Method, device, equipment and medium for predicting water pollution transport

Abstract

The present disclosure relates to a water pollutant transport prediction method, device, equipment and medium. The water pollutant transport prediction method comprises: obtaining initial pollutant concentration data and initial background flow field data at a first target time; performing model parameter prediction based on the initial pollutant concentration data and the initial background flow field data through a model parameter prediction model to obtain an optimal value of the model parameter; and performing transport prediction based on the initial pollutant concentration data, the initial background flow field data and the optimal value of the model parameter according to a water pollutant transport prediction model to obtain a target pollutant concentration result at a second target time. According to the embodiment of the present disclosure, the prediction effect and the prediction accuracy of the water pollutant transport prediction model can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of environmental simulation and prediction, and in particular to a method, apparatus, equipment and medium for predicting the transport of pollutants on water. Background Technology

[0002] Water pollution accidents are common sudden incidents on water. Once they occur, they can cause huge casualties, seriously damage the ecological environment, and bring incalculable economic losses to the country. Simulation and prediction of the water pollutant transport process is an important means of emergency response to such accidents.

[0003] For simulating marine pollutant incidents, mathematical models are generally used to characterize the convection and diffusion processes, while numerical simulations are employed to predict their transport in water bodies. Examples include a three-dimensional simulation method and apparatus for marine oil spills (CN112016226 B), and a numerical simulation method and system for underwater oil spill movement (CN 108268751 A). However, the ability to simulate and predict marine pollutant transport processes depends on the accuracy of solving the convection and diffusion equations. Therefore, effectively solving the governing equations of marine pollutant diffusion models and improving the simulation accuracy of marine pollutant diffusion models are the primary problems that need to be addressed. Summary of the Invention

[0004] To address the aforementioned technical problems, this disclosure provides a method, apparatus, equipment, and medium for predicting the transport of pollutants on water.

[0005] In a first aspect, this disclosure provides a method for predicting the transport of pollutants on waterways, including:

[0006] Acquire the initial pollutant concentration data and initial background flow field data at the first target time;

[0007] The optimal values ​​of the model parameters are obtained by predicting the model parameters based on the initial pollutant concentration data and the initial background flow field data.

[0008] Based on the waterborne pollutant transport prediction model, transport prediction is performed using initial pollutant concentration data, initial background flow field data, and optimal model parameters to obtain the target pollutant concentration at the second target time.

[0009] Secondly, this disclosure provides a device for predicting the transport of pollutants on water, comprising:

[0010] The data acquisition module is used to acquire the initial pollutant concentration data and initial background flow field data at the first target time.

[0011] The first processing module is used to predict the model through model parameters, based on the initial pollutant concentration data and the initial background flow field data, to obtain the optimal values ​​of the model parameters.

[0012] The second processing module is used to predict the transport of pollutants based on the initial pollutant concentration data, initial background flow field data and optimal values ​​of model parameters according to the water pollutant transport prediction model, and obtain the target pollutant concentration result at the second target time.

[0013] Thirdly, this disclosure provides a device for predicting the transport of pollutants on water, comprising:

[0014] processor;

[0015] Memory, used to store executable instructions;

[0016] The processor is used to read executable instructions from memory and execute the executable instructions to implement the first aspect of the method for predicting the transport of waterborne pollutants.

[0017] Fourthly, this disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the waterborne pollutant transport prediction method of the first aspect.

[0018] The technical solution provided in this disclosure has the following advantages compared with the prior art:

[0019] The waterborne pollutant transport prediction method, apparatus, equipment, and medium of this disclosure can acquire initial pollutant concentration data and initial background flow field data at a first target time. Then, based on the initial pollutant concentration data and initial background flow field data, the model parameters are predicted using a model parameter prediction model to obtain the optimal values ​​of the model parameters. Finally, based on the waterborne pollutant transport prediction model, the transport is predicted using the initial pollutant concentration data, initial background flow field data, and optimal values ​​of the model parameters to obtain the target pollutant concentration result at a second target time. Thus, the optimal values ​​of the model parameters can be obtained first through the model parameter prediction model, and then the target pollutant concentration result can be obtained based on the waterborne pollutant transport prediction model, thereby improving the prediction effect and accuracy of the waterborne pollutant transport prediction model. Attached Figure Description

[0020] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0021] Figure 1A flowchart illustrating a method for predicting the transport of pollutants on water, provided in an embodiment of this disclosure;

[0022] Figure 2 This is a schematic diagram of the structure of a waterborne pollutant transport prediction device provided in an embodiment of the present disclosure;

[0023] Figure 3 This is a schematic diagram of a waterborne pollutant transport prediction device provided in an embodiment of this disclosure. Detailed Implementation

[0024] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0025] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0026] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0027] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0028] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0029] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0030] Water pollution accidents are common sudden incidents on water. Once they occur, they can cause huge casualties, seriously damage the ecological environment, and bring incalculable economic losses to the country. Simulation and prediction of the water pollutant transport process is an important means of emergency response to such accidents.

[0031] For simulating marine pollutant incidents, mathematical models are generally used to characterize the convection and diffusion processes, while numerical simulations are employed to predict their transport in water bodies. Examples include a three-dimensional simulation method and apparatus for marine oil spills (CN112016226 B), a numerical simulation method for underwater oil spill movement, and an underwater oil spill behavior simulation system (CN 108268751 A). In general simulation methods, pollutants are considered as oil clumps composed of a group of oil particles of varying sizes. Particle definitions are established, and the particle velocity, migration displacement, and diffusion displacement are calculated to determine the particle's position at time t, thereby determining the pollutant's diffusion range and concentration distribution. Models simulating pollutant incidents can also use the classic three-dimensional convection and diffusion equations. Solving these models typically involves methods such as the "Lagrange particle tracking method" and the "mass-conservation improved Euler-Lagrange integral method," whose calculation methods better reflect the movement mechanism of underwater pollutants, more realistically reproducing their movement characteristics and further improving the reliability of calculating the underwater pollutant diffusion range and concentration distribution.

[0032] The "Lagrange particle tracking method" determines the outer envelope range of the pollutant distribution area and the outer envelope lines of the longitudinal and cross sections based on the spatial positions of all pollutant particles at time t, thereby obtaining data on the diffusion range of pollutants. By calculating the content of pollutants in each grid cell within the water body, the concentration distribution data of pollutants in the water body is obtained.

[0033] However, the ability to simulate and predict the transport of pollutants on water depends on the accuracy of solving the convection-diffusion equations. Therefore, how to effectively solve the governing equations of the marine pollutant diffusion model and improve the simulation accuracy of the marine pollutant diffusion model are the first problems that need to be solved.

[0034] Therefore, (1) the ability to simulate and predict the transport process of pollutants on water depends on the accuracy of solving the convection-diffusion equation. How to effectively solve the governing equations of the simulated marine pollutant diffusion model is the first problem to be solved. (2) Existing solution methods cannot well describe the internal motion of the entire flow field, and the computational load is relatively large when solving the pollutant concentration at each time step at each grid point in three-dimensional space, which requires high computing power. (3) The above methods do not provide a way to improve the simulation accuracy of the marine pollutant diffusion model, such as effectively combining the model solution with the measured data.

[0035] To address the aforementioned problems, this disclosure provides a method, apparatus, device, and medium for predicting the transport of pollutants on water. The following embodiments, in conjunction with… Figure 1 and Figure 2 The method for predicting the transport of pollutants on water provided in the embodiments of this disclosure will be described in detail.

[0036] Figure 1 A flowchart illustrating a method for predicting the transport of pollutants on water provided in an embodiment of this disclosure is shown.

[0037] In this embodiment of the disclosure, the method for predicting the transport of pollutants on water can be executed by an electronic device. This electronic device may include, but is not limited to, mobile terminals such as mobile phones, in-vehicle devices, tablets, wearable devices, and smart home devices, or cloud servers or cloud server clusters.

[0038] like Figure 1 As shown, the method for predicting the transport of pollutants on water may include the following steps.

[0039] S110. Obtain the initial pollutant concentration data and initial background flow field data at the first target time.

[0040] In this embodiment of the disclosure, the electronic device can acquire initial pollutant concentration data and initial background flow field data at a first target time.

[0041] Optionally, the first target time can be a preset time. For example, the first target time can be 10 seconds, 20 seconds, etc. before the current time, and there is no limitation here.

[0042] Optionally, the initial pollutant concentration data can be data used to represent pollutant concentrations. Specifically, the initial pollutant concentration data can be high spatiotemporal resolution observation data within the target area.

[0043] Optionally, the initial pollutant concentration data refers to the pollutant concentration data at all spatial locations within the target area at the first target time. This initial pollutant concentration data can be obtained through simulation and prediction by a waterborne pollutant transport prediction model.

[0044] Optionally, the target area can be a predefined area.

[0045] For example, the initial pollutant concentration data is pollutant concentration observation data obtained through a combination of satellite remote sensing, large-area observation, buoys, and shore-based observation stations. This takes into account the large scope of satellite remote sensing observation and the high precision of shore-based observation stations, buoys, and large-area observation stations, thus obtaining pollutant concentration observation data with a large scope and high spatiotemporal resolution.

[0046] Optionally, the initial background flow field data can be flow field data used to represent the background of the waterborne pollutant transport process. This initial background flow field data may include flow field data and water depth topography data.

[0047] Optionally, the initial background flow field data consists of flow field data at all spatial locations within the target region at the first target time. This initial background flow field data can be obtained through simulation and prediction using an ocean dynamics model.

[0048] Optionally, the flow field data refers to the data on the flow field formed by the space occupied by fluid motion. This flow field data can be obtained through simulation and prediction using ocean dynamics models. These models are not limited to Finite-Volume Coastal Ocean Models (FVCOM), Regional Ocean Model Systems (ROMS), or the Hydrostatic Approximate High-Resolution Model (MITgcm), and utilize methods such as optimal interpolation to assimilate observed data such as water level and flow velocity.

[0049] For example, the basic pattern of using the optimal interpolation method is as follows:

[0050] x a =x b +W[y o -Hx b ]

[0051] Where, x a It is the optimal analysis field for the pattern variables, x b It is the background field (i.e., the prediction field obtained initially based on historical data and the transport model), W is the optimal weight matrix, and y o H is the observation field (real-time observation data), and H is the observation operator, which is an m×n matrix, where m is the number of observations and n is the number of grid points.

[0052] Optionally, the water depth topographic data may be a fusion of the General Bathymetric Chart of the Oceans (GEBCO), the Global Bathymetric Elevation Dataset (ETOPO Global Relief Model, ETOPO), and some water depth topographic data obtained from field surveys.

[0053] Specifically, electronic devices can acquire initial pollutant concentration data and initial background flow field data at the first target moment. This means that pollutant concentration observation data can be obtained through joint observation by satellite remote sensing, large-area observation, buoys, shore-based observation stations, etc. The initial background flow field data can include flow field data and water depth topography data.

[0054] S120. Predict the model by using model parameters. Based on the initial pollutant concentration data and the initial background flow field data, predict the model parameters to obtain the optimal values ​​of the model parameters.

[0055] In this embodiment of the disclosure, the electronic device can use the model parameter prediction model to predict the model parameters of the initial pollutant concentration data and the initial background flow field data, and obtain the optimal values ​​of the model parameters.

[0056] Optionally, the model parameter prediction model can be a model used to calculate the optimal values ​​of the model parameters.

[0057] Optionally, the optimal values ​​of the model parameters can be the optimal model parameters.

[0058] Optionally, model parameter prediction can be used to calculate the optimal values ​​of the model parameters.

[0059] Specifically, electronic devices can predict model parameters based on initial pollutant concentration data and initial background flow field data. In other words, electronic devices can input initial pollutant concentration data and initial background flow field data into the model parameter prediction model, which can then predict model parameters based on these data and output the optimal values ​​of the model parameters.

[0060] S130. Based on the waterborne pollutant transport prediction model, transport prediction is performed based on the initial pollutant concentration data, the initial background flow field data, and the optimal values ​​of the model parameters to obtain the target pollutant concentration result at the second target time.

[0061] In this embodiment of the disclosure, the electronic device can use a waterborne pollutant transport prediction model to predict transport based on initial pollutant concentration data, initial background flow field data, and optimal model parameter values, and obtain the target pollutant concentration result at the second target time.

[0062] Optionally, the waterborne pollutant transport prediction model can be a model used to predict the waterborne pollutant transport process.

[0063] Alternatively, water pollutants can refer to crude oil or other pollutants that leak from the work site or storage tank into water (sea, river, etc.) during the exploration, development, refining, transportation and storage of oil due to accidents or operational errors.

[0064] Optionally, transport forecasting can be a prediction of the transport process of pollutants on water.

[0065] Optionally, the second target time can be similar to the first target time. For example, the second target time can be 10 seconds or 20 seconds after the current time, etc., without limitation here.

[0066] Optionally, the target pollutant concentration result can be the pollutant concentration result at the second target time.

[0067] Specifically, electronic devices can use a waterborne pollutant transport prediction model to predict transport based on initial pollutant concentration data, initial background flow field data, and optimal model parameter values. In other words, electronic devices can input initial pollutant concentration data, initial background flow field data, and optimal model parameter values ​​into the waterborne pollutant transport prediction model. The waterborne pollutant transport prediction model can then predict transport based on these data, thereby outputting the target pollutant concentration result at the second target time.

[0068] Therefore, in this embodiment of the present disclosure, it is possible to obtain the initial pollutant concentration data and the initial background flow field data at the first target time. Then, through the model parameter prediction model, the model parameters of the initial pollutant concentration data and the initial background flow field data are predicted to obtain the optimal values ​​of the model parameters. Finally, the waterborne pollutant transport prediction model is used to predict the transport based on the initial pollutant concentration data, the initial background flow field data and the optimal values ​​of the model parameters to obtain the target pollutant concentration result at the second target time. Thus, the optimal values ​​of the model parameters can be obtained first through the model parameter prediction model, and then the target pollutant concentration result can be obtained based on the waterborne pollutant transport prediction model, thereby improving the prediction effect and prediction accuracy of the waterborne pollutant transport prediction model.

[0069] Optionally, prior to S120, the method for predicting the transport of pollutants on water may specifically include: constructing a predictive model for the transport of pollutants on water based on the three-dimensional convection-diffusion equation and the characteristic line method according to the mass conservation law.

[0070] Alternatively, the electronic device can construct a prediction model for the transport of pollutants on water based on the three-dimensional convection-diffusion equation and the method of characteristics.

[0071] Alternatively, the three-dimensional convection-diffusion equation can be:

[0072]

[0073] In the formula, C is the pollutant concentration, t is time, x and y are horizontal coordinates, σ is the vertical coordinate, H is the total water depth including still water depth and sea surface undulation, u, v, and w are the flow velocities in the x, y, and σ directions, respectively, and K... H K V These are the horizontal and vertical diffusion coefficients, respectively. sms(C) is the source and sink term of the pollutant state variable, including processes such as pollutant removal and biodegradation that cause changes in pollutant quality.

[0074] The boundary conditions for the prediction model of waterborne pollutant transport are: constant inflow length and no gradient outflow.

[0075]

[0076] Specifically, electronic devices can construct a prediction model for the transport of pollutants on water based on the three-dimensional convection-diffusion equation and the method of characteristics.

[0077] Alternatively, the method of characteristics can be a characteristic line technique that maintains the conservation of mass.

[0078] For example, For pollutants in t n = nΔt time at The concentration, assuming at t=t n Given the pollutant concentration data, find the solution t = t n+1 Given pollutant concentration data at a given time, and defining the characteristic direction as τ, the characteristic curve... Satisfy the equation:

[0079]

[0080]

[0081] Furthermore, using Indicates characteristic lines With t n The intersection of time layers. Then the calculation yields... The expression is:

[0082]

[0083] exist Pollutant concentration data at the location It can be obtained from pollutant concentration data at nearby points, and the characteristic difference scheme is constructed as follows:

[0084]

[0085] in,

[0086]

[0087] Specifically, electronic devices can construct a prediction model for the transport of pollutants on water based on the three-dimensional convection-diffusion equation and according to the method of characteristics, i.e., the formula above.

[0088] Therefore, in this embodiment of the present disclosure, the characteristic line method that maintains mass conservation is used for numerical solution, which can both maintain mass conservation and use a large time step for solution, reduce computational cost, and improve the computational accuracy of the prediction model for waterborne pollutant transport.

[0089] Optionally, prior to S120, the method for predicting the transport of pollutants on water may further include: obtaining optimized spatiotemporal variation model parameters through an adjoint assimilation method; and training the model parameter prediction model to be trained based on the spatiotemporal variation model parameters, initial pollutant concentration data, and initial background flow field data to obtain the model parameter prediction model.

[0090] In this embodiment of the disclosure, the electronic device can obtain optimized spatiotemporal variation model parameters through the accompanying assimilation method.

[0091] Alternatively, the accompanying assimilation method can be a method used to assimilate pollutant concentration observation data.

[0092] Optionally, the spatiotemporal variation model parameters can be model parameters used to construct a model parameter prediction model.

[0093] Optionally, the electronic device optimizes the spatiotemporal variation model parameters by adding a penalty function term and a regularization term, and adopts a downscaling localization strategy to obtain the optimized spatiotemporal variation model parameters within a preset target time period.

[0094] For example, based on the Lagrange multiplier method, considering both the range of model parameter variations and smoothness, a cost function is constructed:

[0095]

[0096] Where J is the cost function, C is the simulated pollutant concentration, and Cm is the cost function. obs These are the observed pollutant concentration values, i.e., the initial pollutant concentration data; K represents the weight matrix corresponding to the observed data; S and These are the parameters optimized in the current iteration and the parameters optimized in the previous iteration, respectively. Specifically, S represents the optimized spatiotemporal variation model parameters, and γ is a coefficient. This term ensures that the spatiotemporal variation model parameter S does not change too much between the two iterations during the optimization and inversion process, thus ensuring the smoothness of the spatiotemporal variation model parameter S obtained from the optimization and inversion; μ is the penalty coefficient, g(S) is the penalty function, and the term μg(S) ensures that the spatiotemporal variation model parameter S remains within a reasonable range of variation during the optimization and inversion process; Ω is the model grid region.

[0097] The specific expression for the weight matrix K is as follows:

[0098]

[0099] The expression for the penalty function g(S) is:

[0100]

[0101] Among them, S max and Smin These are the maximum and minimum values ​​of the spatiotemporal variation model parameter S within a reasonable range of variation.

[0102] Next, using the waterborne pollutant transport prediction model as constraints, the model's governing equations are rewritten as follows:

[0103]

[0104] The Lagrange function L is defined as follows:

[0105]

[0106] Where λ is the associated variable of pollutant concentration. According to the Lagrange multiplier method, the cost function reaches its minimum when the first derivative of the defined Lagrange function with respect to pollutant concentration data, associated variables, and spatiotemporal variation model parameters is zero. The specific expression is as follows:

[0107]

[0108]

[0109]

[0110] When the first derivative of the Lagrange function with respect to pollutant concentration is zero, the adjoint model can be derived as follows:

[0111]

[0112] Numerical solutions are obtained using the method of characteristics that preserves mass conservation, and the corresponding characteristic difference scheme is as follows:

[0113]

[0114] in,

[0115]

[0116] When the first derivative of the Lagrangian function with respect to the model parameters is zero, the gradient expression of the model parameters can be obtained. The value of the gradient expression can be calculated from the pollutant concentration and its associated variables. However, when the model parameters are considered to be spatiotemporally variable, the corresponding model parameters are four-dimensional, and the number of model parameters that need to be optimized and adjusted will be very large. In order to further maintain the smoothness of the model parameters during the optimization and adjustment process, a downscaling localization strategy will be used, that is: representative points are uniformly selected on the horizontal domain of the model grid. Assuming that the influence area of ​​each representative point is a circle, the center of the circle corresponds to an independent point, there are i×j grid points GP(i,j) and ii×jj representative points IP(ii,jj) in the horizontal domain. SGP(i,j) and SIP(ii,jj) are the parameter values ​​corresponding to grid point GP(i,j) and representative point IP(ii,jj), respectively. Then the relationship between SGP(i,j) and SIP(ii,jj) is as follows:

[0117]

[0118] Where ω(ii,jj) is the influence region of the representative point IP(ii,jj); W(i,j,ii,jj) is the weight coefficient, and its specific expression is as follows:

[0119]

[0120] Where R represents the local radius of point IP(ii,jj), r i,j,ii,jj It is the distance between the representative point IP(ii,jj) and the grid point GP(i,j).

[0121] Meanwhile, the gradient of the cost function with respect to the model parameters at the representative point IP(ii,jj) can be obtained by interpolating the gradient of the cost function with respect to the parameters at the grid point GP(i,j), as shown in the following formula:

[0122]

[0123] Then, based on the optimized spatiotemporal variation model parameters at representative points along the negative gradient direction using the steepest descent algorithm, the expression for the steepest descent method is as follows:

[0124]

[0125] in, It is a vector composed of parameters S of a spatiotemporal change model that needs to be adjusted, arranged in a certain order; m is the m-th iteration step of parameter optimization; β is the adjustment step size, which is a positive value.

[0126] Furthermore, after obtaining the optimized spatiotemporal variation model parameters, the electronic device can train the model parameter prediction model to be trained based on the spatiotemporal variation model parameters, initial pollutant concentration data, and initial background flow field data, thus obtaining the model parameter prediction model.

[0127] Specifically, after obtaining the optimized spatiotemporal variation model parameters, i.e., after obtaining the spatiotemporal variation model parameters S, the electronic device can train the model parameter prediction model to be trained based on the spatiotemporal variation model parameters, initial pollutant concentration data, and initial background flow field data to obtain the model parameter prediction model.

[0128] Optionally, before training the model parameter prediction model to be trained, the waterborne pollutant transport prediction method may further include: acquiring preset model training data for a preset target time period; and constructing a model parameter prediction model to be trained based on the preset model training data.

[0129] In this embodiment of the disclosure, the electronic device can acquire preset model training data for a preset target time period.

[0130] Optionally, the preset target time period can be any time period.

[0131] Optionally, the preset model training data may include preset background flow field data and preset pollutant concentration data.

[0132] Optionally, the preset background flow field data is the flow field data at all spatial locations within the target area during the preset target time period, which is obtained by simulation and prediction using an ocean dynamics model.

[0133] Optionally, the preset pollutant concentration data are the pollutant concentration data at all spatial locations within the target area during the preset target time period, which are obtained by simulation and prediction by the waterborne pollutant transport prediction model.

[0134] Specifically, the electronic device can acquire preset model training data for a preset target time period and construct a model parameter prediction model to be trained based on the preset model training data.

[0135] Figure 2 A schematic diagram of the structure of a waterborne pollutant transport prediction device provided in an embodiment of this disclosure is shown.

[0136] like Figure 2 As shown, the waterborne pollutant transport prediction device 200 may include a data acquisition module 210, a first processing module 220, and a second processing module 230.

[0137] The data acquisition module 210 can be used to acquire the initial pollutant concentration data and the initial background flow field data at the first target time.

[0138] The first processing module 220 can be used to predict the model through model parameters, and to predict the model parameters based on the initial pollutant concentration data and the initial background flow field data, so as to obtain the optimal value of the model parameters.

[0139] The second processing module 230 can be used to predict the transport of pollutants based on the initial pollutant concentration data, the initial background flow field data and the optimal values ​​of the model parameters according to the water pollutant transport prediction model, and obtain the target pollutant concentration result at the second target time.

[0140] Therefore, in this embodiment of the present disclosure, it is possible to obtain the initial pollutant concentration data and the initial background flow field data at the first target time. Then, the model parameter prediction model is used to predict the model parameters based on the initial pollutant concentration data and the initial background flow field data to obtain the optimal values ​​of the model parameters. Finally, the waterborne pollutant transport prediction model is used to predict the transport based on the initial pollutant concentration data, the initial background flow field data, and the optimal values ​​of the model parameters to obtain the target pollutant concentration result at the second target time. Thus, the optimal values ​​of the model parameters can be obtained first through the model parameter prediction model, and then the target pollutant concentration result can be obtained according to the waterborne pollutant transport prediction model, thereby improving the prediction effect and prediction accuracy of the waterborne pollutant transport prediction model.

[0141] In some embodiments of this disclosure, the initial pollutant concentration data are high spatiotemporal resolution observation data within the target area, and the initial background flow field data include flow field data and water depth topography data.

[0142] In some embodiments of this disclosure, the initial pollutant concentration data are the pollutant concentration data at all spatial locations within the target area at the first target time, which are simulated and predicted by a waterborne pollutant transport prediction model; the initial background flow field data are the flow field data at all spatial locations within the target area at the first target time, which are simulated and predicted by an ocean dynamics model.

[0143] In some embodiments of this disclosure, the waterborne pollutant transport prediction device 200 may further include a model building module.

[0144] This model building module can be used to construct a waterborne pollutant transport prediction model based on the three-dimensional convection-diffusion equation and the characteristic line method with mass conservation before predicting the model parameters based on the initial pollutant concentration data and initial background flow field data and obtaining the optimal values ​​of the model parameters.

[0145] In some embodiments of this disclosure, the waterborne pollutant transport prediction device 200 may further include a third processing module and a fourth processing module.

[0146] This third processing module can be used to obtain optimized spatiotemporal variation model parameters through the adjoint assimilation method before predicting model parameters for initial pollutant concentration data and initial background flow field data through model parameter prediction model to obtain the optimal values ​​of model parameters.

[0147] The fourth processing module can be used to train the model parameter prediction model to be trained based on the spatiotemporal variation model parameters, initial pollutant concentration data and initial background flow field data, and obtain the model parameter prediction model.

[0148] In some embodiments of this disclosure, the third processing module can be specifically used to optimize the spatiotemporal variation model parameters by adding a penalty function term and a regularization term, and to obtain the optimized spatiotemporal variation model parameters within a preset target time period by adopting a downscaling localization strategy.

[0149] In some embodiments of this disclosure, the waterborne pollutant transport prediction device 200 may further include a fifth processing module and a sixth processing module.

[0150] The fifth processing module can be used to obtain preset model training data for a preset target time period before training the model parameter prediction model to be trained.

[0151] This sixth processing module can be used to construct a model for predicting the parameters of the model to be trained based on the preset model training data.

[0152] In some embodiments of this disclosure, the preset model training data includes preset background flow field data and preset pollutant concentration data. The preset background flow field data is the flow field data at all spatial locations within the target area during the preset target time period, which is obtained by simulation and prediction using an ocean dynamics model. The preset pollutant concentration data is the pollutant concentration data at all spatial locations within the target area during the preset target time period, which is obtained by simulation and prediction using a waterborne pollutant transport prediction model.

[0153] It should be noted that, Figure 2 The waterborne pollutant transport prediction device 200 shown can perform... Figures 1 to 2 The various steps in the method embodiment shown are implemented. Figures 1 to 2 The processes and effects in the method embodiments shown are not described in detail here.

[0154] Figure 3 A schematic diagram of the structure of a waterborne pollutant transport prediction device provided in an embodiment of this disclosure is shown.

[0155] In some embodiments of this disclosure, Figure 3The waterborne pollutant transport prediction device shown can be an electronic device. This electronic device can include, but is not limited to, mobile terminals such as mobile phones, in-vehicle devices, tablets, wearable devices, and smart home devices, or cloud servers or cloud server clusters.

[0156] like Figure 3 As shown, the waterborne pollutant transport prediction device may include a processor 301 and a memory 302 storing computer program instructions.

[0157] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0158] Memory 302 may include a large-capacity storage for information or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway device. In a particular embodiment, memory 302 is a non-volatile solid-state memory. In a particular embodiment, memory 302 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (Electrically Programmable ROM, EPROM), an electrically erasable programmable PROM (EEPROM), an electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0159] The processor 301 reads and executes computer program instructions stored in the memory 302 to perform the steps of the waterborne pollutant transport prediction method provided in the embodiments of this disclosure.

[0160] In one example, the waterborne pollutant transport prediction device may further include a transceiver 303 and a bus 304. Wherein, as... Figure 3As shown, the processor 301, memory 302 and transceiver 303 are connected via bus 304 and communicate with each other.

[0161] Bus 304 may include hardware, software, or both. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 304 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0162] This disclosure also provides a non-volatile computer-readable storage medium that can store a computer program. When the computer program is executed by a processor, the processor enables the processor to implement the waterborne pollutant transport prediction method provided in this disclosure.

[0163] The aforementioned storage medium may, for example, include a memory 302 containing computer program instructions, which can be executed by the processor 301 of the waterborne pollutant transport prediction device to complete the waterborne pollutant transport prediction method provided in this embodiment. Optionally, the storage medium may be a non-transitory computer-readable storage medium, such as a ROM, random access memory (RAM), compact disc ROM (CD-ROM), magnetic tape, floppy disk, and optical data storage device.

[0164] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the term "comprising" is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0165] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the transport of pollutants on waterways, characterized in that, include: Acquire the initial pollutant concentration data and initial background flow field data at the first target time; The model parameters are predicted by using the model parameters. Based on the initial pollutant concentration data and the initial background flow field data, the model parameters are predicted to obtain the optimal values ​​of the model parameters. Based on the waterborne pollutant transport prediction model, transport prediction is performed using the initial pollutant concentration data, the initial background flow field data, and the optimal values ​​of the model parameters to obtain the target pollutant concentration at the second target time. The method further includes: Based on the three-dimensional convection-diffusion equation, the water pollutant transport prediction model is constructed according to the characteristic line method with mass conservation. Before predicting the model parameters based on the initial pollutant concentration data and the initial background flow field data to obtain the optimal values ​​of the model parameters, the method further includes: The optimized spatiotemporal variation model parameters are obtained through the adjoint assimilation method; Based on the spatiotemporal variation model parameters, the initial pollutant concentration data, and the initial background flow field data, the model parameter prediction model to be trained is trained to obtain the model parameter prediction model; The optimized spatiotemporal variation model parameters include: Based on the Lagrange multiplier method, and taking into account both the range of variation and smoothness of the model parameters, the cost function is constructed as follows: Where J is the cost function, C is the simulated pollutant concentration, and C obs These are the observed pollutant concentration values, i.e., the initial pollutant concentration data; K represents the weight matrix corresponding to the observed data; S and These are the parameters optimized in the current iteration and the parameters optimized in the previous iteration, respectively. Specifically, S represents the optimized spatiotemporal variation model parameters, and γ is a coefficient. To ensure that the spatiotemporal variation model parameter S does not change significantly between the two iterations during the optimization and inversion process, the spatiotemporal variation model parameter S obtained from the optimization and inversion is smooth; μ is the penalty coefficient, g(S) is the penalty function, and μg(S) ensures that the spatiotemporal variation model parameter S remains within a reasonable range of variation during the optimization and inversion process; Ω is the model grid region; By adding penalty function terms and regularization terms, the spatiotemporal variation model parameters are optimized, and a downscaling localization strategy is adopted to obtain the optimized spatiotemporal variation model parameters within a preset target time period.

2. The method according to claim 1, characterized in that, The initial pollutant concentration data are the pollutant concentration data at all spatial locations within the target area at the first target time, which are obtained by simulation and prediction of the waterborne pollutant transport prediction model; the initial background flow field data are the flow field data at all spatial locations within the target area at the first target time, which are obtained by simulation and prediction of the ocean dynamics model.

3. The method according to claim 1, characterized in that, Before training the model parameter prediction model to be trained, the method further includes: Obtain the preset model training data for the preset target time period; The model parameter prediction model to be trained is constructed based on the preset model training data.

4. The method according to claim 3, characterized in that, The preset model training data includes preset background flow field data and preset pollutant concentration data. The preset background flow field data is the flow field data at all spatial locations within the target area during the preset target time period, which is obtained by simulation and prediction using an ocean dynamics model. The preset pollutant concentration data is the pollutant concentration data at all spatial locations within the target area during the preset target time period, which is obtained by simulation and prediction using a waterborne pollutant transport prediction model.

5. A device for predicting the transport of pollutants on water, characterized in that, include: The data acquisition module is used to acquire the initial pollutant concentration data and initial background flow field data at the first target time. The first processing module is used to predict the model through model parameters, and to predict the model parameters based on the initial pollutant concentration data and the initial background flow field data to obtain the optimal values ​​of the model parameters. The second processing module is used to perform transport prediction based on the initial pollutant concentration data, the initial background flow field data and the optimal value of the model parameters according to the water pollutant transport prediction model, and to obtain the target pollutant concentration result at the second target time. The device further includes: The model building module is used to construct the waterborne pollutant transport prediction model based on the three-dimensional convection-diffusion equation and the characteristic line method with mass conservation. The third processing module is used to obtain the optimized spatiotemporal variation model parameters through the adjoint assimilation method; The fourth processing module is used to train the model parameter prediction model to be trained based on the spatiotemporal variation model parameters, the initial pollutant concentration data and the initial background flow field data, so as to obtain the model parameter prediction model. The optimized spatiotemporal variation model parameters include: Based on the Lagrange multiplier method, and taking into account both the range of variation and smoothness of the model parameters, the cost function is constructed as follows: Where J is the cost function, C is the simulated pollutant concentration, and C obs These are the observed pollutant concentration values, i.e., the initial pollutant concentration data; K represents the weight matrix corresponding to the observed data; S and These are the parameters optimized in the current iteration and the parameters optimized in the previous iteration, respectively. Specifically, S represents the optimized spatiotemporal variation model parameters, and γ is a coefficient. To ensure that the spatiotemporal variation model parameter S does not change significantly between the two iterations during the optimization and inversion process, the spatiotemporal variation model parameter S obtained from the optimization and inversion is smooth; μ is the penalty coefficient, g(S) is the penalty function, and μg(S) ensures that the spatiotemporal variation model parameter S remains within a reasonable range of variation during the optimization and inversion process; Ω is the model grid region; The third processing module is specifically used to optimize the parameters of the spatiotemporal variation model by adding a penalty function term and a regularization term, and to obtain the optimized spatiotemporal variation model parameters within a preset target time period by adopting a downscaling localization strategy.

6. A device for predicting the transport of pollutants over water, characterized in that, include: processor; Memory, used to store executable instructions; The processor is configured to read the executable instructions from the memory and execute the executable instructions to implement the method for predicting the transport of waterborne pollutants as described in any one of claims 1-4.

7. A non-volatile computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, causes the processor to implement the method for predicting the transport of waterborne pollutants as described in any one of claims 1-4.

Images