Water pollution tracing method based on artificial intelligence
The AI-based water pollution source tracing method using Delft3D and PCE models addresses the limitations of traditional methods by providing rapid and accurate source identification, enhancing emergency response capabilities.
Patent Information
- Application Number
- CN202411209281.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-30
AI Technical Summary
Existing methods for tracing water pollution sources are labor-intensive, time-consuming, and unable to timely respond to sudden water pollution incidents, lacking the flexibility and accuracy needed for effective emergency management.
A water pollution source tracing method based on artificial intelligence using the Delft3D model, Hadoop MapReduce, and PCE models to simulate water pollution events, optimize models, and invert data sets for rapid source identification.
Enables rapid and accurate identification of pollution sources, reducing computational costs and improving response times to sudden water pollution incidents while maintaining model accuracy and flexibility.
Smart Images

Figure FT_1 
Figure FT_2 
Figure FT_3
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of water quality monitoring, and in particular to a water pollution source tracing method based on artificial intelligence. Background Art
[0002] Sudden water pollution accidents are one of the most severe environmental pollution accidents, with diverse forms, suddenness, serious harm, difficulty in handling, randomness and uncertainty. In order to minimize the losses caused by sudden water pollution, after a sudden accident occurs, it is necessary to quickly locate and identify the leakage source and simulate the spatial impact range and degree of pollutants after the accident, so as to assist in taking control measures.
[0003] At present, the research methods for tracing the source of water pollution are generally field sampling and measurement methods or mathematical model methods. The field sampling method mainly uses tracing technologies such as isotope tracing, water pattern recognition, and ultraviolet spectroscopy to trace the source of pollutants. Although these methods have high stability and accuracy, they are labor-intensive, time-consuming, and difficult to timely investigate pollution sources, so they cannot effectively control pollution accidents in a timely manner. The mathematical model method has the advantages of flexibility, speed, and strong operability. It helps decision-makers understand the migration, diffusion, and temporal changes of pollutants in the water environment, and understand the impact of pollutants on the water environment, so as to make timely and accurate responses to emergencies. Summary of the invention
[0004] The present invention aims to solve at least one of the technical problems existing in the related art. Due to the concealment and suddenness of water pollution incidents, it is often impossible to know in advance the spatial location and intensity of the leakage source, and it is urgent to develop a method for quickly locating and identifying the leakage source. To this end, the present invention provides a water pollution source tracing method based on artificial intelligence.
[0005] The present invention provides a water pollution source tracing method based on artificial intelligence, comprising: S1: Select the basic model, simulate water pollution events under various discharge conditions, and obtain the water quality diffusion relationship data set; S2: establishing a first PCE model based on the water quality diffusion relationship data set; S3: Optimizing the approximation capability of the first PCE model according to the basic model to obtain a second PCE model; S4: inverting the water quality diffusion relationship data set by using the second PCE model, optimizing the second PCE model according to the inversion result, and obtaining a third PCE model; S5: Use the third PCE model as a source tracing model to trace the water pollution event to be traced.
[0006] According to the artificial intelligence-based water pollution source tracing method provided by the present invention, the basic model selected in step S1 is the Delft3D model.
[0007] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, in step S1, the step of simulating water pollution events under various discharge conditions further includes: S11: defining a simulation area, dividing the simulation area into grids, and obtaining a plurality of simulation grids; S12: setting simulation parameters and simulation variables, sampling the simulation parameters and selecting grids in the downstream area of the simulation area as observation point locations to obtain a variety of emission conditions; S13: Based on the distributed computing framework integrating the basic model, a variety of discharge conditions are substituted to simulate water pollution events and obtain a water quality diffusion relationship data set.
[0008] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, the simulation parameters in step S12 include simulation duration, calculation time step, gravity acceleration, and water body density, wherein the gravity acceleration and the water body density are the default values of the basic model; The simulation variables in step S12 include the discharge start time, discharge end time, and discharge amount of the water pollution event.
[0009] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, step S13 further includes: S131: Select Hadoop MapReduce as the distributed computing framework, integrate the basic model, and build a simulated cluster computing system; S132: using the simulation cluster computing system, respectively simulating water pollution events under various discharge conditions to obtain simulation results; S133: Obtaining original binary data of the simulation result, parsing the original binary data and storing it in a column-based database; S134: extracting the record set of the column database through Spark memory grid technology to obtain a water quality diffusion relationship data set.
[0010] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, the control equation of the basic model in step S1 is expressed as:
[0011]
[0012]
[0013] in, is the water level at the observation point, for Directional water flow, for Directional water flow, is the flow rate at the observation point, for Flow rate in direction, for Flow rate in direction, is the density of observation points, is the Coriolis force coefficient, Bottom of water Shear stress in the direction Bottom of water Shear stress in the direction is the horizontal viscosity coefficient, The water pressure.
[0014] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, the diffusion equation of the basic model in step S1 is expressed as:
[0015] in, is the concentration of pollutants, for The diffusion coefficient in the direction, for The diffusion coefficient in the direction, is the source-sink term, is the reaction term.
[0016] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, step S2 further includes: S21: fitting the data in the water quality diffusion relationship data set through the configuration point method in the proxy modeling base of the PCE algorithm to obtain a fitting data set; S22: Input the fitting data set into a preset model, and perform model training through the chaospy software library based on the Python programming language to obtain a first PCE model.
[0017] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, step S3 further includes: S31: Perform MC sampling on the simulation parameters to obtain sampling parameters; S32: inputting the sampling parameters into the basic model for simulation to obtain a sampling data set; S33: inputting the sampling parameters into the first PCE model to obtain a first PCE data set; S34: Perform multi-objective optimization on the first PCE model according to the sampling data set and the first PCE data set to obtain a second PCE model.
[0018] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, when performing multi-objective optimization on the first PCE model, the selected objectives include: An accuracy index, wherein the accuracy index is a Nash efficiency coefficient; High efficiency index, the high efficiency index is the time consumed by simulation calculation; Fidelity index, wherein the fidelity index is the Sobol parameter sensitivity index.
[0019] According to a water pollution source tracing method based on artificial intelligence provided by the present invention, step S4 further comprises: S41: randomly selecting an inversion data set from the sampling data set; S42: using the output data set in the inversion data set as observation values, and inputting the observation values into the second PCE model to obtain an inversion result; S43: taking the input data set in the inversion data set as the true value, comparing the inversion result with the true value, and optimizing the second PCE model with the Nash efficiency coefficient as the objective function to obtain a third PCE model.
[0020] The present invention provides a water pollution source tracing method based on artificial intelligence, which aims to infer the location of the pollution source, the time of pollutant emission and the emission intensity, and provide a scientific method for intelligent management and decision-making of the water environment. The present invention uses Delft3D model, big data cluster computing (Hadoop), artificial neural network, multi-objective Bayesian optimization and Monte Carlo simulation (MC) technology, and adopts agent modeling method to approximate and replace the prediction ability of complex water quality models and the ability to support simulation optimization, so as to provide a new method for solving the reverse pollution source inversion. It not only has the advantages of flexibility, speed and strong operability, but also can grasp the impact of pollutants on the water environment, so as to make timely and accurate response, and improve the accuracy of location set identification of hidden and sudden water pollution incidents.
[0021] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0023] Figure 1 This is a schematic flow chart of a water pollution source tracing method based on artificial intelligence provided by an embodiment of the present invention; Figure 2 It is a schematic flow chart of a method for simulating water pollution events under various discharge conditions provided by an embodiment of the present invention; Figure 3 is a schematic flow chart of a method for obtaining a first PCE model provided in an embodiment of the present invention; Figure 4 is a schematic flow chart of a method for obtaining a second PCE model provided in an embodiment of the present invention; Figure 5 It is a flowchart of a method for obtaining a third PCE model provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0024] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present invention. The following embodiments are used to illustrate the present invention, but cannot be used to limit the scope of the present invention.
[0025] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0026] Combine the following Figures 1 to 5 Embodiments of the present invention are described.
[0027] like Figure 1As shown, the present invention provides a water pollution source tracing method based on artificial intelligence, comprising: S1: Select the basic model, simulate water pollution events under various discharge conditions, and obtain the water quality diffusion relationship data set.
[0028] In step S1, a hydrodynamic and water quality model of the study area is first constructed, and scenario simulations are carried out under different emission conditions through experimental design to obtain a large training data set that can characterize the relationship between the pollution source emission pattern and the pollutant concentration time series response. This data set inherently embodies the mechanism model process.
[0029] The basic model selected in step S1 is a Delft3D model.
[0030] Furthermore, the selected Delft3D model is a mature, advanced and open source hydrodynamic and water quality model. As the basic model in this step, it has the characteristics of fast calculation speed and high simulation accuracy, especially for the surface mesh generation technology, it is very suitable for boundary edge processing.
[0031] like Figure 2 As shown, in step S1, the step of simulating water pollution events under various discharge conditions further includes: S11: defining a simulation area, dividing the simulation area into grids, and obtaining a plurality of simulation grids; Step S11 aims to set the regional scope for research, and select the spatial modeling area for hydrodynamic water quality simulation based on factors such as typicality, representativeness and data availability, to ensure that the research on water quality prediction and pollution source inversion in this area is representative. The selected research area must have the characteristics of a concentration of some industrial enterprises, a high risk of sudden pollution accidents, and an urgent need for pollution source inversion.
[0032] Specifically, in step S11, the RGFGRID tool software is first used to perform two-dimensional orthogonal quadrilateral grid division on the study area, the QUICKIN tool software is used to interpolate the underground of the study area, and then the QUICKPLOT tool software is used to visualize the simulation results and perform other post-processing operations. S12: setting simulation parameters and simulation variables, sampling the simulation parameters and selecting grids in the downstream area of the simulation area as observation point locations to obtain a variety of emission conditions; In addition, in order to highlight the feasibility of the method, the boundary conditions and model setting conditions are generalized. In the early stage of implementation, the actual pollution emissions along the line are not considered. Instead, the simulation test method is used to simulate the migration and diffusion process of conservative substances instead of the water quality process.
[0033] The simulation parameters in step S12 include simulation duration, calculation time step, gravity acceleration, and water body density, wherein the gravity acceleration and the water body density are the default values of the basic model; The simulation variables in step S12 include the discharge start time, discharge end time, and discharge amount of the water pollution event.
[0034] When selecting simulation parameters and variables, the urgency of water pollution incidents was taken into consideration. At the same time, in order to conform to the actual situation, the simulation duration was set to 20 days, the calculation time step was generally set to 5 minutes, and other parameters, such as gravity acceleration and water density, adopted the model default values.
[0035] Then, the grids covering the 1 / 4 length of the upstream area of the entire study area were set as optional locations for emission points, together with different emission start times, end times, and different emission amounts as parameter variables, and the Latin hypercube sampling method was used to sample the above four parameters. Finally, the grids at three equally spaced center locations within the 1 / 4 length of the downstream area of the study area were selected as observation point locations.
[0036] In order to simplify the complex water quality change process into a conservative substance migration and diffusion process, after step S12, different emission scenarios are established using the above parameter sampling.
[0037] S13: Based on the distributed computing framework integrating the basic model, a variety of discharge conditions are substituted to simulate water pollution events and obtain a water quality diffusion relationship data set.
[0038] Wherein, step S13 further comprises: S131: Select Hadoop MapReduce as the distributed computing framework, integrate the basic model, and build a simulated cluster computing system; In step S131, firstly, based on the Hadoop MapReduce distributed computing framework, a water environment simulation cluster computing system is built by integrating the Delft3D model computing engine. The hardware configuration performance and computing node roles are shown in Table 1.
[0039] Table 1 Software and hardware environment of water environment simulation cluster computing system Machine Name Hardware Configuration Roles in the cluster Master HP ProLiant DL580 g7, 32 cores, 2.67 GHZ, Memory: 64 GB, Hard Drive Capacity: 900 GB NameNode, JobTracker Master2 HP ProLiant DL580 g7, 32 cores, 2.67 GHZ, Memory: 64GB, Hard Drive Capacity: 900 GB Secondary NameNode s201~s205 ThinkCentre M8411t, 8 cores, 3.4 GHZ, RAM: 32 GB, HDD capacity: 2 TB DataNode, TaskTracker s206~s212 HP ProLiant DL580 g7, 32 cores, 2.67 GHZ, Memory: 64GB, Hard Drive Capacity: 2 TB DataNode, TaskTracker S132: using the simulation cluster computing system, respectively simulating water pollution events under various discharge conditions to obtain simulation results; S133: Obtaining original binary data of the simulation result, parsing the original binary data and storing it in a column-based database; S134: extracting the record set of the column database through Spark memory grid technology to obtain a water quality diffusion relationship data set.
[0040] After step S131, a cluster computing system is used to simulate and calculate the emission scenario represented by the aforementioned Latin hypercube sampling. Then, the Hadoop HDFS file system is used to store the original binary data of the simulation results, and the HBase column database is used to store the parsed records of the simulation results. Finally, the Spark memory grid technology is used to quickly extract the record set in HBase to generate a "four-parameter" input-"three observation points" concentration time series output data set. The four parameters are the simulation parameters in step S12, and the three observation points are the simulation variables in step S12. The generation of this data set is derived from the mechanism model, and therefore inherently characterizes the mechanism process of the pollutant emission input-concentration output response in the study area.
[0041] The control equation of the basic model in step S1 is expressed as:
[0042]
[0043]
[0044] in, is the water level at the observation point, for Directional water flow, for Directional water flow, is the flow rate at the observation point, for Flow rate in direction, for Flow rate in direction, is the density of observation points, is the Coriolis force coefficient, Bottom of water Shear stress in the direction Bottom of water Shear stress in the direction is the horizontal viscosity coefficient, The water pressure.
[0045] The above control equations are the two-dimensional Navier-Stokes equations used by the Delft3D two-dimensional hydrodynamic model.
[0046] The diffusion equation of the basic model in step S1 is expressed as:
[0047] in, is the concentration of pollutants, for The diffusion coefficient in the direction, for The diffusion coefficient in the direction, is the source-sink term, is the reaction term.
[0048] Furthermore, the grid division of the two-dimensional hydrodynamic and water quality model adopts orthogonal curve grids, and the numerical solution of multidimensional problems based on grid division has higher requirements on stability. In order to obtain a format with good stability, the implicit format has more advantages than the explicit format.
[0049] In order to reduce the bandwidth of the coefficient matrix of the implicit format for solving the two-dimensional equation group, the present invention adopts the alternating direction implicit (ADI) method to discretely solve the above-mentioned control equation group. Except for the vertical viscosity term and the convection term which use the central difference method, the rest use the ADI method, i.e. the alternating direction implicit difference method.
[0050] S2: Establishing a first PCE model based on the water quality diffusion relationship data set.
[0051] Furthermore, the PCE on which the first PCE model is based is a chaotic polynomial expansion algorithm, which is a statistical method that uses normally distributed random inputs to describe the uncertainty of a system.
[0052] like Figure 3 As shown, step S2 further includes: S21: fitting the data in the water quality diffusion relationship data set through the configuration point method in the proxy modeling base of the PCE algorithm to obtain a fitting data set; In step S21, firstly, based on the "four parameters" input - "three observation points" concentration time series output data set obtained in the above step S1, the configuration point method in the PCE agent modeling technology is used to fit the above data set to establish a PCE model. The reason for selecting the configuration point method is that this method has no special requirements for the probability distribution of parameters, and it is widely used because of its strong applicability.
[0053] S22: Input the fitting data set into a preset model, and perform model training through the chaospy software library based on the Python programming language to obtain a first PCE model.
[0054] Prepare a data set for training the PCE model, then clean the data, check for missing values, outliers, etc. in the data, and perform corresponding processing. Subsequently, normalize or standardize the data as needed to ensure that all input variables have similar scales; subsequently, you need to define random variables corresponding to your data in chaospy, including uniform distribution, normal distribution, etc., and then select an appropriate polynomial basis according to the requirements of the problem and the characteristics of the data; subsequently, use the functions in the chaospy library to associate the constructed data set with the random variables and the polynomial basis, and then use the prepared data and the defined PCE model to train by calling the corresponding functions or methods; during or after the training process, you can obtain the above-mentioned first PCE model.
[0055] S3: Optimize the approximation capability of the first PCE model according to the basic model to obtain a second PCE model.
[0056] like Figure 4 As shown, step S3 further includes: S31: Perform MC sampling on the simulation parameters to obtain sampling parameters; S32: inputting the sampling parameters into the basic model for simulation to obtain a sampling data set; S33: inputting the sampling parameters into the first PCE model to obtain a first PCE data set; S34: Perform multi-objective optimization on the first PCE model according to the sampling data set and the first PCE data set to obtain a second PCE model.
[0057] The specific process in step S3 is to perform MC sampling on the four parameters of the Delft3D model, and obtain the MC "four parameters" input-"three observation points" concentration time series output data set based on the scale calculation of the cluster computing system. Then, based on the constructed PCE model, the above MC sampling parameters are input to obtain the PCE "four parameters" input-"three observation points" concentration time series output data set. Finally, based on the above two sets of time series sets, the accuracy, efficiency and fidelity indicators are used to evaluate the approximation ability of the PCE model.
[0058] Wherein, when performing multi-objective optimization on the first PCE model, the selected objectives include: An accuracy index, wherein the accuracy index is a Nash efficiency coefficient; High efficiency index, the high efficiency index is the time consumed by simulation calculation; Fidelity index, wherein the fidelity index is the Sobol parameter sensitivity index.
[0059] Furthermore, the above-mentioned Nash efficiency coefficient is an indicator used to quantify the prediction accuracy of simulation models (such as hydrological models). It compares the degree of agreement between the simulation results and the observed data to evaluate the accuracy of the model; the simulation calculation time refers to the total time required to complete a simulation calculation. It is an important indicator to measure the calculation efficiency of the model, reflecting the operation speed and efficiency of the model under given computing resources (such as processor speed, memory size, etc.); the Sobol parameter sensitivity index is a global sensitivity analysis tool used to quantify the contribution of input variables to the variability of model outputs. By analyzing the influence of input variables on model outputs, the fidelity of the model (that is, the degree to which the model reflects real-world phenomena) is evaluated.
[0060] S4: Invert the water quality diffusion relationship data set through the second PCE model, optimize the second PCE model according to the inversion result, and obtain a third PCE model.
[0061] like Figure 5 As shown, step S4 further includes: S41: randomly selecting an inversion data set from the sampling data set; S42: using the output data set in the inversion data set as observation values, and inputting the observation values into the second PCE model to obtain an inversion result; S43: taking the input data set in the inversion data set as the true value, comparing the inversion result with the true value, and optimizing the second PCE model with the Nash efficiency coefficient as the objective function to obtain a third PCE model.
[0062] In step S4, similar to step S31 in step S3, firstly, MC sampling is performed on the four parameters of the Delft3D model, and the MC "four parameters" input-"three observation points" concentration time series output data set is obtained based on the scale calculation of the cluster computing system; secondly, one of the data sets is randomly selected, and its "three observation points" concentration time series data set is used as the observation value, and its "four parameters" input data set is used as the "true value". The Nash efficiency coefficient is used as the objective function, and the PCE model and the multi-objective Bayesian optimization algorithm are combined to invert the "four parameters" input values such as emission point, start time, end time, and emission amount; then, the multi-objective optimization algorithm is repeated multiple times to obtain the probability distribution of the four parameters, and by comparing with the "true value", the ability of the PCE model to support simulation optimization is verified and the model is optimized according to the results.
[0063] S5: Use the third PCE model as a source tracing model to trace the water pollution event to be traced.
[0064] After steps S1 to S4, a third PCE model can be obtained that has been optimized multiple times and has been verified for reliability through pollution source inversion, that is, the corresponding water pollution incident source tracing work can be carried out to obtain data such as the emission start time, emission end time, emission amount, and emission source location of the water pollution incident.
[0065] The present invention provides a water pollution source tracing method based on artificial intelligence, which transforms the traditional model-driven perspective into a data-driven perspective, establishes a complex mechanism model, can greatly reduce the model calculation amount while maintaining the accuracy of the mechanism model, and can effectively support the uncertainty analysis and simulation optimization of the complex model. The regulation and optimization based on the complex numerical model of the present invention has always been a common problem faced by multiple disciplines and an urgent need in practical applications. The present invention integrates numerical simulation (Delft3D), big data cluster computing (Hadoop), artificial intelligence (artificial neural network, probabilistic agent model), multi-objective optimization (multi-objective Bayesian optimization) and Monte Carlo simulation (MC) technology, gives full play to the advantages of multidisciplinary cross-disciplinary, and adopts an agent modeling method to approximate and replace the prediction ability of complex water quality models and the ability to support simulation optimization, so as to provide a new method for solving the reverse pollution source inversion. Under the limited computing resource budget and the limited number of model iteration calculations, it can perform complex pollution source inversion, can replace the existing complex water quality model, improve the accuracy and efficiency of pollution inversion, and reduce the computing cost.
[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A water pollution source tracing method based on artificial intelligence, characterized in that: include: S1: Select the basic model, simulate water pollution events under various discharge conditions, and obtain the water quality diffusion relationship data set; S2: establishing a first PCE model based on the water quality diffusion relationship data set; S3: Optimizing the approximation capability of the first PCE model according to the basic model to obtain a second PCE model; S4: inverting the water quality diffusion relationship data set by using the second PCE model, optimizing the second PCE model according to the inversion result, and obtaining a third PCE model; S5: Use the third PCE model as a source tracing model to trace the water pollution event to be traced.
2. The water pollution source tracing method based on artificial intelligence according to claim 1 is characterized in that: The base model selected in step S1 is the Delft3D model.
3. The water pollution source tracing method based on artificial intelligence according to claim 1 is characterized in that: In step S1, the step of simulating water pollution events under various discharge conditions further includes: S11: defining a simulation area, dividing the simulation area into grids, and obtaining a plurality of simulation grids; S12: setting simulation parameters and simulation variables, sampling the simulation parameters and selecting grids in the downstream area of the simulation area as observation point locations to obtain a variety of emission conditions; S13: Based on the distributed computing framework integrating the basic model, a variety of discharge conditions are substituted to simulate water pollution events and obtain a water quality diffusion relationship data set.
4. The water pollution source tracing method based on artificial intelligence according to claim 3 is characterized in that: The simulation parameters in step S12 include simulation duration, calculation time step, gravity acceleration, and water body density, wherein the gravity acceleration and the water body density are default values of the basic model; The simulation variables in step S12 include the discharge start time, discharge end time, and discharge amount of the water pollution event.
5. The water pollution source tracing method based on artificial intelligence according to claim 3 is characterized in that: Step S13 further comprises: S131: Select Hadoop MapReduce as the distributed computing framework, integrate the basic model, and build a simulated cluster computing system; S132: using the simulation cluster computing system, respectively simulating water pollution events under various discharge conditions to obtain simulation results; S133: Obtaining original binary data of the simulation result, parsing the original binary data and storing it in a column-based database; S134: extracting the record set of the column database through Spark memory grid technology to obtain a water quality diffusion relationship data set.
6. The water pollution source tracing method based on artificial intelligence according to claim 1 is characterized in that: The control equation of the basic model in step S1 is expressed as: in, is the water level at the observation point, for Directional water flow, for Directional water flow, is the flow rate at the observation point, for Flow rate in direction, for Flow rate in direction, is the density of observation points, is the Coriolis force coefficient, Bottom of water Shear stress in the direction Bottom of water Shear stress in the direction is the horizontal viscosity coefficient, is the water pressure; The diffusion equation of the basic model in step S1 is expressed as: in, is the concentration of pollutants, for The diffusion coefficient in the direction, for The diffusion coefficient in the direction, is the source-sink term, is the reaction term.
7. The method for tracing the source of water pollution based on artificial intelligence according to claim 1 is characterized in that: Step S2 further comprises: S21: fitting the data in the water quality diffusion relationship data set through the configuration point method in the proxy modeling base of the PCE algorithm to obtain a fitting data set; S22: Input the fitting data set into a preset model, and perform model training through the chaospy software library based on the Python programming language to obtain a first PCE model.
8. The method for tracing the source of water pollution based on artificial intelligence according to claim 3 is characterized in that: Step S3 further comprises: S31: Perform MC sampling on the simulation parameters to obtain sampling parameters; S32: inputting the sampling parameters into the basic model for simulation to obtain a sampling data set; S33: inputting the sampling parameters into the first PCE model to obtain a first PCE data set; S34: Perform multi-objective optimization on the first PCE model according to the sampling data set and the first PCE data set to obtain a second PCE model.
9. The method for tracing the source of water pollution based on artificial intelligence according to claim 8 is characterized in that: When performing multi-objective optimization on the first PCE model, the selected objectives include: An accuracy index, wherein the accuracy index is a Nash efficiency coefficient; High efficiency index, the high efficiency index is the time consumed by simulation calculation; Fidelity index, wherein the fidelity index is the Sobol parameter sensitivity index.
10. The method for tracing the source of water pollution based on artificial intelligence according to claim 8 is characterized in that: Step S4 further comprises: S41: randomly selecting an inversion data set from the sampling data set; S42: using the output data set in the inversion data set as observation values, and inputting the observation values into the second PCE model to obtain an inversion result; S43: taking the input data set in the inversion data set as the true value, comparing the inversion result with the true value, and optimizing the second PCE model with the Nash efficiency coefficient as the objective function to obtain a third PCE model.
Citation Information
Patent Citations
Tracing method for sudden river point source pollution
CN105956664A
Emergency plan generation method involving water pollution source tracing
CN116011711A
Optimization-based spatial-temporal scale multi-level water body toxic pollutant tracing method
CN116384147A
Data-driven rapid traceability method for air pollutants in small-scale regionals
US20230194755A1
Tracing method and apparatus for gas fluid pollutant emission source
WO2024108812A1
Cited By
Method and system for tracing PPCPs in surface water based on discharge source characteristic indexes
CN120142429A
