Underground water quality factor dynamic migration analysis method and system

Abstract

The invention relates to the technical field of water quality monitoring and analysis, in particular to an underground water quality factor dynamic migration analysis method and system. The method comprises the following steps: acquiring historical monitoring data of underground water, including acquiring water quality factor concentration, water level, water temperature, permeability and porosity; obtaining the concentration of the fed electrolyte, and calculating the underground water flow velocity according to the electrolyte diffusion velocity; constructing a groundwater quality factor migration model based on a convection-dispersion equation according to the groundwater flow velocity, and introducing a neural network and a generative adversarial network to carry out model parameter optimization on the groundwater quality factor migration model; the optimized groundwater quality factor migration model is subjected to model solution, the calibrated groundwater quality factor migration model is used for water quality factor migration trajectory analysis, the water quality factor concentration value of each grid point at different moments can be obtained in real time, and analysis of groundwater quality factor dynamic migration is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality monitoring and analysis, and in particular to a method and system for analyzing the dynamic migration of groundwater quality factors. Background Art

[0002] In the existing technology, groundwater is a key freshwater resource, and its quality is closely related to human health and sustainable economic development. With the progress of industry and the increase in population, the problem of groundwater pollution has become increasingly serious, making the tracking and tracing of groundwater pollution sources a key task in the field of environmental monitoring and governance. Traditional groundwater monitoring methods rely on setting up groundwater monitoring wells and then taking samples and analyzing them regularly. However, this method has obvious drawbacks. It not only consumes a lot of time and manpower, but also makes it difficult to achieve real-time analysis of the dynamic migration of groundwater quality factors.

[0003] Among the existing technologies, some solutions based on the Internet of Things and big data technologies can collect groundwater quality data in real time by deploying sensors in groundwater, and transmit it to the data processing center for storage and analysis with the help of wireless networks. Although such technologies can realize real-time analysis of the dynamic migration of groundwater quality factors, they still have many shortcomings. On the one hand, these technologies focus on data collection and storage, and lack simulation of the migration trajectory of groundwater quality factors, resulting in the inability to accurately determine the source location, diffusion range and potential impact of pollutants. On the other hand, these technologies are costly and require a large number of sensors, wireless network equipment, and the construction and maintenance of data processing centers, all of which require high costs. At this stage, a method and system for analyzing the dynamic migration of groundwater quality factors is needed. Summary of the invention In order to solve the problems of lack of simulation of groundwater quality factor migration trajectory and high construction cost in the prior art, the present invention provides a groundwater quality factor dynamic migration analysis method and system.

[0004] In the first aspect, the present invention provides a method for analyzing the dynamic migration of groundwater quality factors, which adopts the following technical solution: A method for analyzing the dynamic migration of groundwater quality factors, comprising: Obtain historical groundwater monitoring data, including water quality factor concentration, water level, water temperature, permeability and porosity; Obtaining the concentration of the electrolyte to be added, and calculating the electrolyte diffusion rate based on the historical monitoring data obtained, and calculating the groundwater flow rate according to the electrolyte diffusion rate; According to the groundwater flow velocity, a groundwater quality factor migration model based on the convection-diffusion equation is constructed, and neural networks and generative adversarial networks are introduced to optimize the model parameters of the groundwater quality factor migration model. Solving the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; Calibrate the parameters of the groundwater quality factor transport model based on the solution results; The calibrated groundwater quality factor migration model is used to analyze the migration trajectory of water quality factors, and the source location and diffusion range of pollutants are determined based on the analysis results.

[0005] Further, the calculating of groundwater flow velocity according to the electrolyte diffusion velocity includes calculating the diffusion flux of the electrolyte in groundwater using Fick's law, and calculating the groundwater flow velocity according to the diffusion flux. The formula for calculating the groundwater flow velocity is: , in, is represented by diffusion flux, C is represented by the average concentration of electrolyte, v is represented by groundwater flow velocity, x is represented by spatial coordinate, It is expressed as the electrolyte migration flux due to diffusion caused by concentration gradient.

[0006] Furthermore, the groundwater quality factor migration model based on the convection-diffusion equation is constructed according to the groundwater flow velocity, including calculating the source-sink term based on the acquired historical monitoring data and using the Langmuir adsorption model, and substituting the source-sink term into the source-sink term to construct the groundwater quality factor migration model. The groundwater quality factor migration model expression is: , in, It is represented by the concentration of water quality factors, t is represented by time, x is represented by spatial coordinates, v is represented by groundwater flow velocity, D is represented by diffusion coefficient, S is represented by source and sink terms, is expressed as the diffusion effect due to the concentration gradient, Expressed as the effect of convection on the migration of water quality factors.

[0007] Furthermore, the introduction of neural networks and generative adversarial networks optimizes model parameters of the groundwater quality factor migration model, including constructing a multi-layer perceptron neural network consisting of an input layer, a hidden layer, and an output layer for predicting the diffusion coefficient D and the groundwater flow velocity v, and then introducing an adversarial network structure consisting of multiple generators and discriminators, and optimizing the model parameters through adversarial training of the generator and the discriminator.

[0008] Furthermore, the finite difference method is used to discretize the groundwater quality factor migration model to obtain a discretized equation, including dividing the solution area into a series of grid points and discretizing the partial differential equation on these grid points, wherein for the time derivative Using forward difference, for the second-order spatial derivative and the first-order spatial derivative The central difference is used to obtain the discretized equation. Through continuous iterative calculation and step-by-step advancement of time, the water quality factor concentration value of each grid point at different times is solved.

[0009] Furthermore, the parameters of the groundwater quality factor migration model are calibrated according to the solution results, including obtaining the calculated value and the actual measured concentration value of the water quality factor concentration, using the root mean square error as the objective function to measure the difference between the calculated value and the actual measured concentration value, calculating the gradient of the objective function to the parameter, and updating the parameter according to a certain step size.

[0010] Furthermore, the calibrated groundwater water quality factor migration model is used to analyze the water quality factor migration trajectory, including plotting the water quality factor concentration values ​​of each grid point at different times into a three-dimensional coordinate graph, extracting the change curve of the central coordinate of the water quality factor concentration over time and space, and quantitatively describing the migration trajectory based on the change curve.

[0011] In the second aspect, a groundwater quality factor dynamic migration analysis system includes: The data acquisition module is configured to: acquire historical monitoring data of groundwater, including acquiring water quality factor concentration, water level, water temperature, permeability and porosity; The calculation module is configured to: obtain the concentration of the electrolyte, calculate the electrolyte diffusion rate based on the historical monitoring data obtained, and calculate the groundwater flow rate according to the electrolyte diffusion rate; The model module is configured to: construct a groundwater quality factor transport model based on the convection-diffusion equation according to the groundwater flow velocity, and introduce a neural network and a generative adversarial network to optimize the model parameters of the groundwater quality factor transport model; The model solving module is configured to: solve the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; The calibration module is configured to: calibrate the parameters of the groundwater quality factor migration model according to the solution results; The output module is configured to: use the calibrated groundwater quality factor migration model to analyze the migration trajectory of water quality factors, and determine the source location and diffusion range of pollutants based on the analysis results.

[0012] In a third aspect, the present invention provides a computer-readable storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device as described in the method for analyzing the dynamic migration of groundwater quality factors.

[0013] In a fourth aspect, the present invention provides a terminal device including a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executing the method for dynamic migration analysis of groundwater quality factors.

[0014] In summary, the present invention has the following beneficial technical effects: 1. The present invention takes into account a variety of factors, including historical monitoring data, electrolyte diffusion rate, groundwater flow rate, etc., and introduces advanced neural networks and generative adversarial networks to optimize model parameters. It can more accurately simulate the dynamic migration of groundwater quality factors and improve the accuracy of analysis. By using the calibrated model to analyze the migration trajectory of water quality factors, the source location and diffusion range of pollutants can be more accurately determined, providing more targeted measures for pollution prevention and control.

[0015] 2. The present invention divides the solution area into grid points and discretizes them, which can convert continuous partial differential equations into discrete algebraic equations, thereby facilitating rapid solution on a computer and improving computing efficiency. By continuously iterative calculations and gradually advancing time, the water quality factor concentration value of each grid point at different times can be obtained in real time, and the dynamic migration of groundwater quality factors can be analyzed, reducing the installation of sensors and a series of network equipment, saving manpower, material resources and time costs to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the overall process of a method for analyzing the dynamic migration of groundwater quality factors according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0018] Example 1 Reference Figure 1 , a method for analyzing the dynamic migration of groundwater quality factors in this embodiment includes: Obtain historical groundwater monitoring data, including water quality factor concentration, water level, water temperature, permeability and porosity; Obtaining the concentration of the electrolyte to be added, and calculating the electrolyte diffusion rate based on the historical monitoring data obtained, and calculating the groundwater flow rate according to the electrolyte diffusion rate; According to the groundwater flow velocity, a groundwater quality factor migration model based on the convection-diffusion equation is constructed, and neural networks and generative adversarial networks are introduced to optimize the model parameters of the groundwater quality factor migration model. Solving the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; Calibrate the parameters of the groundwater quality factor transport model based on the solution results; The calibrated groundwater quality factor migration model is used to analyze the migration trajectory of water quality factors, and the source location and diffusion range of pollutants are determined based on the analysis results.

[0019] Specifically, a method for analyzing the dynamic migration of groundwater quality factors in this embodiment includes the following steps: like Figure 1 As shown, S1. Obtain historical monitoring data of groundwater, including water quality factor concentration, water level, water temperature, permeability and porosity; Collect historical groundwater monitoring data, including water quality factor concentration (such as heavy metal ion concentration, organic matter concentration, etc.), water level, water temperature, aquifer characteristics (such as permeability, porosity), geological structure, etc. Use database management system to organize and store these data to ensure data integrity and accuracy. S2. Obtaining the concentration of the electrolyte, and calculating the electrolyte diffusion rate based on the historical monitoring data obtained, and calculating the groundwater flow rate according to the electrolyte diffusion rate; A specific electrolyte solution, such as sodium chloride solution, is placed in the monitoring well. Highly sensitive electrodes are carefully set in adjacent monitoring wells to accurately measure the current between the monitoring wells. When calculating the diffusion rate of the electrolyte in the groundwater, Fick's law is used: , in, It is expressed as diffusion flux, which represents the amount of substance passing through a unit area per unit time, that is, the diffusion rate of the electrolyte; C represents the average concentration of the electrolyte, and x represents the spatial coordinate. Expressed as a concentration gradient, for the concentration gradient , it is necessary to measure the electrolyte concentration C at different locations and then calculate the rate of change of the concentration in space. For example, at two adjacent monitoring points and The concentrations measured at and , then the concentration gradient can be approximately calculated as: ,Then the relationship between the electrolyte diffusion rate and the groundwater flow rate is established. Because the migration of electrolytes is affected by both diffusion and convection under the action of groundwater flow, the relationship between the migration of electrolytes and groundwater flow is described by a formula based on convection-diffusion theory. The formula is expressed as: , Then the calculation formula of groundwater flow velocity is obtained: , in, is represented by diffusion flux, C is represented by the average concentration of electrolyte, v is represented by groundwater flow velocity, x is represented by spatial coordinate, It is expressed as the electrolyte migration flux caused by diffusion due to concentration gradient, taking into account the characteristics of the aquifer (such as permeability, porosity), geological structure (such as rock layer distribution, faults) and water chemical conditions (such as pH, ionic strength), water level, water temperature and other factors. This is because these factors will affect the diffusion coefficient D, concentration distribution and the speed and direction of groundwater flow, so the calculated groundwater flow velocity needs to be corrected and optimized accordingly. It is necessary to determine the various parameters in the above relationship through multiple measurements and data analysis to improve the accuracy of groundwater flow velocity calculation.

[0020] S3. According to the groundwater flow velocity, a groundwater quality factor migration model based on the convection-diffusion equation is constructed, and neural networks and generative adversarial networks are introduced to optimize the model parameters of the groundwater quality factor migration model; In the groundwater quality factor migration model, the source-sink term is a quantity that comprehensively considers various factors that cause the concentration of water quality factors to increase or decrease. "Source" refers to factors that increase the concentration of water quality factors, such as pollutants produced by industrial and agricultural activities, and "sink" refers to factors that reduce the concentration of water quality factors. Soil and rock particles in the aquifer may adsorb some water quality factors, causing them to separate from the water, thereby reducing the concentration in the water. The process of determining the source-sink term, first, conduct detailed field investigations and surveys to understand the land use type, industrial distribution, agricultural activity pattern, landfill location and other information in the study area, preliminarily determine the possible types and locations of pollution sources, and then collect historical groundwater monitoring data on the spot. Finally, the Langmuir adsorption model is used and the parameters obtained on the spot are substituted into it to calculate the preliminary estimate of the source-sink term. The Langmuir adsorption model assumes that there are uniform adsorption sites on the adsorbent surface and there is no interaction between adsorbate molecules. Its expression is: , in, is the concentration of the solute in the solution, q is the amount of solute adsorbed per unit mass of the adsorbent, is the maximum adsorption capacity, is a constant related to adsorption energy. In addition, the first-order reaction kinetic model is used for the natural degradation of microorganisms. The value of the source-sink term is equal to the input of the pollution source minus the removal of water quality factors by the adsorption of the aquifer and the degradation of microorganisms. Substituting the source-sink term into the groundwater quality factor migration model, the groundwater quality factor migration model is expressed as: , in, It is expressed as the concentration of water quality factors (the concentration of a specific pollutant in water quality, including heavy metal concentration, etc.), t is time, x is spatial coordinates, v is groundwater flow velocity, D is diffusion coefficient, S is source and sink term, is expressed as the diffusion effect due to the concentration gradient, It is expressed as the effect of convection on the migration of water quality factors, Indicates the concentration of water quality factors The rate of change over time t, then the neural network and generative adversarial network are introduced to optimize the model parameters of the groundwater quality factor migration model, and a multi-layer perceptron neural network consisting of an input layer, a hidden layer, and an output layer is constructed. The number of input layer nodes matches the number of input features, that is, the number of factors such as geological structure, aquifer characteristics, and water temperature. The number of output layer nodes is 2, corresponding to the predicted values ​​of the diffusion coefficient and groundwater flow velocity, respectively, which are used to predict the diffusion coefficient D and groundwater flow velocity v. The historical groundwater monitoring data obtained are divided into training set, validation set, and test set. The training set data is used to train the multi-layer perceptron neural network, and the network is calculated through forward propagation. The output is then back-propagated based on the difference between the output and the actual value to update the network weights and biases. The trained multi-layer perceptron neural network can accept new input features (geological structure, aquifer characteristics, water temperature, etc.) and predict the corresponding diffusion coefficient and groundwater flow rate. Finally, an adversarial network consisting of multiple generators and discriminators is introduced. The generator tries to generate a realistic distribution of pollution sources, and the discriminator determines whether the distribution it generates is real. Each generator is connected to all discriminators, which means that the pollution source distribution generated by the generator will be sent to all discriminators at the same time. Similarly, the discriminator is also connected to all generators, and it will receive outputs from all generators. In this way, the discriminator can compare and judge the distributions generated by different generators, so as to better identify the difference between the real pollution source distribution and the generated distribution. The discriminator will independently evaluate the distribution generated by each generator and give corresponding feedback. First, the generator randomly generates some pollution source distribution samples, and the discriminator judges these generated samples and the real pollution source distribution samples, and gives true and false labels. According to the feedback from the discriminator, the generator adjusts its own parameters to generate a more realistic pollution source distribution, trying to deceive the discriminator. The discriminator also continuously updates its parameters to improve its judgment ability. This adversarial process is continuously iterated. The specific iterative training method can be to first fix the parameters of the discriminator and train the generator. At this stage, the generator will adjust its parameters according to the feedback from the discriminator to generate a distribution that is more in line with the discriminator's expectations. Then, the parameters of the generator are fixed and the discriminator is trained. The discriminator will update its parameters according to the distribution generated by the generator to improve its judgment ability. This process is repeated over and over again to continuously optimize the performance of the generator and discriminator until the model converges and the generator can generate diversified pollution source distribution scenarios that are in line with actual conditions. The model parameters are optimized through adversarial training between the two. Multiple generators generate a variety of different pollution source distribution patterns, thereby enriching the output possibilities of the model. Multiple discriminators evaluate and provide feedback on the distribution generated by the generator from different angles.

[0021] S4, solving the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; In order to more accurately analyze and calculate the changes in water quality factors in this area, we divide the land into many small grids, which are the so-called grid points. For example, suppose the groundwater area we want to study is a rectangle, and a certain number of equally spaced points are divided in the horizontal direction (x direction) and the vertical direction (y direction). These points constitute a grid. Each grid point has its own coordinate position and corresponding value (such as the concentration value of the water quality factor). Through this division, we can discretize the continuous solution area and transform the complex partial differential equation into algebraic equations on these discrete grid points. The solution area is divided into a series of grid points, and the partial differential equations are discretized on these grid points, thereby transforming them into discretized equations. In one-dimensional space, the spatial range is divided into equally spaced , the time step is , for the time derivative of the groundwater quality factor transport model The forward difference approximation is , for the second-order spatial derivative The central difference approximation is First-order spatial derivative The central difference approximation is , thus obtaining the discretized equation: , in, represents the water quality factor concentration at position i at the nth time step, It represents the concentration of water quality factors at position i at the n+1th time step. And so on. By continuously iterating and calculating this discretized equation and advancing time step by step, the concentration value of each grid point at different times can be solved, which can be used to simulate the migration process of groundwater quality factors.

[0022] S5. Calibrate the parameters of the groundwater quality factor migration model according to the solution results; Obtain accurate measured values ​​of water quality factor concentrations at multiple monitoring points and at different time points , using the root mean square error as a measure of the calculated value of water quality factor concentration and the actual measured concentration value The objective function of the difference, the calculation formula of the root mean square error is: , is the concentration value calculated by the model at the i-th data point, is the actual monitored concentration value of the ith data point, N represents the total number of data points, and then the parameter value is gradually adjusted along the negative gradient direction of the objective function through the gradient descent algorithm, so that the objective function value gradually decreases, and then the parameter is updated according to a certain step size, where the step size is set according to the difference value.

[0023] S6. Use the calibrated groundwater quality factor migration model to analyze the migration trajectory of water quality factors, and determine the source location and diffusion range of pollutants based on the analysis results.

[0024] The water quality factor concentration value of each grid point at different times is plotted into a three-dimensional coordinate graph. The three axes of the three-dimensional coordinate graph represent the three dimensions of space respectively. The coordinates (x, y, z) of each grid point are associated with the corresponding water quality factor concentration value, and the water quality factor concentration is marked in the three-dimensional coordinates. The center position of the area with the highest concentration value is calculated, and the area with the highest concentration value is found. Assuming that the area contains multiple grid points, the average value of the coordinates of these grid points is calculated as the coordinates of the center position of the concentration, and the curve of the change of the center coordinates of the water quality factor concentration with time and space is extracted. At different time points, the above steps are repeated to obtain the change of the coordinates of the center position of the concentration with time. These coordinates are plotted in a coordinate system with time as the horizontal axis and space coordinates as the vertical axis to obtain the curve of the change of the center coordinates of the water quality factor concentration with time and space. The change curve is analyzed, and the moving distance and direction of the concentration center between adjacent time points are calculated to obtain the moving speed and direction. The residence time of the concentration center in different areas and the change of the concentration value are observed. Based on this information, the possible source location of the pollutant can be determined. For example, if the concentration center gradually moves along a certain direction and the diffusion range gradually expands, it can be inferred that the pollution source may be upstream of this direction. The migration trajectory is quantitatively described according to the change curve, which shows the movement trajectory of the concentration center in space at different time points.

[0025] Example 2 The difference between this embodiment and embodiment 1 is that this embodiment provides a groundwater quality factor dynamic migration analysis system, including: The data acquisition module is configured to: acquire historical monitoring data of groundwater, including acquiring water quality factor concentration, water level, water temperature, permeability and porosity; The calculation module is configured to: obtain the concentration of the electrolyte, calculate the electrolyte diffusion rate based on the historical monitoring data obtained, and calculate the groundwater flow rate according to the electrolyte diffusion rate; The model module is configured to: construct a groundwater quality factor transport model based on the convection-diffusion equation according to the groundwater flow velocity, and introduce a neural network and a generative adversarial network to optimize the model parameters of the groundwater quality factor transport model; The model solving module is configured to: solve the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; The calibration module is configured to: calibrate the parameters of the groundwater quality factor migration model according to the solution results; The output module is configured to: use the calibrated groundwater quality factor migration model to analyze the migration trajectory of water quality factors, and determine the source location and diffusion range of pollutants based on the analysis results.

[0026] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for analyzing the dynamic migration of groundwater quality factors.

[0027] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executing the method for dynamic migration analysis of groundwater quality factors.

[0028] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for analyzing the dynamic migration of groundwater quality factors, characterized in that: include: Obtain historical groundwater monitoring data, including water quality factor concentration, water level, water temperature, permeability and porosity; Obtaining the concentration of the electrolyte to be added, and calculating the electrolyte diffusion rate based on the historical monitoring data obtained, and calculating the groundwater flow rate according to the electrolyte diffusion rate; According to the groundwater flow velocity, a groundwater quality factor migration model based on the convection-diffusion equation is constructed, and neural networks and generative adversarial networks are introduced to optimize the model parameters of the groundwater quality factor migration model. Solving the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; Calibrate the parameters of the groundwater quality factor transport model based on the solution results; The calibrated groundwater quality factor migration model is used to analyze the migration trajectory of water quality factors, and the source location and diffusion range of pollutants are determined based on the analysis results.

2. A method for analyzing the dynamic migration of groundwater quality factors according to claim 1, characterized in that: The calculating of groundwater flow velocity according to the electrolyte diffusion velocity includes calculating the diffusion flux of the electrolyte in groundwater using Fick's law, and calculating the groundwater flow velocity according to the diffusion flux. The formula for calculating the groundwater flow velocity is: , in, is represented by diffusion flux, C is represented by the average concentration of electrolyte, v is represented by groundwater flow velocity, x is represented by spatial coordinate, Expressed as the electrolyte migration flux due to diffusion caused by concentration gradient.

3. A method for analyzing the dynamic migration of groundwater quality factors according to claim 2, characterized in that: The groundwater quality factor migration model based on the convection-diffusion equation is constructed according to the groundwater flow velocity, including calculating the source and sink terms based on the acquired historical monitoring data and using the Langmuir adsorption model, and substituting the source and sink terms into the groundwater quality factor migration model. The groundwater quality factor migration model is expressed as: ， in, It is represented by the concentration of water quality factors, t is represented by time, x is represented by spatial coordinates, v is represented by groundwater flow velocity, D is represented by diffusion coefficient, S is represented by source and sink terms, is expressed as the diffusion effect due to the concentration gradient, Expressed as the effect of convection on the migration of water quality factors.

4. A method for analyzing the dynamic migration of groundwater quality factors according to claim 3, characterized in that: The method introduces a neural network and a generative adversarial network to optimize the model parameters of the groundwater quality factor migration model, including constructing a multi-layer perceptron neural network composed of an input layer, a hidden layer and an output layer, which is used to predict the diffusion coefficient D and the groundwater flow velocity v. Then, an adversarial network structure composed of multiple generators and discriminators is introduced, and the model parameters are optimized through adversarial training of the generator and the discriminator.

5. A method for analyzing the dynamic migration of groundwater quality factors according to claim 4, characterized in that: The method of using the finite difference method to discretize the groundwater quality factor migration model to obtain a discretized equation includes dividing the solution area into a series of grid points and discretizing the partial differential equation on these grid points, wherein the time derivative Using forward difference, for the second-order spatial derivative and the first-order spatial derivative The central difference is used to obtain the discretized equation. Through continuous iterative calculation and step-by-step advancement of time, the water quality factor concentration value of each grid point at different times is solved.

6. A method for analyzing the dynamic migration of groundwater quality factors according to claim 5, characterized in that: The method of calibrating the parameters of the groundwater quality factor migration model based on the solution results includes obtaining the calculated value and the actual measured concentration value of the water quality factor concentration, using the root mean square error as the objective function to measure the difference between the calculated value and the actual measured concentration value, calculating the gradient of the objective function to the parameter, and updating the parameter according to a certain step size.

7. A method for analyzing the dynamic migration of groundwater quality factors according to claim 6, characterized in that: The calibrated groundwater water quality factor migration model is used to analyze the water quality factor migration trajectory, including plotting the water quality factor concentration value of each grid point at different times into a three-dimensional coordinate graph, extracting the change curve of the central coordinate of the water quality factor concentration over time and space, and quantitatively describing the migration trajectory based on the change curve.

8. A groundwater quality factor dynamic migration analysis system, executing the method according to any one of claims 1 to 7, characterized in that: include: The data acquisition module is configured to: acquire historical monitoring data of groundwater, including acquiring water quality factor concentration, water level, water temperature, permeability and porosity; The calculation module is configured to: obtain the concentration of the electrolyte, calculate the electrolyte diffusion rate based on the historical monitoring data obtained, and calculate the groundwater flow rate according to the electrolyte diffusion rate; The model module is configured to: construct a groundwater quality factor transport model based on the convection-diffusion equation according to the groundwater flow velocity, and introduce a neural network and a generative adversarial network to optimize the model parameters of the groundwater quality factor transport model; The model solving module is configured to: solve the optimized groundwater quality factor transport model, including discretizing the groundwater quality factor transport model using a finite difference method to obtain a discretized equation; The calibration module is configured to: calibrate the parameters of the groundwater quality factor migration model according to the solution results; The output module is configured to: use the calibrated groundwater quality factor migration model to analyze the migration trajectory of water quality factors, and determine the source location and diffusion range of pollutants based on the analysis results.

9. A computer-readable storage medium storing a plurality of instructions, characterized in that: The instructions are suitable for being loaded by a processor of a terminal device and executing a method for analyzing the dynamic migration of groundwater quality factors as described in any one of claims 1-7.

10. A terminal device, comprising a processor and a computer-readable storage medium, wherein the processor is used to implement each instruction; and the computer-readable storage medium is used to store multiple instructions, characterized in that: The instructions are suitable for being loaded by a processor and executing a method for analyzing the dynamic migration of groundwater quality factors as described in any one of claims 1-7.

Images