Water pollution tracing and emergency decision linkage method and system

Abstract

The application discloses a water pollution source tracing and emergency decision linkage method and system, relates to the technical field of water pollution source tracing and emergency decision linkage, and comprises data acquisition and enhancement, dynamic fluorescence feature registration, pollutant feature, water power fluorescence coupling source tracing and emergency decision linkage. Through analysis of surface enhanced fluorescence (SEF) and three-dimensional fluorescence spectrum (EEM), the water pollution source tracing and emergency decision linkage are analyzed, the limitation problem in the current water pollution source tracing and emergency decision linkage analysis process is solved, the spatial correlation mapping of dynamic fluorescence feature registration is constructed, the source tracing weight matrix is generated in combination with a pollutant diffusion water power model, and finally, the pollution source positioning and emergency scheme optimization are realized through a multi-factor decision tree model, so that the problems of insufficient sensitivity, high overlap rate of pollutant fingerprint features, disconnection between source tracing and emergency response and the like in traditional fluorescence detection technology are solved.

Description

Technical Field

[0001] This application relates to the field of water pollution source tracing and emergency decision-making linkage technology, specifically to a method and system for water pollution source tracing and emergency decision-making linkage. Background Technology

[0002] With breakthroughs in spectral enhancement technology, water pollution source tracing technology has made a qualitative leap. However, traditional water pollution source tracing technologies suffer from limitations such as single-dimensional monitoring, isolated hydrodynamic models, and a disconnect between source tracing and decision-making. Therefore, the linkage between water pollution source tracing and emergency decision-making still requires further research.

[0003] Existing technologies, such as the invention application patent with announcement number CN113869798A, disclose a method for rapid source tracing of water pollution by automatically screening upstream risk sources. This method includes: constructing a hydrological spatial scenario database based on DEM data; constructing a risk source discharge characteristic database by analyzing the discharge characteristics of pollution risk sources; automatically screening upstream risk sources of monitoring points based on the spatial relationships of hydrological elements using spatial query technology; comparing the monitoring results of the monitoring points with the discharge characteristics of suspected risk sources; and calculating the comprehensive discharge contribution rate of each suspected risk source. The greater the comprehensive discharge contribution rate of the suspected risk source, the greater the potential impact of that risk source on the water quality pollution at the monitoring point.

[0004] Regarding the above-mentioned solutions, the inventors of this application have found that the above-mentioned technologies have at least the following technical problems: 1. Currently, there is a lack of dynamic registration between the response characteristics of surface-enhanced fluorescence nanomaterial substrates and 3D-EEM, a lack of elimination of substrate interference by establishing a wavelength shift compensation function, an inability to improve the spectral resolution of pollutant characteristics, and an inability to reduce cross-interference errors; there is a lack of quantification of reconstruction errors by the sum of squares of matrix elements, which cannot ensure the global consistency between the decomposition results and the original data; at the same time, there is a lack of application of sparsity constraints, which cannot force non-zero elements in the feature weight matrix to concentrate on significant feature peaks, and cannot suppress the interference of useless features and enhance feature interpretability.

[0005] 2. Currently, there is a lack of coupling hydrodynamic and fluorescence features to obtain the probability distribution of pollution sources and complete the source tracing of water pollution. Furthermore, the calculation process lacks the use of pollutant concentration fields, fluorescence feature migration coupling coefficients, fluorescence feature migration terms, feature weight gradients, effective diffusion coefficients, turbulence enhancement factors, and correction factors corresponding to turbulence intensity, which to some extent reduces prediction accuracy. Additionally, there is a lack of directly mapping the source tracing results to a multi-objective optimization function, and a lack of dynamic decision optimization through constructing a game-theoretic weight function of emergency response duration and total resource cost. This makes it impossible to reduce emergency response duration or significantly improve resource utilization. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the purpose of this application is to provide a method and system for linking water pollution source tracing with emergency decision-making.

[0007] To solve the above-mentioned technical problems, this application adopts the following technical solution: In the first aspect, this application provides a method for linking water pollution source tracing and emergency decision-making. The method includes the following steps: Step 1, data acquisition and enhancement: Acquire basic data of the current water area, wherein the basic data includes the surface-enhanced fluorescence signal, three-dimensional fluorescence spectral matrix and hydrodynamic data of the current water area, wherein the hydrodynamic data includes the flow velocity and turbulence intensity of the current water area.

[0008] Step 2, Dynamic Fluorescence Feature Registration: Extract the basic data of the current water area, and then analyze the dynamic changes of the surface-enhanced fluorescence signal according to the dynamic fluorescence feature registration function, thereby generating a fused spectral matrix.

[0009] Step 3: Pollutant characteristics: Collect pollutant data, extract the fused spectral matrix, and then perform non-negative matrix decomposition on the fused spectral matrix to obtain the feature weight matrix.

[0010] Step 4: Hydrodynamic-fluorescence coupling source tracing: Extract pollutant data and feature weight matrix, and then couple hydrodynamic and fluorescence features to analyze and obtain the probability distribution of pollution sources.

[0011] Step 5, Emergency Decision-Making Coordination: Extract the probability distribution of pollution sources, and then output the optimal emergency plan based on the multi-objective optimization function combined with the probability distribution of pollution sources.

[0012] Preferably, the basic data includes: recording the surface-enhanced fluorescence signal of the current water area as... The time-series fluorescence intensity is denoted as The three-dimensional fluorescence spectral matrix is ​​denoted as ,in Indicated as the time point of data collection. , and Let these be the excitation wavelength, emission wavelength, and depth, respectively; and let the current flow velocity and turbulence intensity in the hydrodynamic data be denoted as... and .

[0013] Preferably, the step of analyzing the dynamic change of the surface-enhanced fluorescence signal based on the dynamic fluorescence feature registration function includes: based on the dynamic fluorescence feature registration function... The dynamic changes in the surface-enhanced fluorescence signal were obtained. ,in This is represented as a time-dynamic compensation item. Represented as a spatial gradient compensation term, and These are represented as the time sensitivity coefficient and the spatial compensation coefficient, respectively.

[0014] Preferably, generating the fused spectral moments includes: according to the calculation formula Generate fusion spectral matrix .

[0015] Preferably, the pollutant data is the natural diffusion capacity of pollutants under still water conditions.

[0016] Preferably, the step of performing nonnegative matrix decomposition on the fused spectral matrix to obtain the feature weight matrix includes: according to the calculation formula The feature weight matrix is ​​obtained by performing nonnegative matrix decomposition. ,in Represented as a coefficient matrix, It is represented as the sparsity adjustment coefficient, where Represented as a data reconstruction item, It is represented as a sparse constraint term.

[0017] Preferably, the coupling of hydrodynamic and fluorescence characteristics includes: denoting the natural diffusion capacity of pollutants under still water conditions as... Substitute into the calculation formula The hydrodynamic and fluorescence features are coupled, whereby... Pollutant concentration field Represented as the fluorescence feature migration coupling coefficient, Represented as fluorescence feature transfer term, Represented as the feature weight gradient, It is expressed as the effective diffusion coefficient, and ,in Represented as the turbulence enhancement factor, This is expressed as the correction factor corresponding to the turbulence intensity.

[0018] Preferably, the analysis yields the probability distribution of pollution sources, including a calculation formula that couples discretized hydrodynamic and fluorescence characteristics. And iteratively obtain the pollutant concentration field This is then mapped to the probability distribution of pollution sources. .

[0019] Preferably, the step of outputting the optimal emergency response plan based on a multi-objective optimization function combined with the probability distribution of pollution sources includes: constructing a multi-objective optimization function. ,in This indicates the duration of the emergency response. Expressed as total resource cost, and These are respectively represented as the weighting factor corresponding to the emergency response duration and the weighting factor corresponding to the total resource cost, and are combined with... Dynamic adjustment and This allows us to output the optimal emergency response plan corresponding to the multi-objective optimization function.

[0020] In its second aspect, this application provides a system for linking water pollution source tracing and emergency decision-making, comprising: a data acquisition and enhancement module for acquiring basic data of the current water body, wherein the basic data includes surface-enhanced fluorescence signals, three-dimensional fluorescence spectral matrix and hydrodynamic data of the current water body, wherein the hydrodynamic data includes the flow velocity and turbulence intensity of the current water body.

[0021] The dynamic fluorescence feature registration module is used to extract the basic data of the current water area, and then analyze the dynamic changes of the surface-enhanced fluorescence signal according to the dynamic fluorescence feature registration function, thereby generating a fused spectral matrix.

[0022] The pollutant feature module is used to collect pollutant data, extract the fused spectral matrix, and then perform non-negative matrix decomposition on the fused spectral matrix to obtain the feature weight matrix.

[0023] The hydrodynamic-fluorescence coupled source tracing module is used to extract pollutant data and feature weight matrix, and then couple hydrodynamic and fluorescence features to analyze and obtain the probability distribution of pollution sources.

[0024] The emergency decision-making linkage module is used to extract the probability distribution of pollution sources, and then output the optimal emergency plan based on the multi-objective optimization function combined with the probability distribution of pollution sources.

[0025] The beneficial effects of this application are as follows: 1. The water pollution source tracing and emergency decision-making linkage method and system provided in this application analyze the water pollution source tracing and emergency decision-making linkage by combining surface-enhanced fluorescence (SEF) and three-dimensional fluorescence spectroscopy (EEM). This solves the limitations of the current water pollution source tracing and emergency decision-making linkage analysis process. It constructs a spatial correlation mapping for dynamic fluorescence feature registration, generates a source tracing weight matrix by combining a pollutant diffusion hydrodynamic model, and finally realizes pollution source location and emergency plan optimization through a multi-factor decision tree model. This solves the problems of insufficient sensitivity, high overlap rate of pollutant fingerprint features, and disconnect between source tracing and emergency response in traditional fluorescence detection technology.

[0026] 2. This application dynamically registers the response characteristics of surface-enhanced fluorescence nanomaterial substrates with 3D-EEM, and eliminates substrate interference by establishing a wavelength shift compensation function, which greatly improves the spectral resolution of pollutant characteristics and reduces cross-interference error.

[0027] 3. This application quantifies the reconstruction error by using the sum of squares of matrix elements, ensuring global consistency between the decomposition results and the original data; at the same time, by applying sparsity constraints, it forces the non-zero elements in the feature weight matrix to concentrate on the significant feature peaks, suppressing the interference of useless features and enhancing feature interpretability.

[0028] 4. This application couples hydrodynamic and fluorescence features to obtain the probability distribution of pollution sources and complete the source tracing of water pollution. In the calculation process, pollutant concentration field, fluorescence feature migration coupling coefficient, fluorescence feature migration term, feature weight gradient, effective diffusion coefficient, turbulence enhancement factor and correction factor corresponding to turbulence intensity are used to improve the prediction accuracy.

[0029] 5. This application directly maps the source tracing results to a multi-objective optimization function. By constructing a game-theoretic weight function between emergency response time and total resource cost, dynamic decision optimization is achieved, which reduces emergency response time and improves resource utilization. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart illustrating the steps involved in implementing the method described in this application.

[0032] Figure 2 This is a schematic diagram of the system structure connection of this application. Detailed Implementation

[0033] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0034] Please see Figure 1 As shown, this application provides a method for linking water pollution source tracing and emergency decision-making in the first aspect, including: Step 1, data acquisition and enhancement: acquiring basic data of the current water area, wherein the basic data includes surface-enhanced fluorescence signals, three-dimensional fluorescence spectral matrix and hydrodynamic data of the current water area, wherein the hydrodynamic data includes the flow velocity and turbulence intensity of the current water area.

[0035] In one specific example, the underlying data includes: denoting the surface-enhanced fluorescence signal of the current water area as... The time-series fluorescence intensity is denoted as The three-dimensional fluorescence spectral matrix is ​​denoted as ,in Indicated as the time point of data collection. , and Let these be the excitation wavelength, emission wavelength, and depth, respectively; and let the current flow velocity and turbulence intensity in the hydrodynamic data be denoted as... and .

[0036] It should be noted that, This indicates the spatial distribution differences of pollutant characteristics.

[0037] It should be noted that the intensity of the fluorescence signal is enhanced by using nanomaterials such as metal nanoparticles, and the surface-enhanced fluorescence signal is collected using a sensor modified with nanomaterials; a three-dimensional fluorescence spectral matrix is ​​collected using an in-situ fluorescence sensor; the flow velocity is collected using a flow velocity sensor; and the turbulence intensity is collected using a turbulence measurement instrument.

[0038] It should be noted that a microfluidic chip is used to integrate the SEF substrate; 3D-EEM scanning parameters: excitation wavelength 200-550nm (step size 5nm), emission wavelength 250-600nm.

[0039] Step 2, Dynamic Fluorescence Feature Registration: Extract the basic data of the current water area, and then analyze the dynamic changes of the surface-enhanced fluorescence signal according to the dynamic fluorescence feature registration function, thereby generating a fused spectral matrix.

[0040] In a specific example, the step of analyzing the dynamic change of the surface-enhanced fluorescence signal based on the dynamic fluorescence feature registration function includes: based on the dynamic fluorescence feature registration function... The dynamic changes in the surface-enhanced fluorescence signal were obtained. ,in This is represented as a time-dynamic compensation item. Represented as a spatial gradient compensation term, and These are represented as the time sensitivity coefficient and the spatial compensation coefficient, respectively.

[0041] It should be noted that, Represents surface-enhanced fluorescence signal Over time The change reflects the transient response of pollutant concentration; spatial gradient compensation term It is a Laplace calculation of the three-dimensional fluorescence spectral matrix, which enhances spectral resolution and separates overlapping peaks, representing the three-dimensional fluorescence spectral matrix. In the excitation-emission wavelength plane The gradient reflects the spatial distribution characteristics of pollutants.

[0042] It should be noted that, and Represented as the substrate feature compensation coefficient, this coefficient was determined through calibration experiments using a metal nanoparticle substrate in a standard contaminant solution. The response rate of the surface-enhanced fluorescence signal to contaminant concentration and the contribution weight of adjusting the gradient of the three-dimensional fluorescence spectral matrix to the fusion result were obtained. The response rate of the surface-enhanced fluorescence signal to contaminant concentration was then used as the time sensitivity coefficient, denoted as... The contribution weight of adjusting the gradient of the three-dimensional fluorescence spectral matrix to the fusion result is denoted as the spatial compensation coefficient. .

[0043] In a specific instance, generating the fused spectral moments includes: according to the calculation formula Generate fusion spectral matrix .

[0044] It should be noted that cross-validation was used to simulate pollution events in the laboratory and compare the fused spectral matrices. The results of the fusion were verified by combining the gas chromatography-mass spectrometry detection results with the results of the gas chromatography-mass spectrometry detection results, and the error was reduced to less than 8%.

[0045] This application dynamically registers the response characteristics of surface-enhanced fluorescence nanomaterial substrates with 3D-EEM, and eliminates substrate interference by establishing a wavelength shift compensation function, which greatly improves the spectral resolution of pollutant characteristics and reduces cross-interference error.

[0046] Step 3: Pollutant characteristics: Collect pollutant data, extract the fused spectral matrix, and then perform non-negative matrix decomposition on the fused spectral matrix to obtain the feature weight matrix.

[0047] In one specific instance, the pollutant data refers to the natural diffusion capacity of pollutants under still water conditions.

[0048] In a specific example, the nonnegative matrix decomposition of the fused spectral matrix to obtain the feature weight matrix includes: according to the calculation formula... The feature weight matrix is ​​obtained by performing nonnegative matrix decomposition. ,in Represented as a coefficient matrix, It is represented as the sparsity adjustment coefficient, where Represented as a data reconstruction item, It is represented as a sparse constraint term.

[0049] It should be noted that, By decomposing and fusing the spectral matrix into a feature weight matrix and a coefficient matrix, pollutant features with physical meaning are extracted. This means making the product of the decomposed feature weight matrix and the coefficient matrix infinitely close to the original fusion matrix, thereby preserving the spatial distribution and intensity information of the spectral features; This is represented by using L1 regularization to sparsify the feature weight matrix, thereby extracting pollutant features with significant discriminative power.

[0050] It should be noted that L1 regularization is an existing technique that achieves regularization by adding the L1 norm of the weights (the sum of the absolute values ​​of the weight vectors) to the model's loss function. L1 regularization tends to produce a sparse weight matrix, that is, pushing some weights to zero, thereby achieving the effect of feature selection.

[0051] It should be noted that the sparsity adjustment coefficient controls the sparsity of the feature weight matrix, balancing data reconstruction accuracy and feature selectivity. Cross-validation is used, and the feature weights are determined through grid search based on a known pollutant spectral database. At that time, the feature recognition accuracy is the highest; based on the signal-to-noise ratio requirements of pollutant characteristic peaks in environmental monitoring standards, the sparsity adjustment coefficient is dynamically adjusted. Values ​​are adapted to different water quality scenarios.

[0052] It should be noted that during the nonnegative matrix decomposition iteration process, the sparsity adjustment coefficient makes the unnecessary elements in the feature weight matrix approach zero, and at the same time defines the minimum significance threshold of the feature weights. Weights below this threshold are suppressed as useless features.

[0053] This application quantifies the reconstruction error by summing the squares of matrix elements, ensuring global consistency between the decomposition results and the original data; at the same time, by applying sparsity constraints, it forces the non-zero elements in the feature weight matrix to concentrate on the significant feature peaks, suppressing the interference of useless features and enhancing feature interpretability.

[0054] Step 4: Hydrodynamic-fluorescence coupling source tracing: Extract pollutant data and feature weight matrix, and then couple hydrodynamic and fluorescence features to analyze the probability distribution of pollution sources and complete the source tracing of water pollution.

[0055] In one specific example, the coupling of hydrodynamic and fluorescence characteristics includes: denoting the natural diffusion capacity of pollutants under still water conditions as... Substitute into the calculation formula The hydrodynamic and fluorescence features are coupled, whereby... Pollutant concentration field Represented as the fluorescence feature migration coupling coefficient, Represented as fluorescence feature transfer term, Represented as the feature weight gradient, It is expressed as the effective diffusion coefficient, and ,in Represented as the turbulence enhancement factor, This is expressed as the correction factor corresponding to the turbulence intensity.

[0056] It should be noted that the turbulence enhancement factor dynamically adjusts the diffusion rate according to the turbulence intensity, and A dataset was established to show the relationship between fluorescence feature migration coupling coefficients and pollutant types, such as benzene series compounds, through laboratory simulation experiments. and heavy metals wait.

[0057] It should be noted that the feature weight gradient is used as an additional driving force for the diffusion of pollution sources, thereby quantifying the spatial migration effect of fluorescence features.

[0058] In a specific example, the analysis yields a probability distribution of pollution sources, including a calculation formula that couples discretized hydrodynamic and fluorescence characteristics. And iteratively obtain the pollutant concentration field This is then mapped to the probability distribution of pollution sources. .

[0059] It should be noted that, according to the calculation formula The pollutant concentration field is converted into a pollution source probability distribution, and the high concentration region of the pollutant concentration field corresponds to the high probability position in the pollution source probability distribution.

[0060] This application couples hydrodynamic and fluorescence features to obtain the probability distribution of pollution sources and complete the source tracing of water pollution. In the calculation process, pollutant concentration field, fluorescence feature migration coupling coefficient, fluorescence feature migration term, feature weight gradient, effective diffusion coefficient, turbulence enhancement factor and correction factor corresponding to turbulence intensity are used to improve prediction accuracy.

[0061] Step 5, Emergency Decision-Making Coordination: Extract the probability distribution of pollution sources, and then output the optimal emergency plan based on the multi-objective optimization function combined with the probability distribution of pollution sources.

[0062] In a specific example, the step of outputting the optimal emergency response plan based on a multi-objective optimization function and the probability distribution of pollution sources includes: constructing a multi-objective optimization function. ,in This indicates the duration of the emergency response. Expressed as total resource cost, and These are respectively represented as the weighting factor corresponding to the emergency response duration and the weighting factor corresponding to the total resource cost, and are combined with... Dynamic adjustment and This allows us to output the optimal emergency response plan corresponding to the multi-objective optimization function.

[0063] It should be noted that emergency response time includes the time spent on pollution control decision-making and resource allocation; total resource costs include the costs of emergency equipment, manpower, and environmental remediation.

[0064] It should be noted that, in combination Dynamic adjustment and Construct a weight allocation model ,in , The Sigmoid function maps probabilities to an interval of... The weight, For the growth slope in control rights, for example when hour, That is, triggering high weight; among which Expressed as the pollution diffusion rate, for example, when diffusion is high, it decreases. Accelerate response, through Restrain resource capital and prevent excessive waste of resources.

[0065] It should be noted that, drive Improvement, will Reduce to minimum; Dynamic inhibition This is to prevent uneven distribution of resources.

[0066] It should be noted that the calculation results of the multi-objective optimization function all have corresponding optimal emergency solutions.

[0067] This application directly maps the source tracing results to a multi-objective optimization function. By constructing a game-theoretic weight function between emergency response time and total resource cost, dynamic decision optimization is achieved, which reduces emergency response time and improves resource utilization.

[0068] Please see Figure 2 As shown, in its second aspect, this application provides a system for a water pollution source tracing and emergency decision-making linkage method, comprising: a data acquisition and enhancement module for acquiring basic data of the current water area, wherein the basic data includes surface-enhanced fluorescence signals, three-dimensional fluorescence spectral matrix and hydrodynamic data of the current water area, wherein the hydrodynamic data includes the flow velocity and turbulence intensity of the current water area.

[0069] The dynamic fluorescence feature registration module is used to extract the basic data of the current water area, and then analyze the dynamic changes of the surface-enhanced fluorescence signal according to the dynamic fluorescence feature registration function, thereby generating a fused spectral matrix.

[0070] The pollutant feature module is used to collect pollutant data, extract the fused spectral matrix, and then perform non-negative matrix decomposition on the fused spectral matrix to obtain the feature weight matrix.

[0071] The hydrodynamic-fluorescence coupled source tracing module is used to extract pollutant data and feature weight matrix, and then couple hydrodynamic and fluorescence features to analyze and obtain the probability distribution of pollution sources.

[0072] The emergency decision-making linkage module is used to extract the probability distribution of pollution sources, and then output the optimal emergency plan based on the multi-objective optimization function combined with the probability distribution of pollution sources.

[0073] This application discloses a method and system for linking water pollution source tracing and emergency decision-making, relating to the field of water pollution source tracing and emergency decision-making linkage technology. This application includes data acquisition and enhancement, dynamic fluorescence feature registration, pollutant characterization, hydrodynamic fluorescence coupling for source tracing, and emergency decision-making linkage. By combining surface-enhanced fluorescence (SEF) with three-dimensional fluorescence spectroscopy (EEM) for analysis, the application analyzes the linkage between water pollution source tracing and emergency decision-making, solving the limitations of current water pollution source tracing and emergency decision-making linkage analysis processes. It constructs a spatial correlation mapping for dynamic fluorescence feature registration, generates a source tracing weight matrix by combining a pollutant diffusion hydrodynamic model, and finally achieves pollution source location and emergency plan optimization through a multi-factor decision tree model. This solves problems such as insufficient sensitivity, high overlap rate of pollutant fingerprint features, and disconnect between source tracing and emergency response in traditional fluorescence detection technologies.

[0074] The above content is merely an example and illustration of the concept of this application. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in this specification, they should all fall within the protection scope of this application.

Claims

1. A water pollution source tracing and emergency decision linkage method, characterized in that, Comprising: Step one, data acquisition and enhancement: collecting the basic data of the current water area, wherein the basic data includes the surface enhanced fluorescence signal, the three-dimensional fluorescence spectrum matrix and the water dynamic data of the current water area, wherein the water dynamic data includes the flow rate and the turbulence intensity of the current water area; The basic data includes: Let the surface enhanced fluorescence signal of the current water body be denoted as Let the time series fluorescence intensity be denoted as Let the three-dimensional fluorescence spectrum matrix be denoted as where is denoted as the acquisition time point, , and are denoted as the excitation wavelength, the emission wavelength and the depth, respectively; and let the flow velocity and the turbulence intensity of the current water body in the hydrodynamic data be denoted as and , respectively; Step two, dynamic fluorescence feature registration: extracting the basic data of the current water area, and then analyzing the dynamic change amount of the surface enhanced fluorescence signal according to the dynamic fluorescence feature registration function, so as to generate a fusion spectrum matrix; The dynamic fluorescence feature registration function includes: Dynamic fluorescence characteristic registration function Deriving a dynamic amount of surface enhanced fluorescence signal wherein is represented as a temporal dynamic compensation term, is represented as a spatial gradient compensation term, and are respectively represented as a temporal sensitivity coefficient and a spatial compensation coefficient; The fusion spectrum matrix includes: According to the calculation formula Generating a fusion spectrum matrix ; Step three, pollutant feature: collecting pollutant data, extracting the fusion spectrum matrix, and then performing non-negative matrix factorization on the fusion spectrum matrix, so as to obtain a feature weight matrix; The pollutant data is the natural diffusion capacity of the pollutant under static water conditions; The non-negative matrix factorization of the fusion spectrum matrix includes: According to the calculation formula performing non-negative matrix factorization to obtain a feature weight matrix wherein is represented as a coefficient matrix, is represented as a sparsity adjustment coefficient, wherein is represented as a data reconstruction term, is represented as a sparsity constraint term; Step four, water dynamic fluorescence coupling tracing: extracting the pollutant data and the feature weight matrix, and then coupling the water dynamic and fluorescence features, so as to analyze the pollution source probability distribution; The coupling of the water dynamic and fluorescence features includes: The natural dispersion capacity of the pollutant under quiescent conditions is noted , into the calculation formula The hydrodynamic and fluorescent features are coupled, where the pollutant concentration field, is expressed as a fluorescent feature migration coupling coefficient, is expressed as a fluorescent feature migration term, is expressed as a feature weight gradient, is expressed as an effective diffusion coefficient, and where is expressed as a turbulence enhancement factor, is expressed as a correction factor corresponding to the turbulence intensity; The analysis of the pollution source probability distribution includes: The discrete water power and fluorescence characteristics are coupled by a calculation formula, and the pollutant concentration field is iteratively obtained , and then mapped to the pollution source probability distribution ; Step five, emergency decision linkage: extracting the pollution source probability distribution, and then combining the pollution source probability distribution according to a multi-objective optimization function, so as to output an optimal emergency scheme.

2. The water pollution tracing and emergency decision linkage method according to claim 1, characterized in that, The multi-objective optimization function includes: Constructing a multi-objective optimization function wherein is represented as an emergency response duration, is represented as a total resource cost, and are respectively represented as a weight factor corresponding to the emergency response duration and a weight factor corresponding to the total resource cost, while combining dynamically adjusting and to thereby output an optimal emergency scheme corresponding to the multi-objective optimization function.

3. A system for performing the water pollution source tracing and emergency decision linkage method of any one of claims 1-2, characterized in that, Comprising: A data acquisition and enhancement module for collecting the basic data of the current water area, wherein the basic data includes the surface enhanced fluorescence signal, the three-dimensional fluorescence spectrum matrix and the water dynamic data of the current water area, wherein the water dynamic data includes the flow rate and the turbulence intensity of the current water area; A dynamic fluorescence feature registration module for extracting the basic data of the current water area, and then analyzing the dynamic change amount of the surface enhanced fluorescence signal according to the dynamic fluorescence feature registration function, so as to generate a fusion spectrum matrix; A pollutant feature module for collecting pollutant data, extracting the fusion spectrum matrix, and then performing non-negative matrix factorization on the fusion spectrum matrix, so as to obtain a feature weight matrix; A water dynamic fluorescence coupling tracing module for extracting the pollutant data and the feature weight matrix, and then coupling the water dynamic and fluorescence features, so as to analyze the pollution source probability distribution; An emergency decision linkage module for extracting the pollution source probability distribution, and then combining the pollution source probability distribution according to a multi-objective optimization function, so as to output an optimal emergency scheme.

Images