Acid mine water tracing method and device

Abstract

The invention provides an acid mine water traceability method and device, and relates to the technical field of pollution traceability, the method comprehensively considers water chemical fingerprints, environmental isotope composition, microbial community structure and multivariate data characteristics of soluble organic molecules, breaks through the limitation that a traditional method depends on a single index for traceability, and improves the traceability accuracy. Fusion analysis of multi-dimensional and multi-source information is realized; on the basis, a characteristic parameter combination capable of effectively distinguishing each end member is screened, a Bayesian end member mixed model constructed based on the law of conservation of mass and an MCMC solving method are combined, the probability relation between priori knowledge and actually measured data is fully utilized, and under the complex condition of mutual superposition and interference of pollution source signals, the probability relation between the priori knowledge and the actually measured data is fully utilized. The contribution ratio of each acid mine water end member to the mixed water sample and the confidence interval thereof are accurately deduced, and the scientificity, precision and reliability of a traceability result are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of pollution source tracing technology, and in particular to a method and apparatus for tracing the source of acidic mine water. Background Technology

[0002] Acid mine water (AMD) is a highly acidic, heavy metal-rich polluted water body generated during mining operations. Its formation is mainly related to the oxidation of sulfide minerals (such as pyrite). Accurately identifying the sources of AMD is crucial for developing targeted prevention and control measures during environmental remediation and ecological restoration in mining areas.

[0003] Traditional source tracing techniques for AMD (Advanced Melting Point) mainly include water chemical analysis, stable isotope tracing, and mineralogical characterization. These methods are typically based on the chemical composition or isotopic characteristics of a single water sample, identifying the pollution source by comparing characteristic parameters of potential pollution sources. However, in actual mining environments, AMD often results from a mixture of multiple pollution sources, leading to a complex superposition effect in the chemical composition of the water sample. When tracing mixed water samples, existing technologies struggle to accurately distinguish the contribution ratios of different pollution sources due to mutual interference of characteristic signals, significantly reducing the accuracy and reliability of the tracing results. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for tracing the source of acidic mine water, so as to alleviate the technical problems of low accuracy and poor reliability of existing methods for tracing the source of acidic mine water.

[0005] In a first aspect, the present invention provides a method for tracing the source of acidic mine water, comprising: acquiring multi-dimensional data features of each acidic mine water terminus within a study area; wherein, the multi-dimensional data features include: water chemical fingerprint, environmental isotope composition, microbial community structure, and dissolved organic matter molecules; based on the multi-dimensional data features of each acidic mine water terminus, screening characteristic parameter combinations used to distinguish each acidic mine water terminus; constructing a corresponding prior distribution for the content of each characteristic parameter of each acidic mine water terminus, obtaining a prior distribution set; acquiring the content of each characteristic parameter in the mixed water sample of acidic mine water to be traced, obtaining multi-dimensional measured characteristic values; and then... The posterior probability distribution set is used as the input data for the Bayesian endmember mixture model, and the multidimensional measured eigenvalues ​​are used as the observed values ​​of the likelihood function in the Bayesian endmember mixture model. The Monte Carlo Markov Chain (MCMC) method is used to solve the Bayesian endmember mixture model to obtain the posterior probability distribution of the contribution ratio of each acid mine water endmember to the acid mine water mixture sample. The Bayesian endmember mixture model is a model constructed based on the law of conservation of mass. The source tracing results of the acid mine water mixture sample are determined based on the posterior probability distribution. The source tracing results include: the target contribution ratio of each acid mine water endmember to the acid mine water mixture sample and its confidence interval.

[0006] In an optional implementation, based on the multi-data characteristics of each acidic mine water terminal element, a combination of feature parameters for distinguishing each acidic mine water terminal element is selected, including: standardizing the multi-data characteristics of all acidic mine water terminal elements to obtain multi-data standard features for each acidic mine water terminal element; removing redundant feature parameters from the multi-data standard features using a preset statistical method to obtain a first feature parameter combination; using a preset classification model to select a second feature parameter combination from multiple non-empty subsets of the first feature parameter combination, where the classification accuracy is greater than a first preset threshold and the increase in classification accuracy after adding new feature parameters is less than a second preset threshold; and using the second feature parameter combination as the feature parameter combination for distinguishing each acidic mine water terminal element.

[0007] In an optional implementation, a prior distribution is constructed for the content of each characteristic parameter of each acidic mine water terminal element, including: obtaining multiple measurement results of the content of multivariate data characteristics of the target acidic mine water terminal element; wherein, the target acidic mine water terminal element represents any one of the acidic mine water terminal elements in the study area; calculating the mean and standard deviation of the content of the target characteristic parameter of the target acidic mine water terminal element based on the multiple measurement results; wherein, the target characteristic parameter represents any one of the multivariate data characteristics; and using a normal distribution with mean and standard deviation as the prior distribution of the content of the target characteristic parameter of the target acidic mine water terminal element.

[0008] In an optional implementation, the prior distribution set is used as the input data of the Bayesian endmember mixture model, and the multidimensional measured feature values ​​are used as the observed values ​​of the likelihood function in the Bayesian endmember mixture model. The Bayesian endmember mixture model is solved using the Monte Carlo Markov Chain (MCMC) method, including: randomly initializing the contribution ratio of each acid mine water endmember to the acid mine water mixture sample to obtain a candidate contribution ratio set; under the constraint of the prior distribution set, initializing the content of each feature parameter of each acid mine water endmember to obtain a candidate content set; calculating the content of each feature parameter in the candidate mixture sample determined based on the candidate contribution ratio set and the candidate content set to obtain multidimensional candidate feature values; calculating the function value of the likelihood function in the Bayesian endmember mixture model based on the multidimensional candidate feature values ​​and the multidimensional measured feature values; retaining the candidate contribution ratio set whose function value exceeds a third preset threshold; repeating the iteration until a preset number of iterations is reached to determine the posterior probability distribution of the contribution ratio of each acid mine water endmember to the acid mine water mixture sample based on all retained candidate contribution ratio sets.

[0009] In an optional implementation, the source tracing result of the acidic mine water mixture sample is determined based on the posterior probability distribution, including: determining the expected value and 95% confidence interval of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixture sample based on the posterior probability distribution of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixture sample; wherein, the target acidic mine water end-member represents any one of the acidic mine water end-members in the study area; the expected value includes any one of the following: mean, median; the expected value and 95% confidence interval of the contribution ratio of all acidic mine water end-members to the acidic mine water mixture sample are used as the source tracing result.

[0010] In an optional implementation, the Bayesian endmember mixture model satisfies the following mass conservation relation: ;in, Indicating the first [item] in the mixed acidic mine water sample The content of various characteristic parameters, Indicates the first The contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample. Indicates the first The first acidic mine water end element The content of various characteristic parameters, This represents the total number of acidic mine water end-units. Indicates the first The model residuals of various characteristic parameters.

[0011] In an optional implementation, the formula for the likelihood function value in the Bayesian endmember mixture model is: ;in, This indicates the number of feature parameters in the feature parameter combination. Indicates the first Model residuals of various characteristic parameters Represents the th eigenvalue in the multidimensional measured eigenvalues The content of various characteristic parameters, Represents the first eigenvalue among multidimensional candidate eigenvalues The content of various characteristic parameters.

[0012] Secondly, the present invention provides an acid mine water tracing device, comprising: a first acquisition module for acquiring multi-data characteristics of each acid mine water terminus within a study area; wherein the multi-data characteristics include: water chemical fingerprint, environmental isotope composition, microbial community structure, and dissolved organic matter molecules; a screening module for screening characteristic parameter combinations used to distinguish each acid mine water terminus based on the multi-data characteristics of each acid mine water terminus; a construction module for constructing a corresponding prior distribution for the content of each characteristic parameter of each acid mine water terminus, thereby obtaining a prior distribution set; and a second acquisition module for acquiring the content of each characteristic parameter in the mixed water sample of acid mine water to be traced, thereby obtaining... The system comprises: a multidimensional measured eigenvalue module; a solution module, which uses the prior distribution set as input data for the Bayesian endmember mixture model and the multidimensional measured eigenvalues ​​as observed values ​​of the likelihood function in the Bayesian endmember mixture model; a Monte Carlo Markov Chain (MCMC) method to solve the Bayesian endmember mixture model, obtaining the posterior probability distribution of the contribution ratio of each acid mine water endmember to the acid mine water mixture sample; wherein, the Bayesian endmember mixture model is a model constructed based on the law of conservation of mass; and a determination module, which determines the source tracing results of the acid mine water mixture sample based on the posterior probability distribution; wherein, the source tracing results include: the contribution ratio of each acid mine water endmember to the acid mine water mixture sample and its confidence interval.

[0013] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the acidic mine water tracing method described in any of the foregoing embodiments.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the acid mine water tracing method described in any of the foregoing embodiments.

[0015] The acid mine water source tracing method provided by this invention overcomes the limitations of traditional methods that rely on a single indicator for source tracing by comprehensively acquiring multi-dimensional and multi-source data characteristics, including water chemical fingerprints, environmental isotope composition, microbial community structure, and dissolved organic matter molecules. It achieves multi-dimensional and multi-source information fusion analysis. Furthermore, by screening characteristic parameter combinations that can effectively distinguish each end-member, and combining a Bayesian end-member mixture model with the MCMC solution method, it fully utilizes the probabilistic relationship between prior knowledge and measured data. Under complex conditions of superimposed and interfering pollution source signals, it can accurately deduce the contribution ratio and confidence interval of each acid mine water end-member to the mixed water sample, significantly improving the scientific rigor, accuracy, and reliability of the source tracing results. Particularly noteworthy is that the Bayesian end-member mixture model, constructed based on the law of mass conservation, conforms to physical mechanism constraints, enhancing the model's rationality and interpretability. This effectively solves the technical problem of decreased source tracing accuracy due to feature signal aliasing when facing multi-source mixed pollution, providing strong technical support for source tracing and remediation of environmental pollution in mining areas. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 A flowchart of a method for tracing the source of acidic mine water provided in an embodiment of the present invention; Figure 2 A flowchart illustrating a method for solving a Bayesian endmember mixture model using the Monte Carlo Markov Chain (MCMC) method, as provided in this embodiment of the invention. Figure 3 A functional module diagram of an acidic mine water tracing device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0020] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0021] Example 1 Figure 1 A flowchart of a method for tracing the source of acidic mine water provided in an embodiment of the present invention is shown below. Figure 1 As shown, the method specifically includes the following steps: Step S102: Obtain the multi-data characteristics of each acidic mine water terminal element within the study area.

[0022] The multivariate data features include: water chemical fingerprint, environmental isotope composition, microbial community structure, and dissolved organic matter molecules.

[0023] To accurately trace the source of acidic mine water samples collected within the study area, this invention first requires analyzing the multi-dimensional data characteristics of each acidic mine water end-member within the study area. In mine water research, an "end-member" refers to the original water component that constitutes the mixed mine water and has a "stable chemical composition and clear origin," representing an indivisible "basic unit" in the mine water mixing process. Examples of acidic mine water end-members include seepage from different goaf areas, surface runoff, and background groundwater.

[0024] Among the multivariate data characteristics, the water chemical fingerprint includes: basic physicochemical parameters, common major ions, and characteristic heavy metals and trace elements. Basic physicochemical parameters include: pH value, electrical conductivity (EC), redox potential (Eh), etc. These describe the macroscopic chemical environment of the water body, such as acidity / alkalinity, total dissolved solids content, and redox state; common major ions include: K+. + Na + Ca 2+ Mg 2+ Cl - SO4 2- HCO3 - Their relative proportions can reflect the rock types through which the water flows and processes such as evaporation and concentration; characteristic heavy metals and trace elements include: Fe, Mn, Al, Cu, Zn, Cd, Pb, As, etc., which are often "indicators" of specific pollution sources.

[0025] Environmental isotopic composition includes: stable isotopes (δ¹²) in the water sample.2 H, δ 18 O, δ 34 S-SO4, δ 18 O-SO4, δ 13 C-DIC) and metal isotopes (δ) 56 Based on the differences in isotopic composition among different water bodies, Fe can help determine the origin of the water body, the geochemical processes it has undergone, and the mixing ratio. For example, the hydrogen and oxygen isotopes in water (δ¹⁰) 2 H, δ 18 O) can indicate the characteristics of water supply; sulfur isotopes (δ) 34 S-SO4) and iron isotopes (δ) 56 The combination of Fe and pyrite can accurately trace the oxidation source of sulfides (such as pyrite), and through the coupling relationship between the two, correct and resolve the isotopic fractionation effect caused by different sulfur sources (such as pyrite oxidation, gypsum dissolution) or different oxidation pathways, thus solving the ambiguity problem of tracing the source of a single sulfur isotope; carbon isotopes (δ¹⁰) 13 C-DIC can be used to help determine the impact of processes such as carbonate rock dissolution.

[0026] Microbial community structure analysis involves extracting microbial DNA from water samples and then using high-throughput sequencing technologies (such as 16S rRNA gene amplicon sequencing and metagenomic sequencing) to analyze the microbial community structure and functional genes, as well as to identify microbial groups specific to or with significantly different abundances in different acidic mine water termins, thereby constructing a microbial fingerprint of the water sample. Optionally, for the strongly acidic (pH < 4) conditions of AMD, the microbial community structure includes: microbial markers specific to acidic environments, such as 16S rRNA gene fragments of acidophilic sulfur oxidizing bacteria.

[0027] Dissolved organic matter molecules are obtained by analyzing the composition and molecular characteristics of dissolved organic matter in water samples using techniques such as three-dimensional fluorescence spectroscopy (EEMs) and / or Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). As an auxiliary tracer, it can help identify organic matter components associated with specific pollution sources or geochemical processes.

[0028] Step S104: Based on the multi-data characteristics of each acidic mine water terminal element, filter the combination of characteristic parameters used to distinguish each acidic mine water terminal element.

[0029] As described above, multivariate data features include a large number of feature parameters. While using more feature parameters leads to more accurate results when tracing the source of mixed acidic mine water samples, this embodiment of the invention filters the multivariate data features, removing feature parameters that have little or no impact on distinguishing acidic mine water end-members, retaining only combinations of feature parameters that effectively differentiate each acidic mine water end-member. Optionally, statistical methods can be used to filter the feature parameters, such as difference analysis, correlation analysis, principal component analysis, etc.

[0030] Step S106: Construct a prior distribution for the content of each characteristic parameter of each acidic mine water terminal element, and obtain a prior distribution set.

[0031] To accurately quantify the contribution ratio of each acidic mine water end-member in a mixed water sample, this invention employs a multi-parameter end-member mixing model based on Bayesian inference (i.e., the Bayesian end-member mixing model hereinafter). Solving this model requires prior information. Therefore, after determining the combination of characteristic parameters, it is necessary to further construct a corresponding prior distribution for the content of each characteristic parameter of each acidic mine water end-member. In other words, the content of each characteristic parameter of each acidic mine water end-member is not a fixed value, but a probability distribution.

[0032] For example, suppose there are 10 acidic mine water end-members in the study area, and the selected characteristic parameter combination is {Mg 2+ δ 18 O-SO4, δ 56 Fe}, that is, 3 characteristic parameters. In this step, it is necessary to sample each acidic mine water terminal element multiple times to obtain multiple measurements of the content of each characteristic parameter, so as to construct a prior distribution of the content of the 3 characteristic parameters for each acidic mine water terminal element. That is, a total of 10×3=30 prior distributions need to be constructed for 10 acidic mine water terminal elements. The 30 prior distributions constitute the prior distribution set.

[0033] Step S108: Obtain the content of each characteristic parameter in the acidic mine water mixed sample to be traced, and obtain multidimensional measured characteristic values.

[0034] Continuing with the example above, if the combination of characteristic parameters is {Mg} 2+ δ 18 O-SO4, δ 56 If {Fe}, then after obtaining the mixed acidic mine water sample to be traced, it is only necessary to obtain {Mg} from it. 2+ δ 18 O-SO4, δ 56The content of Fe is sufficient; the contents of the above three characteristic parameters are the multidimensional measured characteristic values.

[0035] Step S110: The prior distribution set is used as the input data of the Bayesian endmember mixture model, and the multidimensional measured feature values ​​are used as the observed values ​​of the likelihood function in the Bayesian endmember mixture model. The Bayesian endmember mixture model is solved using the Monte Carlo Markov Chain (MCMC) method to obtain the posterior probability distribution of the contribution ratio of each acid mine water endmember to the acid mine water mixture sample.

[0036] Among them, the Bayesian endmember mixture model is a model built based on the law of conservation of mass. Specifically, the Bayesian endmember mixture model satisfies the following mass conservation relation: ;in, Indicating the first [item] in the mixed acidic mine water sample The content of various characteristic parameters, Indicates the first The contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample. Indicates the first The first acidic mine water end element The content of various characteristic parameters, This represents the total number of acidic mine water end-units. Indicates the first The model residuals for each characteristic parameter represent measurement error and variation that the model fails to explain.

[0037] According to Bayes' theorem, the posterior distribution of a parameter is proportional to the product of the likelihood function and the prior distribution. However, due to the complexity of this posterior distribution, it is difficult to obtain an analytical solution. Therefore, this embodiment of the invention employs the Monte Carlo Markov Chain (MCMC) method for numerical solution. In solving the Bayesian endmember mixture model using the MCMC method, the multidimensional measured feature values ​​of the acidic mine water mixture sample to be traced are used as the observed input of the likelihood function. The goodness of fit of each candidate parameter combination (the contribution ratio of each endmember and the content of each feature parameter in each endmember) to the mixed water sample data is evaluated, and the set of contribution ratios with a high goodness of fit (the set composed of the contribution ratios of all endmembers) is retained, thereby obtaining the posterior probability distribution of the contribution ratio of each acidic mine water endmember to the acidic mine water mixture sample.

[0038] Step S112: Determine the source tracing results of the acidic mine water mixed sample based on the posterior probability distribution.

[0039] The source tracing results include: the target contribution ratio of each acid mine water end-member to the acid mine water mixture sample and its confidence interval.

[0040] Optionally, based on the posterior probability distribution of each contribution proportion, its statistical characteristic value can be calculated, including: expected value (mean or median) and confidence interval (such as 95% confidence interval). Then, the expected value is used as the best estimate of the contribution proportion of that endmember, that is, the aforementioned target contribution proportion, and the uncertainty range of the target contribution proportion is quantified according to the confidence interval.

[0041] Through the above steps, the embodiments of the present invention can not only quantitatively identify the main sources of acidic mine water mixed samples, but also provide the contribution ratio of each acidic mine water end-member and its uncertainty assessment, thereby achieving accurate and robust quantification of the source tracing results.

[0042] The acid mine water source tracing method provided in this invention overcomes the limitations of traditional methods that rely on a single indicator for source tracing by comprehensively acquiring multi-dimensional and multi-source data features, including water chemical fingerprints, environmental isotope composition, microbial community structure, and dissolved organic matter molecules. It achieves multi-dimensional and multi-source information fusion analysis. Furthermore, by screening feature parameter combinations that effectively distinguish each end-member, and combining a Bayesian end-member mixture model with the MCMC solution method, it fully utilizes the probabilistic relationship between prior knowledge and measured data. Under complex conditions of superimposed and interfering pollution source signals, it accurately infers the contribution ratio and confidence interval of each acid mine water end-member to the mixed water sample, significantly improving the scientific rigor, accuracy, and reliability of the source tracing results. Particularly noteworthy is that the Bayesian end-member mixture model, constructed based on the law of mass conservation, conforms to physical mechanism constraints, enhancing the model's rationality and interpretability. This effectively solves the technical problem of decreased source tracing accuracy due to feature signal aliasing when facing multi-source mixed pollution, providing strong technical support for source tracing and remediation of environmental pollution in mining areas.

[0043] In an optional implementation, step S104 above, based on the multi-data characteristics of each acidic mine water terminal element, filters the combination of characteristic parameters used to distinguish each acidic mine water terminal element, specifically including the following steps: Step S1041: Standardize the multi-data features of all acid mine water terminal elements to obtain multi-data standard features for each acid mine water terminal element.

[0044] In the default embodiment of the present invention, the multi-data features of each acidic mine water terminal element obtained in step S102 have no missing values ​​or outliers. If the pre-processing steps are not performed, then after obtaining the original measurement data, the missing value completion and outlier removal operations need to be performed.

[0045] Due to the significant differences in the dimensions of various characteristic parameters (e.g., pH range 1-6, SO42-2000), 2-(Concentration range 100-5000 mg / L) To eliminate the influence of dimensions and ensure comparability between different characteristic parameters, this embodiment of the invention requires standardization of the multi-data characteristics of all acidic mine water end-members to unify the scale. Standardization can be performed using Z-score or Min-Max normalization, etc., and this embodiment of the invention does not specifically limit them.

[0046] Step S1042: Use a preset statistical method to remove redundant feature parameters from the multivariate standard data features to obtain the first feature parameter combination.

[0047] The preset statistical methods can be difference analysis, correlation analysis, principal component analysis, etc., and this embodiment of the invention does not specifically limit them. Users can choose according to their actual needs. After processing by statistical methods, redundant feature parameters that have no significant impact on endmember differentiation can be initially eliminated.

[0048] Taking principal component analysis as an example, firstly, principal component analysis is performed on the features of multivariate standard data. Then, the contribution rate of each principal component is calculated, and the top k principal components with a cumulative contribution rate of more than 85% are selected. Next, the loading matrix is ​​analyzed to identify the original variables with high loadings on the principal components (i.e., parameters that contribute greatly to data variation). Finally, the original parameters with absolute loading values ​​greater than a threshold (such as 0.6) are retained as candidate feature sets, which are the first feature parameter combinations mentioned above.

[0049] Step S1043: Using a preset classification model, select a second feature parameter combination from multiple non-empty subsets of the first feature parameter combination whose classification accuracy is greater than the first preset threshold and whose improvement in classification accuracy after adding new feature parameters is less than the second preset threshold.

[0050] Step S1044: Use the second combination of characteristic parameters as the combination of characteristic parameters to distinguish each acidic mine water terminal element.

[0051] In order to balance the characteristics of the data structure with the requirements of the classification target and avoid subjective selection bias, after obtaining the first feature parameter combination through unsupervised dimensionality reduction, this embodiment of the invention further utilizes supervised optimization to perform deep screening on the first feature parameter combination.

[0052] Specifically, all non-empty subsets of the first feature parameter combination are identified, and these non-empty subsets are designated as the "potentially effective feature set". Next, a preset classification model is used for feature selection; the preset classification model can be Random Forest, Gradient Boosting Tree (XGBoost / LightGBM), etc. Then, the "potentially effective feature set" is used as input, and the "endmember category" is used as output to divide the model into training and testing sets. After model training, the models are sorted from high to low according to their "feature importance score" (e.g., sorting result: Fe > pH > SO4). 2- >ORP>Mn>EC).

[0053] Finally, feature parameters are added incrementally in order of importance, and the "endmember classification accuracy" is calculated. If adding a certain feature parameter results in an increase in classification accuracy of less than 1% (or begins to decrease), then the addition is stopped. This feature combination is the "optimal discriminative combination," also known as the second feature parameter combination. For example, consider the combination Fe + pH + SO4. 2- The combination of feature parameters achieves a classification accuracy of 92%. Adding ORP only improves the accuracy by 0.5%. Therefore, the final combination is the first 3 features.

[0054] In an optional implementation, step S106 above, which constructs a corresponding prior distribution for the content of each characteristic parameter of each acidic mine water terminal element, specifically includes the following steps: Step S1061: Obtain multiple measurement results of the content of multiple data features of the target acidic mine water terminal element; wherein, the target acidic mine water terminal element represents any one of the acidic mine water terminal elements in the study area.

[0055] Step S1062: Calculate the mean and standard deviation of the content of the target characteristic parameters of the target acidic mine water end-member based on multiple measurement results; wherein, the target characteristic parameter represents any one of the data features in the multivariate data features shown.

[0056] Step S1063: The normal distribution of the mean and standard deviation is used as the prior distribution of the content of the target characteristic parameter of the target acidic mine water terminal element.

[0057] Specifically, to construct a prior distribution for the content of each characteristic parameter of the target acidic mine water terminal element, it is necessary to collect water samples from the target acidic mine water terminal element multiple times and measure the content of each characteristic parameter in each water sample. Based on this, each characteristic parameter corresponds to multiple measurement results. This embodiment of the invention uses a normal distribution to describe the prior distribution of the characteristic parameters, i.e. ,in, This indicates the first measurement determined based on multiple measurements. The first acidic mine water end element The mean content of the characteristic parameters, This indicates the first measurement determined based on multiple measurements. The first acidic mine water end element The standard deviation of the content of a characteristic parameter.

[0058] In one alternative implementation, such as Figure 2 As shown, step S110 above uses the prior distribution set as the input data of the Bayesian endmember mixture model, and the multidimensional measured feature values ​​as the observed values ​​of the likelihood function in the Bayesian endmember mixture model. The Monte Carlo Markov Chain (MCMC) method is used to solve the Bayesian endmember mixture model, specifically including the following steps: Step S1101: Randomly initialize the contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample to obtain a candidate contribution ratio set.

[0059] It is known that the sum of the contribution ratios of all acidic mine water end-members to the acidic mine water mixture is 1. Therefore, this embodiment of the invention uses a Dirichlet distribution as the joint prior distribution of the contribution ratios. This naturally satisfies the constraint condition of the contribution ratio (non-negative and summing to 1) and allows setting it as an uninformative prior when there is no explicit prior information. That is, the candidate contribution ratio set obtained by random initialization should ensure that the contribution ratio of each acidic mine water end-member to the acidic mine water mixture is non-negative and sums to 1.

[0060] Step S1102: Under the constraints of the prior distribution set, initialize the content of each characteristic parameter of each acidic mine water end element to obtain the candidate content set.

[0061] The above steps have constructed a priori distributions for the content of each characteristic parameter in each acidic mine water terminal element. Therefore, during the model solution process, a random sample is drawn from each prior distribution, that is, the content of each characteristic parameter in each acidic mine water terminal element is determined. The samples randomly drawn from all prior distributions constitute the candidate content set, that is, a set of possible characteristic parameter contents is obtained.

[0062] Step S1103: Calculate the content of each characteristic parameter in the candidate mixed water sample determined based on the candidate contribution ratio set and the candidate content set, and obtain multidimensional candidate characteristic values.

[0063] After determining the candidate contribution ratio set and the candidate content set, the corresponding mixed water sample (i.e., candidate mixed water sample) can be determined based on these two sets. This embodiment of the invention utilizes a formula... Calculate the first candidate mixed water sample The content of various characteristic parameters; among them, Represents the first in the set of candidate contribution proportions The contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample. Indicates the first in the candidate content set The first acidic mine water end element The content of various feature parameters. Based on this formula, multidimensional candidate feature values ​​can be obtained: , This indicates the number of feature parameters in the feature parameter combination.

[0064] Step S1104: Calculate the likelihood function value in the Bayesian endmember mixture model based on the multidimensional candidate feature values ​​and the multidimensional measured feature values.

[0065] The embodiments of the present invention do not specifically limit the expression of the likelihood function. It is only necessary to ensure that the higher the degree of fit between the multidimensional candidate feature values ​​and the multidimensional measured feature values, the larger the function value of the likelihood function.

[0066] In one optional implementation, the formula for the likelihood function value in the Bayesian endmember mixture model is: ;in, This indicates the number of feature parameters in the feature parameter combination. Indicates the first Model residuals of various characteristic parameters Represents the th eigenvalue in the multidimensional measured eigenvalues The content of various characteristic parameters, Represents the first eigenvalue among multidimensional candidate eigenvalues The content of various characteristic parameters.

[0067] Step S1105: Retain the set of candidate contribution ratios whose function values ​​exceed the third preset threshold.

[0068] Repeat steps S1101-S1105 until the preset number of iterations is reached, and then execute step S1106.

[0069] Step S1106: Determine the posterior probability distribution of the contribution ratio of each acid mine water end-member to the acid mine water mixed sample based on the set of all retained candidate contribution ratios.

[0070] Through multiple iterations, a variety of candidate mixed water samples that are similar to the acidic mine water mixed water sample to be traced can be screened from a large number of sampled samples. Each candidate mixed water sample corresponds to a set of candidate contribution ratios. Then, the contribution ratio of each acidic mine water end-member to the acidic mine water mixed water sample to be traced corresponds to multiple values. Based on this, the posterior probability distribution of its contribution ratio can be constructed.

[0071] For example, if the set of contribution proportions is represented as After solving the Bayesian endmember mixture model using the Monte Carlo Markov Chain (MCMC) method, if the retained set of W contribution proportions is represented as... Then, the contribution ratio of each acidic mine water end-member to the acidic mine water mixture sample corresponds to W selectable values, where W are... The value of constitutes the first Posterior probability distribution of the contribution ratio of each acidic mine water end-member to the acidic mine water mixture: Therefore, the total result after this step is completed is... The posterior probability distribution of each contribution proportion.

[0072] In an optional implementation, step S112 above, which determines the source tracing result of the acidic mine water mixed sample based on the posterior probability distribution, specifically includes the following: First, based on the posterior probability distribution of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixed sample, the expected value and 95% confidence interval of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixed sample are determined; where the target acidic mine water end-member represents any one of the acidic mine water end-members in the study area; the expected value includes any one of the following: mean, median.

[0073] Then, the expected value and 95% confidence interval of the contribution ratio of all acid mine water end-members to the acid mine water mixture were used as the source tracing results.

[0074] As described above, the source tracing result output by this embodiment of the invention is no longer "end-member A contributes 70%". Instead, it uses the mean or median of the "posterior probability distribution of the contribution ratio of end-member A" as the best estimate. Furthermore, it directly calculates the confidence interval of the contribution ratio from the posterior probability distribution of the contribution ratio of end-member A to the acidic mine water mixed sample. This interval clearly indicates the possible range of the source tracing result, providing decision-makers with a basis for risk assessment. For example, the result can be expressed as "the contribution ratio of end-member A is 65%, and its 95% confidence interval is [55%, 75%]". This source tracing result is far more informative and scientifically rigorous than a single numerical value.

[0075] In summary, the present invention has the following significant beneficial effects: High accuracy in tracing: By integrating multi-source information such as water chemistry, isotopes, microorganisms and organic matter, it overcomes the limitations of single indicators under complex environmental conditions, and can more comprehensively capture the characteristic differences of different end-members, significantly improving the accuracy and reliability of end-member identification in acidic mine water.

[0076] Quantitative identification: This invention can not only qualitatively determine the source of pollution, but also quantitatively calculate the contribution ratio of each end element to the mixed mine water, providing a direct quantitative basis for pollution load assessment and liability determination.

[0077] It offers a wide range of information dimensions: integrating physicochemical indicators (water chemistry, isotopes), biological indicators (microorganisms), and organic geochemical indicators, it provides a deeper understanding of mine water systems and helps to reveal the geochemical and microbial driving mechanisms of AMD formation from different sources.

[0078] Example 2 This invention also provides an acid mine water tracing device, which is mainly used to execute the acid mine water tracing method provided in Embodiment 1 above. The following is a detailed description of the acid mine water tracing device provided in this invention.

[0079] Figure 3 This is a functional block diagram of an acidic mine water tracing device provided in an embodiment of the present invention, as shown below. Figure 3 As shown, the device mainly includes: a first acquisition module 11, a filtering module 12, a construction module 13, a second acquisition module 14, a solving module 15, and a determination module 16, wherein: The first acquisition module 11 is used to acquire multi-data characteristics of each acidic mine water end element in the study area; among which, the multi-data characteristics include: water chemical fingerprint, environmental isotope composition, microbial community structure and dissolved organic matter molecules.

[0080] The filtering module 12 is used to filter the combination of feature parameters used to distinguish each acid mine water terminal element based on the multi-data characteristics of each acid mine water terminal element.

[0081] Module 13 is used to construct a prior distribution for the content of each characteristic parameter of each acidic mine water terminal element, and obtain a prior distribution set.

[0082] The second acquisition module 14 is used to acquire the content of each characteristic parameter in the acidic mine water mixed sample to be traced, and obtain multidimensional measured characteristic values.

[0083] The solution module 15 is used to take the prior distribution set as the input data of the Bayesian endmember mixture model, the multidimensional measured feature values ​​as the observed values ​​of the likelihood function in the Bayesian endmember mixture model, and solve the Bayesian endmember mixture model using the Monte Carlo Markov Chain (MCMC) method to obtain the posterior probability distribution of the contribution ratio of each acid mine water endmember to the acid mine water mixture sample; wherein, the Bayesian endmember mixture model is a model constructed based on the law of conservation of mass.

[0084] The determination module 16 is used to determine the source tracing results of the acid mine water mixture sample based on the posterior probability distribution; wherein, the source tracing results include: the contribution ratio of each acid mine water end-member to the acid mine water mixture sample and its confidence interval.

[0085] The acid mine water tracing device provided in this invention overcomes the limitations of traditional methods that rely on a single indicator for tracing by comprehensively acquiring multi-dimensional and multi-source data features, including water chemical fingerprints, environmental isotope composition, microbial community structure, and dissolved organic matter molecules. It achieves multi-dimensional and multi-source information fusion analysis. Furthermore, by screening feature parameter combinations that effectively distinguish each end-member, and combining a Bayesian end-member mixture model with the MCMC solution method, it fully utilizes the probabilistic relationship between prior knowledge and measured data. Under complex conditions of superimposed and interfering pollution source signals, it accurately infers the contribution ratio and confidence interval of each acid mine water end-member to the mixed water sample, significantly improving the scientific rigor, accuracy, and reliability of the tracing results. Particularly noteworthy is that the Bayesian end-member mixture model, based on the law of mass conservation, conforms to physical constraints, enhancing the model's rationality and interpretability. This effectively solves the technical problem of decreased tracing accuracy due to feature signal aliasing when facing multi-source mixed pollution, providing strong technical support for tracing and controlling environmental pollution in mining areas.

[0086] Optionally, the filtering module 12 is specifically used for: Multivariate data features of all acid mine water end-units are standardized to obtain multivariate standard data features for each acid mine water end-unit.

[0087] Redundant feature parameters are removed from the features of multivariate standard data using a pre-defined statistical method to obtain the first feature parameter combination.

[0088] Using a preset classification model, select a second feature parameter combination from multiple non-empty subsets of the first feature parameter combination whose classification accuracy is greater than a first preset threshold, and whose improvement in classification accuracy after adding new feature parameters is less than a second preset threshold.

[0089] The second characteristic parameter combination is used as the characteristic parameter combination to distinguish each acid mine water terminal element.

[0090] Optionally, module 13 is specifically used for: Multiple measurements of the content of multi-data characteristics of the target acidic mine water terminal element are obtained; where the target acidic mine water terminal element refers to any one of the acidic mine water terminal elements in the study area.

[0091] The mean and standard deviation of the content of target characteristic parameters of the target acidic mine water end-member are calculated based on multiple measurement results; where the target characteristic parameter represents any one of the multivariate data characteristics shown.

[0092] The normal distribution of the mean and standard deviation is used as the prior distribution of the content of the target characteristic parameter of the target acidic mine water terminal component.

[0093] Optionally, the solver module 15 is specifically used for: The contribution ratio of each acid mine water end-member to the acid mine water mixed sample is randomly initialized to obtain a candidate contribution ratio set.

[0094] Under the constraints of the prior distribution set, the content of each characteristic parameter of each acidic mine water terminal element is initialized to obtain the candidate content set.

[0095] The content of each characteristic parameter in the candidate mixed water sample determined based on the candidate contribution ratio set and the candidate content set is calculated to obtain multidimensional candidate characteristic values.

[0096] The likelihood function value in the Bayesian endmember mixture model is calculated based on multidimensional candidate feature values ​​and multidimensional measured feature values.

[0097] The set of candidate contribution proportions whose function values ​​exceed a third preset threshold is retained.

[0098] Repeat the iteration until the preset number of iterations is reached, and determine the posterior probability distribution of the contribution ratio of each acid mine water end member to the acid mine water mixed sample based on the set of all retained candidate contribution ratios.

[0099] Optionally, module 16 is specifically used for: Based on the posterior probability distribution of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixed sample, the expected value and 95% confidence interval of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixed sample are determined; where the target acidic mine water end-member represents any acidic mine water end-member among all acidic mine water end-members in the study area; the expected value includes any of the following: mean, median.

[0100] The expected value and 95% confidence interval of the contribution ratio of all acid mine water end-members to the acid mine water mixture are used as the source tracing results.

[0101] Optionally, the Bayesian endmember mixture model satisfies the following mass conservation relation: ;in, Indicating the first [item] in the mixed acidic mine water sample The content of various characteristic parameters, Indicates the first The contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample. Indicates the first The first acidic mine water end element The content of various characteristic parameters, This represents the total number of acidic mine water end-units. Indicates the first The model residuals of various characteristic parameters.

[0102] Optionally, the formula for calculating the likelihood function value in the Bayesian endmember mixture model is: ;in, This indicates the number of feature parameters in the feature parameter combination. Indicates the first Model residuals of various characteristic parameters Represents the th eigenvalue in the multidimensional measured eigenvalues The content of various characteristic parameters, Represents the first eigenvalue among multidimensional candidate eigenvalues The content of various characteristic parameters.

[0103] Example 3 See Figure 4 This invention provides an electronic device, which includes a processor 60, a memory 61, a bus 62, and a communication interface 63. The processor 60, the communication interface 63, and the memory 61 are connected via the bus 62. The processor 60 is used to execute executable modules, such as computer programs, stored in the memory 61.

[0104] The memory 61 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 63 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0105] Bus 62 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0106] The memory 61 is used to store programs. After receiving an execution instruction, the processor 60 executes the program. The method executed by the apparatus defined by the process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 60 or implemented by the processor 60.

[0107] Processor 60 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 60 or by instructions in software form. Processor 60 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 61. Processor 60 reads the information in memory 61 and, in conjunction with its hardware, completes the steps of the above method.

[0108] The computer program product of the acidic mine water tracing method and apparatus provided in this embodiment of the invention includes a computer-readable storage medium storing non-volatile program code executable by a processor. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0109] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0110] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0111] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0112] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. In addition, the terms "first," "second," "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0113] Furthermore, terms such as "horizontal," "vertical," and "sag" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0114] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for tracing the source of acidic mine water, characterized in that, include: The multi-dimensional data characteristics of each acidic mine water element in the study area were obtained; wherein, the multi-dimensional data characteristics include: water chemical fingerprint, environmental isotope composition, microbial community structure and dissolved organic matter molecules; Based on the multi-data characteristics of each acidic mine water terminal element, a combination of characteristic parameters for distinguishing each acidic mine water terminal element is selected. For each characteristic parameter of each acidic mine water terminal element, a corresponding prior distribution is constructed to obtain the prior distribution set; The content of each characteristic parameter in the acidic mine water mixed sample to be traced is obtained, and multidimensional measured characteristic values ​​are obtained. The prior distribution set is used as the input data of the Bayesian endmember mixture model, and the multidimensional measured feature values ​​are used as the observed values ​​of the likelihood function in the Bayesian endmember mixture model. The Bayesian endmember mixture model is solved using the Monte Carlo Markov Chain (MCMC) method to obtain the posterior probability distribution of the contribution ratio of each acidic mine water endmember to the acidic mine water mixture sample. The Bayesian endmember mixture model is a model constructed based on the law of conservation of mass. The source tracing results of the acidic mine water mixture are determined based on the posterior probability distribution; wherein the source tracing results include: the target contribution ratio of each acidic mine water end-member to the acidic mine water mixture and its confidence interval.

2. The method for tracing the source of acidic mine water according to claim 1, characterized in that, Based on the multi-data characteristics of each acidic mine water terminal element, a combination of characteristic parameters for distinguishing each acidic mine water terminal element is selected, including: The multi-data characteristics of all acid mine water end elements are standardized to obtain the multi-data standard characteristics of each acid mine water end element. Redundant feature parameters are removed from the multivariate standard data features using a preset statistical method to obtain the first feature parameter combination; Using a preset classification model, select a second feature parameter combination from multiple non-empty subsets of the first feature parameter combination whose classification accuracy is greater than a first preset threshold, and whose improvement in classification accuracy after adding new feature parameters is less than a second preset threshold; The second combination of characteristic parameters is used as the combination of characteristic parameters to distinguish each acidic mine water terminal element.

3. The method for tracing the source of acidic mine water according to claim 1, characterized in that, Construct a prior distribution for the content of each characteristic parameter of each acidic mine water terminal element, including: Multiple measurements of the content of multi-data features of the target acidic mine water terminal element are obtained; wherein, the target acidic mine water terminal element represents any one of the acidic mine water terminal elements in the study area. Based on the results of the multiple measurements, the mean and standard deviation of the content of the target characteristic parameters of the target acidic mine water terminal element are calculated; wherein, the target characteristic parameter represents any one of the data features in the multivariate data features shown. The normal distribution of the mean and standard deviation is used as the prior distribution of the content of the target characteristic parameter of the target acidic mine water terminal element.

4. The method for tracing the source of acidic mine water according to claim 1, characterized in that, Using the prior distribution set as input data for the Bayesian endmember mixture model, and the multidimensional measured feature values ​​as observed values ​​of the likelihood function in the Bayesian endmember mixture model, the Bayesian endmember mixture model is solved using the Monte Carlo Markov Chain (MCMC) method, including: Randomly initialize the contribution ratio of each of the acidic mine water end-members to the acidic mine water mixed sample to obtain a candidate contribution ratio set; Under the constraints of the prior distribution set, the content of each characteristic parameter of each acidic mine water terminal element is initialized to obtain a candidate content set; The content of each characteristic parameter in the candidate mixed water sample determined based on the candidate contribution ratio set and the candidate content set is calculated to obtain multidimensional candidate characteristic values; Based on the multidimensional candidate feature values ​​and the multidimensional measured feature values, calculate the function value of the likelihood function in the Bayesian endmember mixture model; The set of candidate contribution proportions whose function values ​​exceed a third preset threshold is retained; The iteration is repeated until a preset number of iterations is reached, so as to determine the posterior probability distribution of the contribution ratio of each of the acidic mine water end-members to the acidic mine water mixture sample based on the set of all retained candidate contribution ratios.

5. The method for tracing the source of acidic mine water according to claim 1, characterized in that, The source tracing results of the acidic mine water mixture sample are determined based on the posterior probability distribution, including: Based on the posterior probability distribution of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixture sample, the expected value and 95% confidence interval of the contribution ratio of the target acidic mine water end-member to the acidic mine water mixture sample are determined; wherein, the target acidic mine water end-member represents any one of the acidic mine water end-members in the study area; the expected value includes any one of the following: mean, median; The expected value and 95% confidence interval of the contribution ratio of all acidic mine water end-members to the acidic mine water mixture sample are used as the source tracing results.

6. The method for tracing the source of acidic mine water according to claim 1, characterized in that, The Bayesian endmember mixture model satisfies the following mass conservation relation: ;in, Indicating the first [item] in the mixed acidic mine water sample The content of various characteristic parameters, Indicates the first The contribution ratio of each acidic mine water end-member to the acidic mine water mixed sample. Indicates the first The first acidic mine water end element The content of various characteristic parameters, This represents the total number of acidic mine water end-units. Indicates the first The model residuals of various characteristic parameters.

7. The method for tracing the source of acidic mine water according to claim 4, characterized in that, The formula for calculating the likelihood function value in the Bayesian endmember mixture model is as follows: ;in, This indicates the number of feature parameters in the feature parameter combination. Indicates the first Model residuals of various characteristic parameters Indicates the first of the multidimensional measured eigenvalues The content of various characteristic parameters, Indicates the first of the multidimensional candidate feature values The content of various characteristic parameters.

8. A source tracing device for acidic mine water, characterized in that, include: The first acquisition module is used to acquire multi-data characteristics of each acidic mine water end element in the study area; wherein, the multi-data characteristics include: water chemical fingerprint, environmental isotope composition, microbial community structure and dissolved organic matter molecules; The filtering module is used to filter the combination of feature parameters used to distinguish each acidic mine water terminal element based on the multi-data characteristics of each acidic mine water terminal element. A construction module is used to construct a prior distribution for the content of each characteristic parameter of each acidic mine water terminal element, thereby obtaining a set of prior distributions; The second acquisition module is used to acquire the content of each characteristic parameter in the acidic mine water mixed sample to be traced, and obtain multidimensional measured characteristic values. The solution module is used to take the prior distribution set as input data for the Bayesian endmember mixture model, and the multidimensional measured feature values ​​as observed values ​​of the likelihood function in the Bayesian endmember mixture model. The module then uses the Monte Carlo Markov Chain (MCMC) method to solve the Bayesian endmember mixture model, obtaining the posterior probability distribution of the contribution ratio of each acidic mine water endmember to the acidic mine water mixture sample. The Bayesian endmember mixture model is a model constructed based on the law of conservation of mass. A determination module is used to determine the source tracing result of the acidic mine water mixture sample based on the posterior probability distribution; wherein the source tracing result includes: the contribution ratio of each acidic mine water end-member to the acidic mine water mixture sample and its confidence interval.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the acid mine water tracing method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the acid mine water tracing method according to any one of claims 1 to 7.

Images