Detection and pollution tracing method based on four-dimensional fluorescent fingerprint

Abstract

The invention provides a detection and pollution tracing method and device based on four-dimensional fluorescent fingerprints, and relates to the technical field of detection. The method comprises the following steps: firstly, taking an enterprise involved in a suspected pollution industry as a suspected pollution source; then, collecting a target water body section sample, a target water body upstream unpolluted section sample and a suspected pollution source sample, and testing four-dimensional fluorescent fingerprints of all the samples; performing background signal deduction on the four-dimensional fluorescent fingerprint data based on the four-dimensional fluorescent fingerprint data of the upstream unpolluted section; calculating the fluorescence integral value of each component, and rejecting component data with the fluorescence integral value lower than a threshold value according to the set threshold value to form four-dimensional fluorescence fingerprint data; and calculating weight vectors of each component of the target water body cross section and the suspected pollution source, obtaining comprehensive similarity of each pollution source, and determining the attribution of the pollution source according to the comprehensive similarity. The method can achieve the precise recognition of different pollution emission sources in the same industry, and improves the traceability accuracy.

Description

Technical Field

[0001] This invention relates to the field of detection technology, and in particular to a detection and pollution source tracing method and apparatus based on four-dimensional fluorescent fingerprints. Background Technology

[0002] Source control is the primary measure in water pollution prevention and control, and pollution source tracing is an important prerequisite for source control. Among existing pollution source tracing technologies, a three-dimensional fluorescence fingerprint system is constructed by detecting the excitation wavelength, emission wavelength, and fluorescence intensity of water samples using three-dimensional fluorescence spectroscopy. When tracing pollution sources, public security organs can quickly identify pollution discharge sources by comparing the fluorescence fingerprints of the affected wastewater samples with those of the pollution source water.

[0003] However, in actual source tracing, the fluorescent fingerprints of wastewater from different companies within the same industry are often very similar and difficult to distinguish, resulting in tracing the source only to a specific industry and not to a specific company. This technical deficiency is particularly pronounced in highly polluting industries such as chemical and pharmaceutical manufacturing, where the high similarity in the composition of organic pollutants in their wastewater discharges makes it difficult for traditional methods to overcome the technical bottleneck of industry-level source tracing, leading to low efficiency in emergency response to pollution incidents and high treatment costs. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the first objective of this invention is to propose a detection and pollution source tracing method based on four-dimensional fluorescent fingerprinting, comprising: S1 uses water quality fluorescent fingerprint tracing technology to identify water pollution sources, clarify the industry of pollution source, and determine the suspected pollution source based on the industry of pollution source; S2, collect samples from the target water body section, the unpolluted section upstream of the target water body, and the suspected pollution source, and test the four-dimensional fluorescent fingerprint of each sample; S3, Based on the four-dimensional fluorescence fingerprint data of the upstream unpolluted section, background signal subtraction is performed on the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source, respectively. S4. After removing the background signal, calculate the fluorescence integral value of each component in the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source. Remove component data with fluorescence integral values ​​lower than the threshold according to the set threshold to form four-dimensional fluorescence fingerprint data of the effective components of the target water body section and the suspected pollution source. S5. Based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source, calculate the weight vector of each component of the target water body section and the suspected pollution source respectively. S6. Based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after removing the background signal, the component-level similarity is calculated, and the comprehensive similarity of each pollution source is obtained. The pollution source attribution is determined according to the comprehensive similarity.

[0006] In one embodiment of the present invention, the four-dimensional fluorescent fingerprint includes: components, excitation wavelength, emission wavelength, and fluorescence intensity.

[0007] In one embodiment of the present invention, step S2 further includes: S21 uses a liquid chromatography column to process the sample and separate the ordered components in the sample; S22, obtain the water quality fluorescence fingerprints of each component of the sample, and arrange the water quality fluorescence fingerprints of each component in the order of the components to form the four-dimensional fluorescence fingerprint of the sample.

[0008] In one embodiment of the present invention, step S22 further includes: A three-dimensional fluorescence spectroscopy detection module was used to test the fluorescent fingerprint of water quality.

[0009] In one embodiment of the present invention, step S3 further includes: By subtracting the fluorescence intensity of the same component in the unpolluted section sample upstream of the target water body from the fluorescence intensity of each component in the target water body section sample, the four-dimensional fluorescence fingerprint data of the target water body section sample after removing the background signal is obtained. By subtracting the fluorescence intensity of the same component in the unpolluted section upstream of the target water body from the fluorescence intensity of each component in the suspected pollution source sample, four-dimensional fluorescence fingerprint data of the suspected pollution source after removing background signals is obtained.

[0010] In one embodiment of the present invention, step S4 further includes: S41, Calculate the fluorescence integral value of each component based on the processed four-dimensional fluorescence fingerprint data of the target water body section. The calculation formula is as follows:

[0011] in, Indicates in the sample The In the water fluorescence fingerprint of the components, the first... The first excitation wavelength at the _th The fluorescence intensity corresponding to each emission wavelength n The total number of emission wavelengths, m This represents the total number of laser wavelengths. S42, Remove Samples The fluorescence integral value is less than The component data, among which, aTo set the weights, the values ​​range from [0,1]. After the elimination is completed, the data are sorted by component number to obtain the four-dimensional fluorescence fingerprint data of the effective components.

[0012] In one embodiment of the present invention, step S5 further includes: S51, based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data, calculate the weight vector of each component. The calculation formula is as follows:

[0013] in, For the sample X The first of the effective four-dimensional fluorescent fingerprint data Fluorescence integral values ​​of each component.

[0014] In one embodiment of the present invention, step S6 further includes: S61, Component-level similarity is calculated based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after background signal removal. The calculation formula is as follows:

[0015] in, This indicates the first of the four-dimensional fluorescent fingerprint data of the suspected pollution source. The water quality fluorescence fingerprints of each component and the four-dimensional fluorescence fingerprint data of the target water body section are shown in the figure. The similarity of water quality fluorescent fingerprints of individual components, The fourth fluorescent fingerprint data representing the effective components of the suspected pollution source Water fluorescence fingerprints of individual components The fourth fluorescence fingerprint data representing the effective components of the target water body section. Water fluorescence fingerprints of individual components This indicates that similarity calculations are performed on the two selected water quality fluorescent fingerprint data. S62, calculate the comprehensive similarity between the target water body cross-section sample and each pollution source. The calculation formula is as follows:

[0016] in, Indicates pollution source With the target section Similarity of four-dimensional fluorescent fingerprints.

[0017] To achieve the above objectives, a second aspect of the present invention provides a detection and pollution source tracing device based on four-dimensional fluorescent fingerprinting, comprising: The four-dimensional fluorescent fingerprint acquisition module is used to test the four-dimensional fluorescent fingerprints of target water body cross-section samples, unpolluted cross-section samples upstream of the target water body, and samples from suspected pollution sources. The background signal subtraction module is used to subtract the background signal from the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint of the suspected pollution source sample based on the four-dimensional fluorescence fingerprint data of the unpolluted section upstream of the target water body, so as to obtain the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source sample after deducting the background value. The effective component screening module is used to screen effective four-dimensional fluorescence fingerprint data using the fluorescence integration method, calculate the fluorescence integral value of each component, and remove component data with fluorescence integral values ​​lower than the set threshold according to the set threshold, so as to form four-dimensional fluorescence fingerprint data of effective components of target water body section and suspected pollution source. The weighted calculation module calculates the weight vectors of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source, based on the fluorescence integral values ​​of each component in the target water body section and the suspected pollution source. The pollution source attribution determination module performs component-level similarity calculations based on the four-dimensional fluorescence fingerprint data of suspected pollution sources and target water body sections after removing background signals, and obtains the comprehensive similarity of each pollution source. The pollution source attribution is then determined based on the comprehensive similarity.

[0018] To achieve the above objectives, a third aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0019] The methods, systems, and storage media of this invention can accurately identify different pollution emission sources in the same industry, significantly improving the accuracy and relevance of pollution source tracing.

[0020] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0021] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a detection and pollution source tracing method based on four-dimensional fluorescent fingerprinting according to an embodiment of the present invention; Figure 2 This is a structural diagram of a device for detection and pollution source tracing based on four-dimensional fluorescent fingerprints according to an embodiment of the present invention; Figure 3 This is a water quality fluorescent fingerprint of a surface water section according to an embodiment of the present invention; Figure 4 This is a water quality fluorescent fingerprint of the wastewater of Company A according to an embodiment of the present invention; Figure 5 The water quality fluorescent fingerprint of Enterprise B's wastewater according to an embodiment of the present invention; Figure 6 This is a water quality fluorescence fingerprint of each effective component in a four-dimensional fluorescence fingerprint of a surface water section according to an embodiment of the present invention. Figure 7 This is a water quality fluorescence fingerprint of each effective component in the four-dimensional fluorescence fingerprint of the wastewater of Company A according to an embodiment of the present invention; Figure 8 This is a water quality fluorescence fingerprint of each effective component in the four-dimensional fluorescence fingerprint of the wastewater of Company B according to an embodiment of the present invention. Detailed Implementation

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0024] The following description, with reference to the accompanying drawings, describes a method, system, and computer-readable storage medium for detection and pollution tracing based on four-dimensional fluorescent fingerprints, according to embodiments of the present invention.

[0025] Example 1 Figure 1 This is a flowchart of a four-dimensional fluorescent fingerprint-based detection and pollution source tracing method according to an embodiment of the present invention.

[0026] like Figure 1 As shown, the detection and pollution source tracing method based on four-dimensional fluorescent fingerprinting includes the following steps: S1 uses water quality fluorescent fingerprint tracing technology to identify the source of water pollution, clarify the industry of pollution source, and determine the suspected pollution source based on the industry of pollution source.

[0027] S2, test the four-dimensional fluorescence fingerprint of the target water body section, its upstream unpolluted section, and the suspected pollution source. The four-dimensional fluorescence fingerprint is formed by arranging the ordered components of the water quality fluorescence fingerprint obtained by high performance liquid chromatography in order of component to form a four-dimensional tensor data structure. The water quality fluorescence fingerprint of each component includes three dimensions: excitation wavelength, emission wavelength, and fluorescence intensity.

[0028] Specifically, this step involves testing the target water body section, its upstream unpolluted section, and the four-dimensional fluorescence fingerprint of the suspected pollution source. It is the fundamental data acquisition step of the entire pollution source tracing method. Its technical implementation principle is based on the combined use of high-performance liquid chromatography (HPLC) and three-dimensional fluorescence spectroscopy (EEM, Excitation-Emission Matrix). By separating the organic components in the water sample according to their retention time, and acquiring the three-dimensional fluorescence spectra of each component at excitation wavelength (Ex), emission wavelength (Em), and fluorescence intensity (I), a four-dimensional tensor data structure containing the four dimensions of component, excitation wavelength, emission wavelength, and fluorescence intensity is ultimately formed.

[0029] In some implementations, the testing process for four-dimensional fluorescent fingerprints is as follows: First, the water sample is injected into a high-performance liquid chromatography system, and the sample is separated into... There are several ordered components, among which... Determined by chromatographic conditions (such as mobile phase composition, flow rate, column temperature, etc.), typically in Within the specified range. Subsequently, the eluent of each component was automatically collected and introduced into a three-dimensional fluorescence detection module for excitation wavelength range. emission wavelength range The scanning sampling interval is usually set to To ensure spectral resolution and data integrity, the fluorescence intensity of each component was measured. It is recorded as a two-dimensional matrix, where Indicates the excitation wavelength index. This indicates the emission wavelength index.

[0030] Furthermore, the structure of a four-dimensional fluorescent fingerprint can be represented as a The three-dimensional tensor, in which and These represent the number of sampling points for the excitation wavelength and the emission wavelength, respectively. In practical applications, this step is typically performed in the initial identification phase of pollution source tracing to obtain the target cross-section. Upstream section and suspected pollution sources The complete four-dimensional fingerprint data provides a unified data foundation for subsequent background subtraction (i.e., subtracting the corresponding fluorescence intensity from the unpolluted section sample upstream of the target water body), effective data screening, and similarity calculation, and is a key prerequisite for achieving high-precision pollution source tracing.

[0031] Furthermore, S2 includes: S21 uses a liquid chromatography column to process the sample and separate the ordered components in the sample; Specifically, in some implementations, high-performance liquid chromatography (HPLC) columns are used to separate the sample, separating the sample into... Separating ordered components is a crucial step in constructing a four-dimensional fluorescent fingerprint. This step is based on the principles of high-performance liquid chromatography (HPLC), which uses a chromatographic column to separate different chemical components in the sample, thereby obtaining a component sequence with time-resolved characteristics. Specifically, the sample enters the chromatographic column under the impetus of the mobile phase, and different components are eluted and collected sequentially due to differences in their partition coefficients between the stationary and mobile phases. In this technical solution, the separation capability of the chromatographic column directly affects the resolution and recognition accuracy of the subsequent four-dimensional fluorescent fingerprint. Therefore, a chromatographic column with high separation efficiency and good repeatability must be selected, such as a C18 reversed-phase column or an ion-exchange column. The specific selection depends on the polarity and molecular structure characteristics of the organic matter in the sample.

[0032] At the parameter level, key parameters such as mobile phase composition, flow rate, and column temperature need to be controlled during the separation process. For example, the mobile phase is usually a mixture of water and an organic solvent (such as acetonitrile or methanol), and its ratio can be optimized by gradient based on the polarity of the target pollutant; the flow rate is generally set in the range of 0.5–1.5 mL / min to ensure sufficient separation of components without causing peak broadening; the column temperature is usually maintained between 25–40°C to improve separation stability and reproducibility. After separation, the sample is divided into... There are several ordered components, each corresponding to a specific elution time window, and their data structure is as follows: ,in .

[0033] In practical applications, this step is typically performed in a high-throughput liquid chromatography system with semi-continuous collection capabilities. The system integrates an automated eluent collection module and a three-dimensional fluorescence detection module at the end, enabling automatic separation of components and acquisition of fluorescence fingerprints. By coupling traditional HPLC with three-dimensional fluorescence detection technology, this step not only improves sample separation efficiency but also provides a structured data foundation for the subsequent construction of four-dimensional fluorescence fingerprints.

[0034] In terms of technical effectiveness, this step significantly enhances the resolution of water quality fluorescent fingerprints by introducing a component dimension. This allows for differentiation based on component order and intensity distribution even when different companies in the same industry exhibit highly similar fluorescence characteristics. Therefore, this step plays a crucial supporting role in pollution source tracing systems and is one of the core components for achieving high-precision pollution source identification.

[0035] S22, obtain the water quality fluorescence fingerprints of each component of the sample, and arrange the water quality fluorescence fingerprints of each component in the order of the components to form the four-dimensional fluorescence fingerprint of the sample.

[0036] Specifically, this step involves constructing a four-dimensional fluorescent fingerprint of the water sample. The core of this process is separating the sample into several ordered components using high-performance liquid chromatography (HPLC), and then performing three-dimensional fluorescence spectroscopy (EEM, Excitation-Emission Matrix) detection on each component. Finally, the EEM data of all components are arranged sequentially to form a four-dimensional tensor data structure with four dimensions: component, excitation wavelength, emission wavelength, and fluorescence intensity. It is important to note that if the wavelength of a component passes through the Raman or Rayleigh scattering region, the influence of scattered light needs to be eliminated to obtain more accurate data results. This step is a crucial foundation for achieving accurate pollution source identification, especially when the fluorescent fingerprints of different companies in the same industry are highly similar. By introducing the component dimension, the resolution and source tracing capabilities of the fingerprint are significantly improved.

[0037] In some implementations, this step first involves separating the water sample using an HPLC system. The selection of the chromatographic column and the configuration of the mobile phase need to be optimized based on the properties of the target pollutant. A C18 reversed-phase column is typically used, with a gradient-eluting water-organic solvent system (such as acetonitrile / water) as the mobile phase. The flow rate is controlled within the range of 0.5–1.0 mL / min, and the separation time is approximately 20–40 minutes, thereby separating the sample into q ordered components. The eluent of each component is captured by an automated collection module and then sent to a three-dimensional fluorescence spectroscopy detection module for testing via an automated injection module.

[0038] Three-dimensional fluorescence spectroscopy detection modules are typically configured with an excitation wavelength range of 200–400 nm, an emission wavelength range of 300–600 nm, a step size of 5–10 nm, and an integration time of 0.1–1.0 seconds to ensure the signal-to-noise ratio and resolution of the data. The EEM data for each component can be represented as follows: Its structure is The matrix, where This represents the fluorescence intensity of the k-th component at the i-th excitation wavelength and the j-th emission wavelength.

[0039] Furthermore, the EEM data of all components are arranged in the order of separation to form a... The three-dimensional dataset, combined with the component index dimension, ultimately forms a 4-dimensional tensor, namely... The complete data structure is presented. This tensor not only preserves the spectral information of traditional three-dimensional fluorescent fingerprints, but also introduces the temporal order of component separation, thereby enhancing the fingerprint's feature representation capability.

[0040] In practical applications, this step is suitable for scenarios such as water environment monitoring, industrial wastewater source tracing, and emergency response to sudden pollution incidents. By constructing high-dimensional fingerprint data, a structured, quantifiable, and comparable data foundation is provided for subsequent background deduction, similarity calculation, and pollution source matching, which is a key technical step in achieving high-precision pollution source tracing.

[0041] S3, based on the four-dimensional fluorescence fingerprint data of the upstream unpolluted section, background signal subtraction is performed on the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source.

[0042] Specifically, this step involves subtracting background signals from the four-dimensional fluorescence fingerprint data of the target water body section based on the four-dimensional fluorescence fingerprint data of the unpolluted upstream section, thereby obtaining the four-dimensional fluorescence fingerprint data of the target water body section after background value subtraction. This step is one of the key links in achieving accurate identification of pollution sources. Its technical principle is based on dynamic compensation of background signals to eliminate the interference of natural water body background components on pollution characteristics.

[0043] At the technical implementation level, four-dimensional fluorescence fingerprint data consists of four dimensions: composition, excitation wavelength, emission wavelength, and fluorescence intensity. Its data structure is a four-dimensional tensor. (Target water body cross-section) Compared with the unpolluted section upstream The four-dimensional data were acquired under the same chromatographic separation conditions to ensure consistent component order. For water body sections... Each component The corresponding water quality fluorescent fingerprint Compared with the unpolluted section upstream In At the same excitation-emission wavelength pair Perform point-by-point subtraction on the above, that is:

[0044] in, This represents the fluorescence intensity value after background subtraction. This operation is performed independently on the EEM (excitation-emission matrix) map of each component, thereby preserving anomalous fluorescence signals introduced by contamination in the target section while removing interference from natural background components.

[0045] At the parameter level, the background subtraction process requires upstream cross-sections. With the target section The chromatographic separation conditions are completely consistent, including mobile phase composition, flow rate, column temperature, and injection volume, to ensure component alignment. Furthermore, the excitation wavelength range for fluorescence detection is typically set to... The emission wavelength range is The wavelength interval is To ensure data resolution and comparability, the subtracted data must meet the requirement of non-negative fluorescence intensity; negative values ​​are typically set to zero or ignored.

[0046] At the application level, this step is widely used in tracing the source of water pollution, especially when multiple companies in the same industry have similar emission characteristics. By subtracting the background signal from the upstream section, fluorescence features introduced by specific pollution sources in the target section can be effectively identified, thereby improving the accuracy of subsequent similarity calculations and the precision of source tracing.

[0047] The technical effect of this step is to significantly reduce the interference of background noise on pollution characteristics and improve the contrast sensitivity and specificity of four-dimensional fluorescent fingerprints. Through accurate background subtraction, the ability to identify differences between pollution sources and target sections can be enhanced, providing a high-quality data foundation for subsequent similarity weighting calculations and pollution source matching, thereby achieving accurate source tracing of pollution.

[0048] Furthermore, S3 includes: By subtracting the fluorescence intensity of the same component in the unpolluted section sample upstream of the target water body from the fluorescence intensity of each component in the target water body section sample, the four-dimensional fluorescence fingerprint data of the target water body section sample after removing the background signal is obtained. By subtracting the fluorescence intensity of the same component in the unpolluted section upstream of the target water body from the fluorescence intensity of each component in the suspected pollution source sample, four-dimensional fluorescence fingerprint data of the suspected pollution source after removing background signals is obtained.

[0049] From a parametric perspective, the accuracy of background subtraction directly affects the reliability of subsequent similarity calculations. Typically, good repeatability and stability of the fluorescence intensity data from the upstream section S are required.

[0050] In application scenarios, this step is widely used in pollution source tracing analysis of natural water bodies such as rivers and lakes, especially in areas where multiple enterprises discharge similar industrial wastewater. Background subtraction can significantly enhance the characteristic differences between the target section and the pollution source. For example, in a chemical industrial park, the wastewater from different enterprises may have similar fluorescent fingerprints, but by subtracting the background signal from the upstream section, the emission characteristics of a specific enterprise can be effectively identified.

[0051] The technical advantage of this step lies in improving the signal-to-noise ratio of the four-dimensional fluorescent fingerprint by eliminating background interference, thereby enhancing the resolution and accuracy of pollution source identification. It is a crucial pre-processing step for achieving traceability from industry-level to enterprise-level, laying the data foundation for subsequent similarity-weighted calculations.

[0052] S4. After removing the background signal, calculate the fluorescence integral value of each component in the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source. Remove component data with fluorescence integral values ​​lower than the set threshold to form four-dimensional fluorescence fingerprint data of the effective components of the target water body section and the suspected pollution source.

[0053] Specifically, this step uses the fluorescence integral method to screen the four-dimensional fluorescence fingerprint data to extract effective components with significant fluorescence signals, thereby improving the accuracy and reliability of subsequent pollution source identification. Specifically, this step calculates the fluorescence integral value of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source. The formula for calculating the fluorescence integral value is as follows: ; in, Indicates in the sample The In the water fluorescence fingerprint of the components, the first... The first excitation wavelength at the _th The fluorescence intensity corresponding to each emission wavelength n The total number of emission wavelengths, m This represents the total number of laser wavelengths.

[0054] The components in the four-dimensional fluorescent fingerprint are quantitatively evaluated based on the fluorescence integral value, and a set threshold is set accordingly. Component data with integral values ​​below this threshold are removed, among which, a To set the weights, the values ​​range from [0,1]. Finally, after removing component data with integral values ​​below this threshold, four-dimensional fluorescence fingerprint data of the target water body section and the effective components of the suspected pollution source are obtained.

[0055] At the technical implementation level, the fluorescence integration method involves analyzing the three-dimensional fluorescence spectrum (EEM) of each component at the excitation wavelength. and emission wavelength The total fluorescence intensity of the component is obtained by performing integration operations in two dimensions. The integration operation is achieved through double summation, that is, by... excitation wavelength and The emission wavelengths are accumulated point by point to reflect the overall fluorescence characteristics of the component. This method can effectively filter out components with low intensity and high noise interference, while retaining fluorescence signals that are representative of the pollution characteristics.

[0056] At the parameter level, This is the integration threshold coefficient, and its value range is... These are the screening criteria used to control the selection of effective components. Typically, The value needs to be optimized based on the fluorescence intensity distribution characteristics of the actual sample to ensure that invalid signals are eliminated without losing key component information. Indicates sample The maximum value of the fluorescence integral of all components is used as the normalization reference.

[0057] At the application level, this step is suitable for tracing the source of water pollution. When multiple pollution sources belong to the same industry and their three-dimensional fluorescent fingerprints are highly similar, the introduction of a component dimension enhances the fingerprint's distinguishing ability. For example, in industries with similar organic emissions characteristics, such as papermaking and dyeing, four-dimensional fluorescent fingerprints can effectively identify the emission characteristics of different enterprises. By screening effective components using the fluorescence integral method, data redundancy can be reduced, and computational efficiency and model stability can be improved.

[0058] From a technical perspective, this step quantifies the fluorescence intensity of components to reduce noise and extract features from four-dimensional fluorescence fingerprint data, thereby improving the representativeness and comparability of the fingerprint data. After removing components with low integral values, the remaining effective components can more accurately reflect the chemical composition characteristics of the pollution source, providing a high-quality data foundation for subsequent similarity weighted calculations and significantly enhancing the resolution and reliability of pollution source tracing.

[0059] Furthermore, S4 includes: S41, Calculate the fluorescence integral value of each component based on the processed four-dimensional fluorescence fingerprint data of the target water body section. The calculation formula is as follows:

[0060] in, Indicates in the sample The In the water fluorescence fingerprint of the components, the first... The first excitation wavelength at the _th The fluorescence intensity corresponding to each emission wavelength n The total number of emission wavelengths, m This represents the total number of laser wavelengths. S42, Remove Samples The fluorescence integral value is less than The component data, among which, a To set the weights, the values ​​range from [0,1]. After the elimination is completed, the data are sorted by component number to obtain the four-dimensional fluorescence fingerprint data of the effective components.

[0061] Specifically, the core of this step lies in extracting effective component data from the four-dimensional fluorescent fingerprint using the fluorescence integration method, thereby improving the accuracy and reliability of subsequent pollution source similarity calculations. In some implementations, this step first targets the sample. Each component Calculate its fluorescence integral value. ,in Indicates sample The The components in the first The first excitation wavelength at the _th The integral value represents the fluorescence intensity corresponding to each emission wavelength. This integral value reflects the total fluorescence response intensity of the component under all excitation-emission wavelength combinations and is an important indicator for measuring its information contribution in four-dimensional fingerprints.

[0062] At the parameter level, the calculation of the integral value depends on the excitation wavelength. quantity and emission wavelength quantity These two parameters are typically determined by the scanning settings of the three-dimensional fluorescence spectrometer. For example, in conventional water quality fluorescence fingerprint testing, the excitation wavelength range is usually set to... nm, step size is nm, emission wavelength range is nm, step size is nm, therefore and Can achieve respectively and The result of the integral calculation will form a one-dimensional vector. Each element corresponds to the total fluorescence intensity of a component.

[0063] Furthermore, to eliminate interference from low-signal components in subsequent analyses, a threshold screening mechanism is introduced in this step. Specifically, a scaling factor is set. And calculate the maximum value among all component integrals. Then set the threshold to All components with integral values ​​below this threshold will be discarded, retaining only valid components with values ​​above the threshold. This threshold can be adjusted according to the specific application scenario; for example, in pollution source tracing tasks requiring high sensitivity, Can be set to However, in situations where the signal is strong or the background interference is minimal, It can be appropriately increased to above.

[0064] In practical applications, this step is typically performed after four-dimensional fluorescence fingerprint data acquisition and background subtraction, and is applicable to target water body sections. Upstream section and multiple suspected sources of pollution Data processing. By removing low-signal components, noise interference can be effectively reduced, the signal-to-noise ratio of fingerprint data can be improved, thereby enhancing the discriminative ability of subsequent similarity calculations.

[0065] The technical advantage of this step lies in its ability to reduce noise and simplify four-dimensional fluorescence fingerprint data by quantifying the sum of the fluorescence intensities of the components and setting a dynamic threshold, thereby improving the accuracy and computational efficiency of pollution source identification. Its innovation lies in combining integral screening with a dynamic scaling factor. This makes the method adaptable and robust under different water quality conditions.

[0066] S5. Based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source, calculate the weight vector of each component of the target water body section and the suspected pollution source respectively. Specifically, in this step, a weight vector is calculated based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data. The process of weighting the similarity of water quality fluorescence fingerprints for each component using this weight vector is a key step in achieving accurate pollution source identification. This weight vector includes the weight vector of the target water body section and the weight vector of suspected pollution sources. The core of this step lies in improving the sensitivity and discriminative power of similarity calculation by quantifying the contribution of each component to the overall fingerprint.

[0067] At the technical implementation level, each element in the weight vector... From the formula The calculation yielded the result.

[0068] At the parameter level, the calculation of the weight vector depends on the fluorescence integral value. Distribution characteristics. Summation calculation. Used for normalization processing to ensure weight values Falling Within the range. This normalization method conforms to the standardization conventions in spectral data processing and helps to eliminate interference caused by differences in fluorescence intensity between different components.

[0069] At the application level, this step is typically performed after the acquisition of four-dimensional fluorescent fingerprints and the screening of valid data. This method is particularly suitable for situations where multiple companies in the same industry have similar emission characteristics, as it enhances the ability to identify subtle differences by introducing component weights.

[0070] In terms of technical effectiveness, this step effectively improves the accuracy and robustness of similarity calculation by introducing a dynamic weighting mechanism based on fluorescence integral values. Compared with traditional equal-weighting methods, this method can highlight the characteristics of high fluorescence intensity components, thereby achieving more precise matching in pollution source comparison and providing a more reliable basis for subsequent pollution source tracing.

[0071] The technical advantage of this step lies in its introduction of a dynamic weighting mechanism, which focuses similarity calculations more on components with significant fluorescence characteristics, thereby enhancing the ability to identify pollution source-specific components in fingerprint comparison. Especially when the similarity of fluorescent fingerprints is high between different companies in the same industry, this method can effectively improve discrimination and achieve more accurate pollution source tracing. Furthermore, this step has good scalability, is applicable to comparative analysis of different water quality samples and pollution sources, and possesses high practical value and engineering feasibility.

[0072] S6. Based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after removing the background signal, the component-level similarity is calculated, and the comprehensive similarity of each pollution source is obtained. The pollution source attribution is determined according to the comprehensive similarity.

[0073] Specifically, the core of this step lies in achieving precise source tracing of pollutants in the target water body section through component-level similarity calculation. Specifically, this step compares the effective four-dimensional fluorescent fingerprint data of suspected pollution sources with the background-removed target water body section data component by component, and obtains the comprehensive similarity of each pollution source through weighted summation, thereby determining the attribution of the pollution source.

[0074] At the technical implementation level, the four-dimensional fluorescence fingerprint data was obtained by coupling high-performance liquid chromatography (HPLC) with three-dimensional fluorescence spectroscopy (EEM). Each component... Corresponding to an EEM diagram Its structure is , indicating the first In the component, the The excitation wavelength and the first The fluorescence intensity corresponding to each emission wavelength. In this step, the target cross-section is first analyzed. Compared with the unpolluted section upstream Background subtraction is performed on the data to obtain the target cross-sectional data after background subtraction. Subsequently, the suspected source of pollution was investigated. The data was used to screen for effective components, and fluorescence integral values ​​were removed. Less than the threshold The components retain the effective components and form .

[0075] The similarity calculation function is used to quantize the first The similarity between the EEM maps of the pollution source and the target section within each component. This function can be implemented using methods such as Pearson correlation coefficient, cosine similarity, or normalized cross-correlation, with the specific choice depending on the distribution characteristics of the actual data and the required comparison accuracy.

[0076] Weighting coefficient Based on the fluorescence integral value of this component The ratio of the maximum integral value among all effective components is determined by the ratio of the integral value of the component to the maximum integral value of all effective components. This weight reflects the contribution of each component to the overall fingerprint, thereby improving the accuracy and representativeness of similarity calculation.

[0077] The four-dimensional fluorescent fingerprint-based detection and pollution source tracing method of this invention enables accurate identification of pollution sources from different enterprises in the same industry, improving the resolution and accuracy of water environment pollution source tracing.

[0078] Furthermore, S6 includes: S61, Component-level similarity is calculated based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after background signal removal. The calculation formula is as follows:

[0079] in, This indicates the first of the four-dimensional fluorescent fingerprint data of the suspected pollution source. The water quality fluorescence fingerprints of each component and the four-dimensional fluorescence fingerprint data of the target water body section are shown in the figure. The similarity of water quality fluorescent fingerprints of individual components, The fourth fluorescent fingerprint data representing the effective components of the suspected pollution source Water fluorescence fingerprints of individual components The fourth fluorescence fingerprint data representing the effective components of the target water body section. Water fluorescence fingerprints of individual components This indicates that similarity calculations are performed on the two selected water quality fluorescent fingerprint data. S62, calculate the comprehensive similarity between the target water body cross-section sample and each pollution source. The calculation formula is as follows:

[0080] in, Indicates pollution source With the target section Similarity of four-dimensional fluorescent fingerprints.

[0081] In practical applications, this step is suitable for water pollution source tracing scenarios, especially when multiple pollution sources in the same industry have high similarity in fluorescent fingerprints and are difficult to distinguish. By introducing a component dimension and an integral weighting mechanism, this method significantly improves the resolution and source tracing accuracy of fingerprint comparison, providing reliable technical support for environmental supervision.

[0082] To demonstrate the accuracy of the four-dimensional fluorescent fingerprint-based detection and pollution source tracing proposed in this invention in identifying water pollution sources, specific examples are provided below to illustrate the invention.

[0083] Pollution signals were detected at a certain surface water section, and its water quality fluorescent fingerprint was as follows: Figure 3 As shown. By comparing with local pollution sources, the pollution signal was determined to originate from dye wastewater, focusing on dye companies A and B. Existing methods were used to detect the fluorescent fingerprints of the wastewater from both companies, yielding the following results: Figure 4 and Figure 5 As shown.

[0084] The similarity of the surface water quality fluorescent fingerprints of the cross-sections with those of companies A and B was compared. The results are shown in Table 1. It can be seen that the similarity between company A and company B in their water quality fluorescent fingerprints is only 2.4%, which is quite close and makes it difficult to further determine the source of surface water pollution.

[0085] Table 1. Similarity between surface water cross-section fluorescence fingerprints and the fluorescence fingerprints of wastewater from companies A and B.

[0086] Furthermore, the detection method based on four-dimensional fluorescent fingerprinting proposed in this invention was used to detect samples from surface water sections, wastewater samples from Company A, and wastewater samples from Company B, respectively. After subtracting the upstream background according to the aforementioned steps, the results were then processed according to... A value of 0.01 was used to filter valid data. Finally, water quality fluorescence fingerprints of five effective components were obtained from the four-dimensional fluorescence fingerprint of each sample. These components were labeled as component A, component B, component C, component D, and component E in order of elution time. (See attached image) Figure 6-8 .

[0087] In the four-dimensional fluorescence fingerprint of surface water, the fluorescence integral values ​​of the six effective components are 9858849, 11065326, 23674730, 14428064, and 6222033, respectively, and the weight vector is [0.151, 0.170, 0.363, 0.221, 0.095].

[0088] Table 2 shows the similarity between the effective components in the four-dimensional fluorescent fingerprints of Company A and Company B and the corresponding components in the four-dimensional fluorescent fingerprints of the surface water section.

[0089] Table 2. Similarity of water quality fluorescent fingerprints of various components from companies A and B with those from surface water.

[0090] The similarity between the surface water section and the four-dimensional fluorescent fingerprint of Company A and the surface water section and the four-dimensional fluorescent fingerprint of Company B were calculated by combining the weight vectors, as shown in Table 3.

[0091] Table 3. Similarity between surface water and the four-dimensional fluorescent fingerprints of Company A and Company B

[0092] Therefore, it can be seen that four-dimensional fluorescent fingerprints can be used to further distinguish pollution sources with similar water quality fluorescent fingerprints, thus achieving precise identification of water pollution sources.

[0093] Example 2 To achieve the above embodiments, such as Figure 2 As shown, this embodiment also provides a detection and pollution source tracing device 10 based on four-dimensional fluorescent fingerprints, including: The four-dimensional fluorescent fingerprint acquisition module 100 is used to test the four-dimensional fluorescent fingerprints of target water body cross-section samples, unpolluted cross-section samples upstream of the target water body, and suspected pollution source samples. Background signal subtraction module 200 is used to subtract background signals from the four-dimensional fluorescent fingerprint data of the target water body section and the four-dimensional fluorescent fingerprint of the suspected pollution source sample based on the four-dimensional fluorescent fingerprint data of the unpolluted section upstream of the target water body, so as to obtain the four-dimensional fluorescent fingerprint data of the target water body section and the four-dimensional fluorescent fingerprint data of the suspected pollution source sample after subtracting the background value. The effective component screening module 300 is used to screen effective four-dimensional fluorescence fingerprint data using the fluorescence integration method, calculate the fluorescence integral value of each component, and remove component data with fluorescence integral values ​​lower than the set threshold according to the set threshold, so as to form four-dimensional fluorescence fingerprint data of effective components of target water body section and suspected pollution source. The weighted calculation module 400 calculates the weight vectors of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source based on the fluorescence integral values ​​of each component in the target water body section and the suspected pollution source. The pollution source attribution determination module 500 performs component-level similarity calculations based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after removing the background signal, and obtains the comprehensive similarity of each pollution source, and determines the pollution source attribution based on the comprehensive similarity.

[0094] Furthermore, the four-dimensional fluorescent fingerprint acquisition module 100 is also used for: The sample was separated into q ordered components using a high-performance liquid chromatography column. The water quality fluorescence fingerprint was tested using a three-dimensional fluorescence spectroscopy detection module for each component, and the water quality fluorescence fingerprints of each component were arranged in the order of the components to form a four-dimensional tensor data structure.

[0095] Furthermore, the background signal subtraction module 200 is also used for: The i-th excitation wavelength of the water quality fluorescence fingerprint of each component of the target water body section. j Fluorescence intensity corresponding to each emission wavelength Subtract the fluorescence intensity at the same location on the upstream section. ,get .

[0096] Where T represents the cross-section of the water body. This indicates that the upstream section was not polluted. k Indicates the sample group number. i Indicates the wavelength number of the excitation light. j Indicates the emission wavelength number.

[0097] Furthermore, the effective component screening module 300 is also used for: Calculate the fluorescence integral value for each component. And based on the threshold Component data with integral values ​​below a threshold are removed, where weights are set. The range of values ​​is , Indicates in the sample The In the water fluorescence fingerprint of the components, the first... The first excitation wavelength at the _th The corresponding fluorescence intensity, sample X includes any one or more of the following: water body cross-section sample, unpolluted cross-section sample upstream of the target water body, and suspected pollution source sample.

[0098] Furthermore, the weighted calculation module 400 is also used for: Weight vector Weights of each component It is applied to the similarity calculation of corresponding components, highlighting the characteristic contribution of components with high fluorescence intensity.

[0099] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described detection and pollution source tracing method based on four-dimensional fluorescent fingerprints.

[0100] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0101] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for detection and pollution source tracing based on four-dimensional fluorescent fingerprinting, characterized in that, include: S1 uses water quality fluorescent fingerprint tracing technology to identify the source of water pollution, clarify the industry of pollution source, and identify enterprises involved in suspected pollution industries as suspected pollution sources. S2, collect samples from the target water body section, the unpolluted section upstream of the target water body, and the suspected pollution source, and test the four-dimensional fluorescent fingerprint of each sample; S3, Based on the four-dimensional fluorescence fingerprint data of the upstream unpolluted section, background signal subtraction is performed on the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source, respectively. S4. After removing the background signal, calculate the fluorescence integral value of each component in the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source. Remove component data with fluorescence integral values ​​lower than the threshold according to the set threshold to form four-dimensional fluorescence fingerprint data of the effective components of the target water body section and the suspected pollution source. S5. Based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source, calculate the weight vector of each component of the target water body section and the suspected pollution source respectively. S6. Based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after removing the background signal, the component-level similarity is calculated, and the comprehensive similarity of each pollution source is obtained. The pollution source attribution is determined according to the comprehensive similarity.

2. The method as described in claim 1, characterized in that, The four-dimensional fluorescent fingerprint includes: components, excitation wavelength, emission wavelength, and fluorescence intensity.

3. The method as described in claim 1, characterized in that, Step S2 further includes: S21 uses a liquid chromatography column to process the sample and separate the ordered components in the sample; S22, obtain the water quality fluorescence fingerprints of each component of the sample, and arrange the water quality fluorescence fingerprints of each component in the order of the components to form the four-dimensional fluorescence fingerprint of the sample.

4. The method as described in claim 1, characterized in that, Step S22 further includes: A three-dimensional fluorescence spectroscopy detection module was used to test the fluorescent fingerprint of water quality.

5. The method as described in claim 1, characterized in that, Step S3 further includes: By subtracting the fluorescence intensity of the same component in the unpolluted section sample upstream of the target water body from the fluorescence intensity of each component in the target water body section sample, the four-dimensional fluorescence fingerprint data of the target water body section sample after removing the background signal is obtained. By subtracting the fluorescence intensity of the same component in the unpolluted section upstream of the target water body from the fluorescence intensity of each component in the suspected pollution source sample, four-dimensional fluorescence fingerprint data of the suspected pollution source after removing background signals is obtained.

6. The method as described in claim 1, characterized in that, Step S4 further includes: S41, Calculate the fluorescence integral value of each component based on the processed four-dimensional fluorescence fingerprint data of the target water body section. The calculation formula is as follows: in, Indicates in the sample The In the water fluorescence fingerprint of the components, the first... The first excitation wavelength at the _th The fluorescence intensity corresponding to each emission wavelength n The total number of emission wavelengths, m This represents the total number of laser wavelengths. S42, Remove Sample The fluorescence integral value is less than The component data, among which, a To set the weights, the values ​​range from [0,1]. After the elimination is completed, the data are sorted by component number to obtain the four-dimensional fluorescence fingerprint data of the effective components.

7. The method as described in claim 1, characterized in that, Step S5 further includes: S51, based on the fluorescence integral values ​​of each component in the effective four-dimensional fluorescence fingerprint data, calculate the weight vector of each component. The calculation formula is as follows: in, For the sample X The first of the effective four-dimensional fluorescent fingerprint data Fluorescence integral values ​​of each component.

8. The method as described in claim 1, characterized in that, Step S6 further includes: S61, Component-level similarity is calculated based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after background signal removal. The calculation formula is as follows: in, This indicates the first of the four-dimensional fluorescent fingerprint data of the suspected pollution source. The water quality fluorescence fingerprints of each component and the four-dimensional fluorescence fingerprint data of the target water body section are shown in the figure. The similarity of water quality fluorescent fingerprints of individual components, The fourth fluorescent fingerprint data representing the effective components of the suspected pollution source Water fluorescence fingerprints of individual components The fourth fluorescence fingerprint data representing the effective components of the target water body section. Water fluorescence fingerprints of individual components This indicates that similarity calculations are performed on the two selected water quality fluorescent fingerprint data. S62, calculate the comprehensive similarity between the target water body cross-section sample and each pollution source. The calculation formula is as follows: in, Indicates pollution source With the target section Similarity of four-dimensional fluorescent fingerprints.

9. A detection and pollution source tracing device based on four-dimensional fluorescent fingerprinting, characterized in that, include: The four-dimensional fluorescent fingerprint acquisition module is used to test the four-dimensional fluorescent fingerprints of target water body cross-section samples, unpolluted cross-section samples upstream of the target water body, and samples from suspected pollution sources. The background signal subtraction module is used to subtract the background signal from the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint of the suspected pollution source sample based on the four-dimensional fluorescence fingerprint data of the unpolluted section upstream of the target water body, so as to obtain the four-dimensional fluorescence fingerprint data of the target water body section and the four-dimensional fluorescence fingerprint data of the suspected pollution source sample after deducting the background value. The effective component screening module is used to screen effective four-dimensional fluorescence fingerprint data using the fluorescence integration method, calculate the fluorescence integral value of each component, and remove component data with fluorescence integral values ​​lower than the set threshold according to the set threshold, so as to form four-dimensional fluorescence fingerprint data of effective components of target water body section and suspected pollution source. The weighted calculation module calculates the weight vectors of each component in the effective four-dimensional fluorescence fingerprint data of the target water body section and the suspected pollution source, based on the fluorescence integral values ​​of each component in the target water body section and the suspected pollution source. The pollution source attribution determination module performs component-level similarity calculations based on the four-dimensional fluorescence fingerprint data of the suspected pollution source and the target water body section after removing the background signal, and obtains the comprehensive similarity of each pollution source. The pollution source attribution is determined based on the comprehensive similarity.

10. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method as claimed in any one of claims 1-8.

Images