performing the above operations, in view of the fact that in the existing water quality benchmark modeling methods, the existing methods often use static standards or historical averages as the basis for judgment, ignoring the natural volatility of groundwater quality in different regions and at different time scales and the complex influence of human interference, this "average" modeling method is prone to misjudgment, either false reporting of pollution or omission of abnormal technical problems, this solution creatively uses an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm to perform water quality benchmark modeling, which more scientifically reflects which water quality changes belong to normal fluctuations and which belong to abnormal pollution, thereby providing a more accurate judgment basis for subsequent tracing and early warning.

[0098] Example 4, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S3, the sewage source tracing is enhanced to locate the pollution source and restore the pollution diffusion path. Specifically, based on the multi-dimensional reference data of water quality health and the groundwater sensor optimization data, an improved adversarial generative network method of virtual-real source tracing modeling is used to enhance pollution source tracing and obtain probabilistic groundwater pollution source tracing analysis data, including the following steps:

[0099] Step S31: virtual and real data reconstruction, specifically, based on the water quality health multidimensional reference data, using the groundwater sensor optimization data corresponding to the water quality health multidimensional reference data including medium rating and poor rating in the water quality health comprehensive health rating as the actual data input for the virtual and real data reconstruction, and constructing a groundwater hydrodynamic model, setting boundary conditions and initial pollution sources, and constructing a virtual pollution diffusion data map based on the actual data input to obtain virtual pollution diffusion data;

[0100] The steps of constructing the virtual pollution diffusion data map include calculating the pollution intensity index, generating the pollution heat map interpolation, and generating the pollution mask binarization; the virtual pollution diffusion data specifically refers to the binary pollution high-risk mask map data;

[0101] The calculation formula for generating the pollution mask binarization is:

[0102] ;

[0103] Where M(x,y) is the binary pollution high-risk mask map data, where x is the horizontal pixel index, y is the vertical pixel index, and PI(·) is the pollution intensity index function, which is specifically calculated by the pollution intensity index. is the pollution intensity threshold;

[0104] Preferably, the calculation formula for calculating the pollution intensity index is:

[0105] ;

[0106] Where PI(·) is the pollution intensity index function, is the pollution index sample index, C is the total number of pollutants, c is the pollutant index, w c is the pollution factor weight, is the pollution index value of the cth pollutant corresponding to the i-th pollution index sample, S c It is the water quality standard limit;

[0107] The pollution intensity threshold Set to 95%;

[0108] Step S32: constructing a pollution path generation network, specifically by collecting the real observed pollution diffusion data corresponding to the virtual pollution diffusion data, constructing a generator, obtaining a pollution path generation network, and generating pollution path generation data, the pollution path generation data includes generating predicted pollution path data and generating pollution source diffusion heat map data;

[0109] The generator is constructed by using a dual-branch ResNet to process the data generation of the predicted pollution path data and the pollution source diffusion heat map data respectively, and constructing a feature fusion embedding layer to perform conditional feature embedding to obtain the pollution path generation data;

[0110] Step S33: constructing a dual-path discriminator network, specifically, respectively constructing a coarse-grained discriminator and a fine-grained discriminator, performing dual-path discriminator network construction, and obtaining a discriminator network;

[0111] The coarse-grained discriminator is used to evaluate the overall trend of the pollution path and the regional pollution probability; the fine-grained discriminator is used to analyze the path change characteristics of local pollution diffusion;

[0112] Step S34: loss improvement design, specifically, performing loss improvement design for generative adversarial training by constructing adversarial loss, path consistency loss and pollution graph guided loss, and obtaining a comprehensive loss function by weighted combination of the adversarial loss, path consistency loss and pollution graph guided loss;

[0113] The adversarial loss adopts the standard generative adversarial loss function as the adversarial loss;

[0114] The path consistency loss is specifically calculated by performing L2 norm difference calculation between the generated predicted contaminated path data and the actual contaminated path as the path consistency loss;

[0115] The pollution map guided loss is specifically calculated by combining the generated pollution source diffusion heat map data with the binary pollution high risk mask map data to make the generated model focus on the core area of ​​pollution as the pollution map guided loss;

[0116] The calculation formula of the comprehensive loss function is:

[0117] ;

[0118] In the formula, is the comprehensive loss function, is the adversarial loss weight, is the standard generative adversarial loss function, is the path consistency loss weight, is the path consistency loss function, is the pollution graph guided loss weight, is the pollution graph guided loss function;

[0119] Preferably, the adversarial loss weight The value of is 0.1, and the path consistency loss weight The value of is 0.4, and the pollution map guides the loss weight The value of is 0.5;

[0120] Step S35: Sewage source tracing model training, specifically, through the virtual and real data reconstruction, the construction of the pollution path generation network, the construction of the dual path discriminator network and the loss improvement design, the generation adversarial method is used to train the sewage source tracing model to obtain the sewage source tracing model Model VRG ;

[0121] Step S36: Sewage source tracing enhancement, specifically, using the sewage source tracing model Model based on the water quality health multi-dimensional reference data and the groundwater sensor optimization data VRG , conduct sewage source tracing enhancement and obtain probabilistic groundwater pollution source tracing analysis data;

[0122] The probabilistic groundwater pollution source tracing analysis data includes pollution source heat map data, pollution path map data and groundwater pollution source tracing credibility score reference data.

[0123] By performing the above operations, in the existing sewage source tracing enhancement methods, there are technical problems such as traditional source tracing methods mostly relying on the concentration inversion of pollutants and inference of hydrogeological maps, which are easily limited by the problems of sparse sample points and incomplete data, and are difficult to cope with complex scenarios of multi-source superposition or hidden pollution, resulting in unclear tracing results and inaccurate path judgment. This solution creatively adopts the improved adversarial generative network method of virtual-real source tracing modeling to enhance pollution source tracing, which not only improves the accuracy of pollution source positioning, but also reveals the diffusion trend of pollutants in groundwater, enabling regulatory units to discover hidden pollution earlier and accurately lock in the source of pollution, significantly enhancing the foresight and pertinence of pollution prevention and control.

[0124] Example 5, see Figure 1 and Figure 2 In step S4, the water quality prediction and warning is used to realize predictive groundwater quality warning, specifically, based on the probabilistic groundwater pollution source tracing analysis data, and by setting the intelligent periodic warning threshold, water quality prediction and warning are performed to obtain groundwater quality prediction and warning reference data;

[0125] The intelligent periodic warning thresholds include short-term cycle (1 to 7 days), medium-term cycle (7 to 30 days) and long-term cycle (more than 30 days);

[0126] By building a dynamic adjustment threshold fitting mechanism, early warning classification based on risk level is carried out to obtain reference data for groundwater quality prediction and early warning;

[0127] The calculation formula of the dynamic adjustment threshold fitting mechanism is:

[0128] ;

[0129] In the formula, is the periodic dynamic warning threshold, is the mean value of the water quality factor, and the water quality factor is used to represent the reference data of the credibility score of groundwater pollution source tracing in the probabilistic groundwater pollution source tracing analysis data, p is the water quality factor index, is the pollution sensitivity coefficient, is the standard deviation of the water quality factor;

[0130] The warning levels include low risk, medium risk and high risk;

[0131] Preferably, Table 2 is a reference table of warning judgment conditions for the warning classification, such as Table, x t It is the credibility score of groundwater pollution source tracing, which is used to represent the pollution prediction value at time t, the warning classification, and water quality prediction and warning.

[0132] Table 2 Reference table of warning judgment conditions for warning classification

[0133]

[0134] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S5, the groundwater quality supervision is used to conduct comprehensive groundwater quality supervision and management in combination with the water quality health, pollution source tracing analysis results and water quality prediction and early warning results. Specifically, based on the multi-dimensional reference data of water quality health, multi-dimensional groundwater health modeling is performed, and basic management of groundwater quality is assisted through visual analysis. In addition, comprehensive groundwater quality supervision is performed based on the probabilistic groundwater pollution source tracing analysis data and the groundwater quality prediction and early warning reference data to obtain a reference plan for groundwater quality supervision.

[0135] Embodiment 7, see Figure 1 and Figure 2 , this embodiment is based on the above embodiment, and the present invention provides an intelligent groundwater quality supervision system, including a heterogeneous sensor acquisition module, a water quality benchmark modeling module, a sewage tracing enhancement module, a water quality prediction and early warning module and a groundwater quality supervision module;

[0136] The heterogeneous sensing acquisition module is used for heterogeneous sensing acquisition, obtains groundwater sensing optimization data through heterogeneous sensing acquisition, and sends the groundwater sensing optimization data to the water quality benchmark modeling module and the sewage tracing enhancement module;

[0137] The water quality benchmark modeling module is used for water quality benchmark modeling, and obtains multi-dimensional reference data of water quality health through water quality benchmark modeling, and sends the multi-dimensional reference data of water quality health to the sewage tracing enhancement module and the groundwater quality supervision module;

[0138] The sewage source tracing enhancement module is used for sewage source tracing enhancement, and obtains probabilistic groundwater pollution source tracing analysis data through sewage source tracing enhancement, and sends the probabilistic groundwater pollution source tracing analysis data to the water quality prediction and early warning module;

[0139] The water quality prediction and warning module is used for water quality prediction and warning, obtains groundwater quality prediction and warning reference data through water quality prediction and warning, and sends the groundwater quality prediction and warning reference data to the groundwater quality supervision module;

[0140] The groundwater quality supervision module is used for groundwater quality supervision, and a groundwater quality supervision reference plan is obtained through groundwater quality supervision.

[0141] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0142] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0143] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent groundwater quality monitoring method, characterized in that: The method comprises the following steps: Step S1: heterogeneous sensor collection to obtain groundwater sensor optimization data; Step S2: water quality benchmark modeling, using an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm to perform water quality benchmark modeling to obtain multi-dimensional reference data of water quality health, including the following steps: Step S21: multi-scale feature extraction, by introducing a void convolution to replace the standard one-dimensional convolution layer, to perform time series perception enhancement improvement; Step S22: spatiotemporal dynamic benchmark modeling; Step S23: water quality health calculation, by calculating the Mahalanobis distance in the groundwater sensor optimization data and the temporal geological feature data, and constructing a multi-dimensional health fusion mechanism to obtain water quality health output; Step S24: water quality benchmark modeling training; Step S25: water quality benchmark modeling; Step S3: Sewage source tracing enhancement, using the improved adversarial generative network method of virtual and real source tracing modeling to enhance pollution source tracing, and obtain probabilistic groundwater pollution source tracing analysis data, including the following steps: Step S31: Virtual and real data reconstruction, by constructing a groundwater hydrodynamic model, setting boundary conditions and initial pollution sources, constructing a virtual pollution diffusion data map, and obtaining virtual pollution diffusion data; Step S32: Constructing a pollution path generation network; Step S33: Constructing a dual-path discriminator network; Step S34: Loss improvement design; Step S35: Sewage source tracing model training; Step S36: Sewage source tracing enhancement; Step S4: water quality prediction and warning, by setting intelligent periodic warning thresholds, water quality prediction and warning are performed to obtain groundwater quality prediction and warning reference data; Step S5: Groundwater quality supervision, obtaining a groundwater quality supervision reference plan.

2. An intelligent groundwater quality monitoring method according to claim 1, characterized in that: In step S1, the heterogeneous sensor acquisition is used to collect the original data set required for groundwater quality supervision, specifically, to obtain the original groundwater quality data set through sensor array data acquisition, and to obtain groundwater sensor optimization data by optimizing and preprocessing the original groundwater quality data set; The original groundwater quality data set includes physical and chemical parameter data, ion concentration data and sensor metadata; the optimization preprocessing steps include data cleaning, data alignment, noise filtering, data standardization, data normalization, feature optimization and data labeling processing; the groundwater sensor optimization data includes optimized physical and chemical parameter data, optimized ion concentration data and optimized sensor metadata.

3. An intelligent groundwater quality monitoring method according to claim 2, characterized in that: In step S2, the water quality benchmark modeling is used to construct a dynamic water quality assessment benchmark model, specifically, based on the groundwater sensor optimization data, an improved spatiotemporal graph convolutional network method combined with a spatiotemporal dynamic baseline algorithm is used to perform water quality benchmark modeling to obtain multidimensional reference data of water quality health, including the following steps: Step S21: multi-scale feature extraction, specifically, extracting multi-scale feature data from the groundwater sensor optimization data by constructing a one-dimensional convolutional long-short term neural network as a hybrid encoder; the multi-scale feature data includes local features and global features; the local features are used to represent hour-level sensor time data, and the global features are used to represent seasonal-level sensor time data; The one-dimensional convolutional long-term short-term neural network is improved by introducing a dilated convolution to replace the standard one-dimensional convolution layer to enhance the temporal perception; Step S22: spatiotemporal dynamic baseline modeling, specifically, performing spatiotemporal dynamic baseline modeling through spatial correlation modeling and temporal dynamic modeling to obtain temporal geological characteristic data; The spatial correlation modeling specifically constructs a groundwater spatial relationship graph structure by modeling the hydrogeological conductivity coefficient and the pollutant diffusion, and calculates the spatial correlation feature data through a standard graph convolution network; The temporal dynamics modeling is specifically performed by using an improved spatiotemporal graph convolutional network to extract spatiotemporal attention. The improved spatiotemporal graph convolutional network specifically introduces a gated spatiotemporal attention mechanism on the basis of a standard spatiotemporal graph convolutional network to perform spatiotemporal feature processing to obtain temporal dynamics modeling feature data; The temporal geological characteristic data is specifically obtained by combining the spatial correlation characteristic data and the temporal dynamic modeling characteristic data to output as a spatiotemporal dynamic baseline, and used as a reference benchmark for water quality health; Step S23: water quality health calculation, specifically, obtaining deviation data by calculating the Mahalanobis distance between the groundwater sensor optimization data and the temporal geological characteristic data, and calculating the water quality health by constructing a multi-dimensional health fusion mechanism to obtain a water quality health output; Step S24: water quality benchmark modeling training, specifically, performing water quality benchmark modeling training through the multi-scale feature extraction, the spatiotemporal dynamic benchmark modeling and the water quality health calculation to obtain a water quality benchmark model Model WHI ; Step S25: water quality benchmark modeling, specifically, using the water quality benchmark model Model based on the groundwater sensor optimization data WHI , conduct water quality benchmark modeling and obtain multi-dimensional reference data on water quality health.

4. An intelligent groundwater quality monitoring method according to claim 3, characterized in that: In step S2, the water quality health multidimensional reference data includes water quality health multidimensional quantitative output and water quality health comprehensive health rating; The multi-dimensional quantitative output of water quality health includes chemical dimension, ecological dimension and toxicity dimension; The water quality health level is a comprehensive health level rating, with specific ratings including excellent, good, fair and poor.

5. An intelligent groundwater quality monitoring method according to claim 4, characterized in that: In step S3, the sewage source tracing enhancement is used to locate the pollution source and restore the pollution diffusion path. Specifically, based on the multi-dimensional reference data of water quality health and the groundwater sensor optimization data, an improved adversarial generative network method of virtual-real source tracing modeling is used to enhance pollution source tracing and obtain probabilistic groundwater pollution source tracing analysis data, including the following steps: Step S31: virtual and real data reconstruction, specifically, based on the water quality health multidimensional reference data, using the groundwater sensor optimization data corresponding to the water quality health multidimensional reference data including medium rating and poor rating in the water quality health comprehensive health rating as the actual data input for the virtual and real data reconstruction, and constructing a groundwater hydrodynamic model, setting boundary conditions and initial pollution sources, and constructing a virtual pollution diffusion data map based on the actual data input to obtain virtual pollution diffusion data; The steps of constructing the virtual pollution diffusion data map include calculating the pollution intensity index, generating the pollution heat map interpolation, and generating the pollution mask binarization; the virtual pollution diffusion data specifically refers to the binary pollution high-risk mask map data; Step S32: constructing a pollution path generation network, specifically by collecting the real observed pollution diffusion data corresponding to the virtual pollution diffusion data, constructing a generator, obtaining a pollution path generation network, and generating pollution path generation data, the pollution path generation data includes generating predicted pollution path data and generating pollution source diffusion heat map data; Step S33: constructing a dual-path discriminator network, specifically, respectively constructing a coarse-grained discriminator and a fine-grained discriminator, performing dual-path discriminator network construction, and obtaining a discriminator network; The coarse-grained discriminator is used to evaluate the overall trend of the pollution path and the regional pollution probability; the fine-grained discriminator is used to analyze the path change characteristics of local pollution diffusion; Step S34: loss improvement design, specifically, performing loss improvement design for generative adversarial training by constructing adversarial loss, path consistency loss and pollution graph guided loss, and obtaining a comprehensive loss function by weighted combination of the adversarial loss, path consistency loss and pollution graph guided loss; Step S35: Sewage source tracing model training, specifically, through the virtual and real data reconstruction, the construction of the pollution path generation network, the construction of the dual path discriminator network and the loss improvement design, the generation confrontation method is used to train the sewage source tracing model to obtain the sewage source tracing model Model VRG ; Step S36: Sewage source tracing enhancement, specifically, using the sewage source tracing model Model based on the water quality health multi-dimensional reference data and the groundwater sensor optimization data VRG , enhance sewage source tracing and obtain probabilistic groundwater pollution source tracing analysis data.

6. An intelligent groundwater quality monitoring method according to claim 5, characterized in that: In step S3, the probabilistic groundwater pollution source tracing analysis data includes pollution source heat map data, pollution path map data and groundwater pollution source tracing credibility score reference data.

7. An intelligent groundwater quality monitoring method according to claim 6, characterized in that: In step S4, the water quality prediction and warning is used to realize predictive groundwater quality warning, specifically, based on the probabilistic groundwater pollution source tracing analysis data, and by setting intelligent periodic warning thresholds, water quality prediction and warning are performed to obtain groundwater quality prediction and warning reference data; In step S5, the groundwater quality supervision is used to conduct comprehensive groundwater quality supervision and management by combining the water quality health, pollution source tracing analysis results and water quality prediction and early warning results. Specifically, based on the multi-dimensional reference data of water quality health, multi-dimensional groundwater health modeling is performed, and basic management of groundwater quality is assisted through visual analysis. Comprehensive groundwater quality supervision is conducted based on the probabilistic groundwater pollution source tracing analysis data and the groundwater quality prediction and early warning reference data to obtain a groundwater quality supervision reference plan.

8. An intelligent groundwater quality monitoring system, used to implement an intelligent groundwater quality monitoring method as claimed in any one of claims 1 to 7, characterized in that: It includes heterogeneous sensor acquisition module, water quality benchmark modeling module, sewage source tracing enhancement module, water quality prediction and early warning module and groundwater quality supervision module; The heterogeneous sensing acquisition module is used for heterogeneous sensing acquisition, obtains groundwater sensing optimization data through heterogeneous sensing acquisition, and sends the groundwater sensing optimization data to the water quality benchmark modeling module and the sewage tracing enhancement module; The water quality benchmark modeling module is used for water quality benchmark modeling, and obtains multi-dimensional reference data of water quality health through water quality benchmark modeling, and sends the multi-dimensional reference data of water quality health to the sewage tracing enhancement module and the groundwater quality supervision module; The sewage source tracing enhancement module is used for sewage source tracing enhancement, and obtains probabilistic groundwater pollution source tracing analysis data through sewage source tracing enhancement, and sends the probabilistic groundwater pollution source tracing analysis data to the water quality prediction and early warning module; The water quality prediction and warning module is used for water quality prediction and warning, obtains groundwater quality prediction and warning reference data through water quality prediction and warning, and sends the groundwater quality prediction and warning reference data to the groundwater quality supervision module; The groundwater quality supervision module is used for groundwater quality supervision, and a groundwater quality supervision reference plan is obtained through groundwater quality supervision.

Images