Pollution source and pollutant distribution inversion method for overwater oil spill and hazardous chemical substance leakage accidents

Abstract

The invention discloses an overwater oil spill and hazardous chemical substance leakage accident pollution source and pollutant distribution inversion method. The method comprises the following steps: obtaining hyperspectral remote sensing data at an accident site; establishing a mathematical model to describe the transmission and diffusion process of the pollutants in the water body; according to a Bayesian statistical inference method, the posterior probability distribution of the pollution source position and the pollutant concentration is updated in combination with the actually measured data; the pollution source position and the pollutant concentration are optimized through an optimization algorithm, so that the prediction result of the mathematical model is matched with the actually measured data as much as possible; and the reasonability and credibility of the inversion result are analyzed and evaluated, and are compared and verified with the actual situation, so that the scientificity and accuracy of the inversion result are ensured. According to the method, a Bayesian statistical inference and optimization method is used, inversion can be carried out more effectively by using measured data, accurate estimation of the position and the initial concentration of the pollution source is realized, and the data-driven method has higher flexibility and adaptability when processing complex environmental conditions and change conditions.

Description

Technical Field

[0001] The present invention relates to the field of environmental monitoring, and in particular to a method for inverting pollution sources and pollutant distribution in water oil spill and hazardous chemical leakage accidents. Background Art

[0002] Oil spills and hazardous chemical leaks on water are one of the major environmental and safety challenges we face today. Such incidents not only pose a direct threat to aquatic ecosystems, but may also endanger nearby communities and public health. For example, chemical leaks may lead to the release of toxic gases or contaminate water sources, causing long-term impacts on the environment and human health.

[0003] Rapid response and effective treatment of oil spills and hazardous chemical leaks on water are crucial to protecting the environment and public safety. Traditional pollution source and pollutant distribution inversion methods often rely on simplified physical models or numerical simulations, such as diffusion models and flow field models. Although these methods have their application value, they have certain limitations when dealing with complex environmental conditions and changes, such as overly simplified model assumptions, insufficient uncertainty processing capabilities, and limited real-time inversion results. Summary of the invention

[0004] The purpose of the present invention is to provide a method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemical leakage accidents on water, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water, comprising the following steps:

[0007] Step S1, acquiring hyperspectral remote sensing data at the accident site, covering visible light and infrared bands, and performing geometric correction and atmospheric correction processing;

[0008] Step S2, based on water velocity, water depth and wind direction factors, a mathematical model is established to describe the transmission and diffusion process of pollutants in the water body;

[0009] Step S3, setting the prior probability distribution of the pollution source location and the pollutant concentration, and updating the posterior probability distribution of the pollution source location and the pollutant concentration in combination with the measured data according to the Bayesian statistical inference method;

[0010] Step S4, formalizing the problem of estimating the pollution source location and pollutant concentration into an inverse problem solution, optimizing the pollution source location and pollutant concentration through an optimization algorithm, so that the prediction results of the mathematical model are as consistent as possible with the measured data;

[0011] Step S5, analyzing and evaluating the rationality and credibility of the inversion results, and comparing and verifying them with the actual situation to ensure the scientificity and accuracy of the inversion results.

[0012] Furthermore, in step S2, describing the diffusion of pollutants includes using the diffusion equation Fick's law, and combining the water velocity and water depth in the hydrodynamic parameters to simulate the transmission and diffusion process of pollutants. The mathematical model also includes a mathematical description of the water flow velocity, flow direction and turbulence factors to capture the movement path and speed changes of pollutants in the water body.

[0013] Fick's law is expressed as follows:

[0014] [\frac{\partial C}{\partial t}=D\nabla^2C]

[0015] Where (C) is the concentration of the pollutant, (t) is the time, and (D) is the diffusion coefficient.

[0016] Furthermore, in step S2, the mathematical model further includes factors based on water temperature, salinity, and suspended particulate matter;

[0017] The Gaussian model is expressed as follows:

[0018] [C(x,y,t)=\frac{Q}{4\pi D t}\exp\left(-\frac{(x-x_s)^2+(y-y_s)^2}{4Dt}

[0019] \right)]

[0020] Where (Q) is the release rate, ((x_s,y_s)) is the coordinate of the pollution source location, (D) is the diffusion coefficient, and (t) is the time.

[0021] Further, in step S3, the prior probability setting is to set the probability distribution of the pollution source location and initial concentration based on the accident background and the expected pollutant type, including normal distribution or uniform distribution, and the Bayesian update is to use the measured data and Bayesian theorem to update the posterior probability distribution of the pollution source location and pollutant concentration to improve the accuracy and credibility of the inversion result;

[0022] The prior probability distribution is expressed as follows:

[0023] The pollution source location ((x_s, y_s)) identifies the prior probability distribution, including a two-dimensional normal distribution or a uniform distribution.

[0024] The pollutant concentration (C_0) represents a set prior probability distribution, including lognormal distribution or uniform distribution.

[0025] Furthermore, in step S3, the Bayesian statistical inference method further includes using variational inference or Markov chain Monte Carlo method to update the posterior probability distribution;

[0026] The updated posterior probability distribution is expressed as follows:

[0027] [P(x_s,y_s|\text{observation data})\propto P(x_s,y_s)\cdot P(\text{observation data}|x_s,

[0028] y_s)]

[0029] Among them, (P(x_s,y_s)) is the prior probability distribution, and (P(\text{observed data}|x_s,y_s)) is the likelihood function of the observed data given the location of the pollution source.

[0030] Furthermore, in step S4, the inverse problem is formalized by converting the estimation of the pollution source location and the pollutant concentration into a mathematical optimization problem to infer the optimal model parameters, and the optimization algorithm is to minimize the error between the prediction result of the mathematical model and the measured data through iterative optimization;

[0031] The optimization function for solving the inverse problem is expressed as follows:

[0032] [\min_{x_s,y_s,C_0}\sum_{i=1}^{N}\left(C_{\text{Model}}(x_i,y_i,t_i,x_s,y_s,

[0033] C_0)-C_{\text{Observation}}(x_i,y_i,t_i)\right)^2]

[0034] Among them, (C_{\text{model}}(x_i,y_i,t_i,x_s,y_s,C_0)) is the pollutant concentration predicted by the model, (C_{\text{observation}}(x_i,y_i,t_i)) is the actual observed pollutant concentration data, and ((x_i,y_i,t_i)) is the spatial coordinates and time point of the observation point.

[0035] Furthermore, in step S4, the optimization algorithm further includes a gradient descent method, a genetic algorithm or a particle swarm algorithm, which is used to optimize the location of the pollution source and the concentration of the pollutant;

[0036] The particle swarm algorithm simulates the behavior of a flock of birds or a school of fish looking for food. Each particle searches for the optimal solution in the solution space. For the pollution source inversion problem, the particle swarm algorithm is expressed as follows:

[0037] [\vec{v}{i}(t+1)=w\vec{v}{i}(t)+c_1r_1(\vec{p}{i}(t)-\vec{x}{i}(t))+c_2

[0038] r_2(\vec{p}{g}(t)-\vec{x}{i}(t))][\vec{x}{i}(t+1)=\vec{x}{i}(t)+

[0039] \vec{v}_{i}(t+1)]

[0040] Among them, (\vec{x}_{i}(t)) is the position vector (pollution source location and pollutant concentration) of particle (i) at time (t), (\vec{v}_{i}(t)) is the velocity vector of particle (i) at time (t), (\vec{p}_{i}(t)) is the best individual historical position vector of particle (i), (\vec{p}_{g}(t)) is the best global historical position vector of the population, (w) is the inertia weight, (c_1) and (c_2) are acceleration factors, and (r_1) and (r_2) are random numbers.

[0041] Furthermore, in step S5, the result evaluation is to evaluate the consistency and rationality of the pollution source location and pollutant concentration distribution map obtained by inversion with the actual data, and compare the inversion result with the actual field observation data or sampling data to verify the effectiveness and accuracy of the inversion method;

[0042] The error evaluation function is expressed as follows:

[0043] [\text{Error}=\frac{1}{N}\sum_{i=1}^{N}\left|C_{\text{Model}}(x_i,y_i,t_i,

[0044] x_s,y_s,C_0)-C_{\text{observation}}(x_i,y_i,t_i)\right|]

[0045] Where (N) is the number of observations.

[0046] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes any one of the methods for inverting the pollution source and pollutant distribution of oil spills and hazardous chemical leakage accidents on water.

[0047] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes any one of the methods for inverting pollution sources and pollutant distribution of oil spills and hazardous chemical leakage accidents on water.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The present invention uses Bayesian statistical inference and optimization methods, which can more effectively utilize measured data for inversion and achieve accurate estimation of the location and initial concentration of pollution sources. This data-driven method has greater flexibility and adaptability when dealing with complex environmental conditions and changes.

[0050] 2. The present invention uses Bayesian statistical methods, which can effectively deal with various uncertainties, such as measurement errors, model parameter uncertainties and initial condition uncertainties, and provide more comprehensive and reliable results through posterior distribution.

[0051] 3. The present invention can provide preliminary inversion results in a relatively short time through optimization methods and data-driven inversion processes, which helps to quickly respond to accidents and take effective measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 The present invention is a flow chart of a method for inverting pollution sources and pollutant distribution in water oil spill and hazardous chemical leakage accidents. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figure 1 , the present invention provides a technical solution:

[0055] A method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water, comprising the following steps:

[0056] Step S1, acquiring hyperspectral remote sensing data at the accident scene, covering visible light and infrared bands, and performing geometric correction and atmospheric correction processing;

[0057] Step S2, based on water velocity, water depth and wind direction factors, a mathematical model is established to describe the transmission and diffusion process of pollutants in the water body;

[0058] Preferably, describing the diffusion of pollutants includes using the diffusion equation Fick's law, and combining the water velocity and water depth in the hydrodynamic parameters to simulate the transmission and diffusion process of pollutants, and the mathematical model also includes a mathematical description of the water flow velocity, flow direction and turbulence factors to capture the movement path and speed change of pollutants in the water body;

[0059] Fick's law is expressed as follows:

[0060] [\frac{\partial C}{\partial t}=D\nabla^2C]

[0061] Where (C) is the concentration of the pollutant, (t) is the time, and (D) is the diffusion coefficient;

[0062] Furthermore, the mathematical model further includes factors based on water temperature, salinity, and suspended particulate matter;

[0063] The Gaussian model is expressed as follows:

[0064] [C(x,y,t)=\frac{Q}{4\pi D t}\exp\left(-\frac{(x-x_s)^2+(y-y_s)^2}{4Dt}

[0065] \right)]

[0066] Where (Q) is the release rate, ((x_s,y_s)) is the coordinate of the pollution source location, (D) is the diffusion coefficient, and (t) is the time;

[0067] In this step, it is assumed that a chemical leak occurs in the center of the lake. Pollutants are released from the point source and then spread along with the flow of lake water. According to Fick's law, the concentration change of pollutants is expressed as:

[0068] [\frac{\partial C}{\partial t}=D\nabla^2C]

[0069] In the two-dimensional case, assuming the lake is flat, Fick's law is defined as:

[0070] [\frac{\partial C}{\partial t}=D\left(\frac{\partial^2C}{\partial x^2}+

[0071] \frac{\partial^2C}{\partial y^2}\right)]

[0072] Here, (C(x,y,t)) is the concentration distribution of the pollutant, and (D) is the diffusion coefficient, which reflects the diffusion rate of the pollutant in the lake water.

[0073] Step S3, setting the prior probability distribution of the pollution source location and the pollutant concentration, and updating the posterior probability distribution of the pollution source location and the pollutant concentration in combination with the measured data according to the Bayesian statistical inference method;

[0074] Preferably, the prior probability setting is to set the probability distribution of the pollution source location and initial concentration based on the accident background and the expected pollutant type, including normal distribution or uniform distribution, and the Bayesian update is to update the posterior probability distribution of the pollution source location and pollutant concentration using measured data and Bayesian theorem to improve the accuracy and credibility of the inversion results;

[0075] The prior probability distribution is expressed as follows:

[0076] The pollution source location ((x_s, y_s)) identifies the prior probability distribution, including a two-dimensional normal distribution or a uniform distribution.

[0077] The pollutant concentration (C_0) represents the set prior probability distribution, including lognormal distribution or uniform distribution;

[0078] Furthermore, the Bayesian statistical inference method further includes using variational inference or Markov chain Monte Carlo method to update the posterior probability distribution;

[0079] The updated posterior probability distribution is expressed as follows:

[0080] [P(x_s,y_s|\text{observation data})\propto P(x_s,y_s)\cdot P(\text{observation data}|x_s,

[0081] y_s)]

[0082] Among them, (P(x_s,y_s)) is the prior probability distribution, (P(\text{observation data}|x_s,y_s)) is the likelihood function of the observation data given the location of the pollution source;

[0083] In this step, it is assumed that multiple observation stations have been arranged around the lake to monitor the concentration changes of pollutants in real time, but the specific location and initial concentration of the pollution source are unknown. The prior distribution of the location ((x_s, y_s)) and initial concentration (C_0) of the pollution source is expressed as:

[0084] [P(x_s,y_s)\sim\mathcal{N}(\mu_{\text{prior}},\Sigma_{\text{prior}})]

[0085] [P(C_0)\sim\mathcal{U}(0,C_{\text{max}})]

[0086] Here, the likelihood function of the observed data (\mathcal{D}={C_{\text{observed}}(x_i,y_i,t_i)}) for a given pollution source location ((x_s,y_s)) and initial concentration (C_0) is expressed as:

[0087] [P(\mathcal{D}|x_s,y_s,C_0)\propto\exp\left(-\frac{1}{2\sigma^2}

[0088] \sum_{i=1}^{N}\left(C_{\text{observed}}(x_i,y_i,t_i)-C_{\text{model}}(x_i,y_i,

[0089] t_i,x_s,y_s,C_0)\right)^2\right)]

[0090] Then, using Bayes’ theorem, update the posterior probability distribution:

[0091] [P(x_s,y_s|\mathcal{D})\propto P(x_s,y_s)\cdot P(C_0)\cdot P(\mathcal{D}|

[0092] x_s,y_s,C_0)].

[0093] Step S4, formalizing the problem of estimating the pollution source location and pollutant concentration into an inverse problem solution, optimizing the pollution source location and pollutant concentration through an optimization algorithm, so that the prediction results of the mathematical model are as consistent as possible with the measured data;

[0094] Preferably, the inverse problem is formalized by converting the estimation of the pollution source location and the pollutant concentration into a mathematical optimization problem to infer the best model parameters, and the optimization algorithm is to minimize the error between the prediction results of the mathematical model and the measured data through iterative optimization;

[0095] The optimization function for solving the inverse problem is expressed as follows:

[0096] [\min_{x_s,y_s,C_0}\sum_{i=1}^{N}\left(C_{\text{Model}}(x_i,y_i,t_i,x_s,y_s,

[0097] C_0)-C_{\text{Observation}}(x_i,y_i,t_i)\right)^2]

[0098] Among them, (C_{\text{model}}(x_i,y_i,t_i,x_s,y_s,C_0)) is the pollutant concentration predicted by the model, (C_{\text{observation}}(x_i,y_i,t_i)) is the actual observed pollutant concentration data, and ((x_i,y_i,t_i)) is the spatial coordinates and time point of the observation point;

[0099] The optimization algorithm further includes a gradient descent method, a genetic algorithm, or a particle swarm algorithm for optimizing the location of pollution sources and the concentration of pollutants;

[0100] The particle swarm algorithm simulates the behavior of a flock of birds or a school of fish looking for food. Each particle searches for the optimal solution in the solution space. For the pollution source inversion problem, the particle swarm algorithm is expressed as follows:

[0101] [\vec{v}{i}(t+1)=w\vec{v}{i}(t)+c_1r_1(\vec{p}{i}(t)-\vec{x}{i}(t))+c_2

[0102] r_2(\vec{p}{g}(t)-\vec{x}{i}(t))][\vec{x}{i}(t+1)=\vec{x}{i}(t)+

[0103] \vec{v}_{i}(t+1)]

[0104] Where, (\vec{x}_{i}(t)) is the position vector (pollution source location and pollutant concentration) of particle (i) at time (t), (\vec{v}_{i}(t)) is the velocity vector of particle (i) at time (t), (\vec{p}_{i}(t)) is the best individual historical position vector of particle (i), (\vec{p}_{g}(t)) is the best global historical position vector of the population, (w) is the inertia weight, (c_1) and (c_2) are acceleration factors, (r_1) and (r_2) are random numbers;

[0105] In this step, the exact location ((x_s,y_s)) and initial concentration (C_0) of the pollution source are inverted by minimizing the error between the model prediction value and the measured data.

[0106] The optimization objective function is expressed as follows:

[0107] [\min_{x_s,y_s,C_0}\sum_{i=1}^{N}\left(C_{\text{model}}(x_i,y_i,t_i,x_s,

[0108] y_s,C_0)-C_{\text{observed}}(x_i,y_i,t_i)\right)^2]

[0109] Here, (C_{\text{model}}(x_i,y_i,t_i,x_s,y_s,C_0)) is the pollutant concentration predicted based on the model, (C_{\text{observed}}(x_i,y_i,t_i)) is the actual observed pollutant concentration data, and ((x_i,y_i,t_i)) is the coordinates and time of the observation point.

[0110] Step S5, analyzing and evaluating the rationality and credibility of the inversion results, and comparing and verifying with the actual situation to ensure the scientificity and accuracy of the inversion results.

[0111] Preferably, the result evaluation is to evaluate the consistency and rationality of the pollution source location and pollutant concentration distribution map obtained by inversion with the actual data, and compare the inversion results with the actual field observation data or sampling data to verify the effectiveness and accuracy of the inversion method;

[0112] The error evaluation function is expressed as follows:

[0113] [\text{Error}=\frac{1}{N}\sum_{i=1}^{N}\left|C_{\text{Model}}(x_i,y_i,t_i,x_s,

[0114] y_s,C_0)-C_{\text{observation}}(x_i,y_i,t_i)\right|]

[0115] Where (N) is the number of observations;

[0116] In the previous step, the location and initial concentration of the pollution source were obtained through Bayesian inference and optimization methods. Now it is necessary to evaluate the accuracy of the inversion results.

[0117] The error evaluation function is expressed as follows:

[0118] [\text{Error}=\frac{1}{N}\sum_{i=1}^{N}\left|C_{\text{model}}(x_i,y_i,t_i,x_s,y_s,C_0)-C_{\text{observed}}(x_i,y_i,t_i)\right|]

[0119] The above error function is used to calculate the mean absolute error between the model prediction value and the actual observation data to evaluate the rationality and credibility of the inversion results.

[0120] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes any one of the methods for inverting the pollution source and pollutant distribution of oil spills and hazardous chemical leakage accidents on water.

[0121] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes any one of the methods for inverting pollution sources and pollutant distribution of oil spills and hazardous chemical leakage accidents on water.

[0122] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0123] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water, characterized in that: The following steps are involved: Step S1, acquiring hyperspectral remote sensing data at the accident scene, covering visible light and infrared bands, and performing geometric correction and atmospheric correction processing; Step S2, based on water velocity, water depth and wind direction factors, a mathematical model is established to describe the transmission and diffusion process of pollutants in the water body; Step S3, setting the prior probability distribution of the pollution source location and the pollutant concentration, and updating the posterior probability distribution of the pollution source location and the pollutant concentration in combination with the measured data according to the Bayesian statistical inference method; Step S4, formalizing the problem of estimating the pollution source location and pollutant concentration into an inverse problem solution, optimizing the pollution source location and pollutant concentration through an optimization algorithm, so that the prediction results of the mathematical model are as consistent as possible with the measured data; Step S5, analyzing and evaluating the rationality and credibility of the inversion results, and comparing and verifying them with the actual situation to ensure the scientificity and accuracy of the inversion results.

2. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S2, describing the diffusion of pollutants includes using the diffusion equation Fick's law, and combining the water velocity and water depth in the hydrodynamic parameters to simulate the transmission and diffusion process of pollutants. The mathematical model also includes a mathematical description of the water flow velocity, flow direction and turbulence factors to capture the movement path and speed changes of pollutants in the water body; Fick's law is expressed as follows: [\frac{\partial C}{\partial t}=D\nabla^2C] Where (C) is the concentration of the pollutant, (t) is the time, and (D) is the diffusion coefficient.

3. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S2, the mathematical model further includes factors based on water temperature, salinity, and suspended particulate matter; The Gaussian model is expressed as follows: [C(x,y,t)=\frac{Q}{4\pi D t}\exp\left(-\frac{(x-x_s)^2+(y-y_s)^2}{4D t}\right)] Where (Q) is the release rate, ((x_s,y_s)) is the coordinate of the pollution source location, (D) is the diffusion coefficient, and (t) is the time.

4. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S3, the prior probability setting is to set the probability distribution of the pollution source location and initial concentration based on the accident background and the expected pollutant type, including normal distribution or uniform distribution, and the Bayesian update is to use the measured data and Bayesian theorem to update the posterior probability distribution of the pollution source location and pollutant concentration to improve the accuracy and credibility of the inversion result; The prior probability distribution is expressed as follows: The pollution source location ((x_s, y_s)) identifies the prior probability distribution, including two-dimensional normal distribution or uniform distribution; The pollutant concentration (C_0) represents a set prior probability distribution, including log-normal distribution or uniform distribution.

5. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S3, the Bayesian statistical inference method further includes using variational inference or Markov chain Monte Carlo method to update the posterior probability distribution; The updated posterior probability distribution is expressed as follows: [P(x_s,y_s|\text{observation data})\propto P(x_s,y_s)\cdot P(\text{observation data}|x_s,y_s)] Among them, (P(x_s,y_s)) is the prior probability distribution, and (P(\text{observed data}|x_s,y_s)) is the likelihood function of the observed data given the location of the pollution source.

6. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S4, the inverse problem is formalized by converting the estimation of the pollution source location and the pollutant concentration into a mathematical optimization problem to infer the optimal model parameters, and the optimization algorithm is to minimize the error between the prediction result of the mathematical model and the measured data through iterative optimization; The optimization function for solving the inverse problem is expressed as follows: [\min_{x_s,y_s,C_0}\sum_{i=1}^{N}\left(C_{\text{Model}}(x_i,y_i,t_i,x_s,y_s,C_0)-C_{\text{Observation}}(x_i,y_i,t_i)\right)^2] Among them, (C_{\text{model}}(x_i,y_i,t_i,x_s,y_s,C_0)) is the pollutant concentration predicted by the model, (C_{\text{observation}}(x_i,y_i,t_i)) is the actual observed pollutant concentration data, and ((x_i,y_i,t_i)) is the spatial coordinates and time point of the observation point.

7. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S4, the optimization algorithm further includes a gradient descent method, a genetic algorithm or a particle swarm algorithm for optimizing the location of pollution sources and the concentration of pollutants; The particle swarm algorithm simulates the behavior of a flock of birds or a school of fish looking for food. Each particle searches for the optimal solution in the solution space. For the pollution source inversion problem, the particle swarm algorithm is expressed as follows: [\vec{v}{i}(t+1)=w\vec{v}{i}(t)+c_1r_1(\vec{p}{i}(t)-\vec{x}{i}(t))+c_2r_2(\ vec{p}{g}(t)-\vec{x}{i}(t))][\vec{x}{i}(t+1)=\vec{x}{i}(t)+\vec{v}_{i}(t+1)] Among them, (\vec{x}_{i}(t)) is the position vector (pollution source location and pollutant concentration) of particle (i) at time (t), (\vec{v}_{i}(t)) is the velocity vector of particle (i) at time (t), (\vec{p}_{i}(t)) is the best individual historical position vector of particle (i), (\vec{p}_{g}(t)) is the best global historical position vector of the population, (w) is the inertia weight, (c_1) and (c_2) are acceleration factors, and (r_1) and (r_2) are random numbers.

8. The method for inverting the pollution sources and pollutant distribution of oil spills and hazardous chemicals leakage accidents on water according to claim 1 is characterized by: In step S5, the result evaluation is to evaluate the consistency and rationality of the pollution source location and pollutant concentration distribution map obtained by inversion with the actual data, and compare the inversion result with the actual field observation data or sampling data to verify the effectiveness and accuracy of the inversion method; The error evaluation function is expressed as follows: [\text{Error}=\frac{1}{N}\sum_{i=1}^{N}\left|C_{\text{Model}}(x_i,y_i,t_i,x_s,y_s,C_0)-C_{\text{Observation}}(x_i,y_i,t_i)\right|] Where (N) is the number of observations.

9. A computer device, characterized in that: It includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the method for inverting the pollution source and pollutant distribution of oil spills and hazardous chemical leakage accidents on water as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes the method for inverting the pollution source and pollutant distribution of an oil spill or hazardous chemical leakage accident on water as described in any one of claims 1 to 8.

Images