A method and device for multi-dimensional source tracing of water pollution

Abstract

This application relates to a method and apparatus for multi-dimensional source tracing of water pollution sources. The method includes: performing cluster analysis on the collected fluorescent fingerprints of the water body to be traced and fingerprints in a typical fluorescent fingerprint library in a pre-set multi-dimensional fluorescent fingerprint database to obtain cluster analysis results. If no suspected pollution sources are found in the cluster analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprints of the water body to be traced, thereby determining the suspected pollution sources of the water body to be traced. This solves the problems of inability to simultaneously trace multiple pollution sources, lack of standardized procedures for sample and fluorescent fingerprint collection, low efficiency, and poor timeliness in related technologies. By standardizing samples and fluorescent fingerprints and adopting online multi-pollution source tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the source tracing method, and the reliability and stability of the source tracing results are improved.

Description

Technical Field

[0001] This invention relates to the field of water environment monitoring technology, and in particular to a method and device for tracing water pollution sources from multiple dimensions. Background Technology

[0002] Currently, pollution source monitoring mainly relies on installing conventional online monitoring equipment at discharge outlets for water quality (COD (Chemical Oxygen Demand), total nitrogen, total phosphorus, etc.). This method has low sensitivity and is difficult to monitor illegal or unauthorized discharges. Source tracing often requires manual investigation, which is time-consuming, labor-intensive, inefficient, and lacks timeliness. Three-dimensional fluorescence spectroscopy analysis, on the other hand, features high sensitivity, simple and rapid testing, and no secondary pollution. It can also reveal the organic composition of wastewater, exhibiting "fingerprint" characteristics, hence the name fluorescent fingerprint. By comparing the similarity between the fluorescent fingerprints of polluted water bodies and pollution sources, rapid source identification can be achieved. Therefore, fluorescent fingerprinting has significant potential in water pollution source tracing.

[0003] In related technologies, the following methods are mainly used to achieve rapid identification of pollution sources by comparing the similarity of fluorescent fingerprints of polluted water bodies and pollution sources:

[0004] (1) A method for tracing the source of water pollution, which first uses online biological detection and early warning technology to identify potential pollution sources, then compares the three-dimensional fluorescence spectra of the pollution sources with those of the polluted water, and then uses gas chromatography-mass spectrometry and liquid chromatography-mass spectrometry for further analysis.

[0005] (2) A pollution source apportionment method, which first collects the three-dimensional fluorescence spectrum of the original sample, and then uses an electrochemical method to disturb the water sample before testing the three-dimensional fluorescence spectrum of the water sample.

[0006] (3) A method for tracing and classifying polluted water bodies using three-dimensional fluorescence spectroscopy. This method requires extracting various features of the fluorescence spectrum, such as peak center coordinates, peak intensity, peak intensity ratio, half-peak width, peak spacing, angle, slope, area, and three-dimensional structure.

[0007] (4) A method for rapid identification and comparison using fluorescence spectral feature information, which requires 34 steps of processing of spectral data.

[0008] However, methods (1) and (2) mentioned above require numerous instruments, are expensive, and involve complex processes, making online source tracing difficult and time-consuming. Furthermore, they do not consider the influence of factors such as concentration and pH on fluorescent fingerprints, thus failing to accurately obtain the fluorescent fingerprints of pollution sources and affecting the accuracy of pollution source tracing. Methods (3) and (4) only utilize a single instrument, three-dimensional fluorescence spectroscopy, and perform some preprocessing on the sample or fluorescence spectrum, but still lack a standardized procedure for fluorescent fingerprint collection and processing, hindering widespread application. In addition, the four rapid pollution source identification methods mentioned above can only trace a single pollution source. In real-world situations where multiple pollution sources contaminate water bodies, simultaneous tracing of multiple pollution sources is often impossible. Therefore, a standardized fluorescent fingerprint tracing technology for multiple pollution sources is urgently needed. Summary of the Invention

[0009] This application provides a method and apparatus for tracing water pollution sources from multiple dimensions, in order to solve the problems of related technologies being unable to trace multiple pollution sources simultaneously, lacking standardized procedures for sample and fluorescent fingerprint collection, and having low efficiency and poor timeliness in the tracing methods.

[0010] The first aspect of this application provides a multi-dimensional comparison and identification method for pollution sources, including the following steps:

[0011] Collect the water body to be traced, and perform cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprint in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results;

[0012] If no suspected pollution source is found in the cluster analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced; and

[0013] The suspected pollution sources of the water body to be traced are determined based on the optimal pollution source determination strategy.

[0014] According to one embodiment of this application, the step of performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint library of a preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results includes:

[0015] If only one typical fluorescent fingerprint of a pollution source exists in the typical fluorescent fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is regarded as the suspected pollution source.

[0016] If the typical fluorescent fingerprints of multiple pollution sources are clustered with the fluorescent fingerprints of the water body to be traced in the typical fluorescent fingerprint database, then the similarity between the typical fluorescent fingerprints of the multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0017] If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0018] According to one embodiment of this application, the optimal pollution source determination strategy based on matching the number of fluorescence peaks and the wavelength of fluorescence peaks in the fluorescence fingerprint of the water body to be traced includes:

[0019] If there is one fluorescence peak or multiple fluorescence peaks with the same emission wavelength in the fluorescent fingerprint of the water body to be traced, it indicates that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0020] If there are multiple fluorescence peaks with different emission wavelengths in the fluorescent fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is less than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database.

[0021] If the fluorescent fingerprint of the water body to be traced contains multiple fluorescence peaks with different emission wavelengths, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is larger than the preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the fluorescence components of the fluorescent fingerprint of the water body to be traced and the fluorescence component library in the preset multi-dimensional fluorescent fingerprint database.

[0022] According to one embodiment of this application, the optimal pollution source determination strategy involves performing cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database. Determining the suspected pollution source of the water body to be traced based on the optimal pollution source determination strategy includes:

[0023] If only one pollution source's sub-fingerprint exists in the dynamic sub-fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is taken as the suspected pollution source.

[0024] If multiple pollution source sub-fingerprints are clustered with the fluorescent fingerprints of the water body to be traced in the dynamic sub-fingerprint database, then the similarity between the multiple pollution source sub-fingerprints and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0025] If the suspected pollution sources of all sub-fluorescent fingerprints of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources.

[0026] If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0027] According to one embodiment of this application, the optimal pollution source determination strategy involves performing cluster analysis based on the fluorescent components of the fluorescent fingerprint of the water body to be traced and the fluorescent component library in the preset multi-dimensional fluorescent fingerprint database. Determining the suspected pollution source of the water body to be traced based on the optimal pollution source determination strategy includes:

[0028] If the fluorescent component library contains only one pollution source whose fluorescent component clusters with the fluorescent fingerprint of the water body to be traced, then the only pollution source is considered the suspected pollution source.

[0029] If the fluorescent components stored in the fluorescent components of multiple pollution sources are clustered with the fluorescent components of the fluorescent fingerprint of the water body to be traced, then the similarity between the fluorescent components of the multiple pollution sources and the fluorescent components of the fluorescent fingerprint of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0030] If the suspected pollution sources of all fluorescent components of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources.

[0031] If the fluorescent components of the fluorescent fingerprint of the water body to be traced do not cluster with the fluorescent components of any pollution source in the fluorescent component database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0032] According to one embodiment of this application, before performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database, the method further includes:

[0033] Collect pollution source samples, water body samples to be traced, clean water body samples, and ultrapure water samples, and perform standardized processing on the pollution source samples, the water body samples to be traced, and the clean water body samples;

[0034] Fluorescent fingerprints of the ultrapure water sample and the standardized pollutant source sample, the water body sample to be traced, and the clean water sample were collected respectively, as well as ultraviolet-visible absorption spectra of the standardized pollutant source sample, the water body sample to be traced, and the clean water sample.

[0035] Based on a preset standardized processing strategy, the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water body sample to be traced are standardized to obtain the fluorescent fingerprints of the target pollution source sample and the target water body sample to be traced.

[0036] Based on a pre-defined parallel factor analysis method, the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample are obtained, and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced are also obtained, so as to construct a fluorescence component library based on the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced.

[0037] At least one typical fluorescent fingerprint of a pollution source sample is selected from the fluorescent fingerprints of the target pollution source sample, and at least one typical fluorescent fingerprint of a water body sample to be traced is selected from the fluorescent fingerprints of the target water body sample to be traced, so as to construct a typical fluorescent fingerprint library based on the typical fluorescent fingerprints of the at least one pollution source sample and the at least one water body sample to be traced.

[0038] Based on the composition characteristics of the fluorescence peaks of the fluorescence fingerprints of the water samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library, wherein each sub-fluorescent fingerprint contains at least one fluorescence peak.

[0039] The multi-dimensional fluorescent fingerprint database is constructed based on the fluorescent component library, the typical fluorescent fingerprint library, and the dynamic sub-fingerprint library.

[0040] According to one embodiment of this application, the standardization processing of the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced, based on a preset standardization processing strategy, includes:

[0041] Blank corrections were performed on the fluorescent fingerprints of the pollution source sample and the water sample to be traced, respectively, to obtain the fluorescent fingerprints of the pollution source sample after blank correction and the fluorescent fingerprints of the water sample to be traced after blank correction.

[0042] Based on a preset correction formula, the fluorescence fingerprint of the pollution source sample after blank correction is corrected by using the ultraviolet-visible absorption spectrum of the pollution source sample to obtain the fluorescence fingerprint of the pollution source sample after internal filtration correction. The fluorescence fingerprint of the water body sample to be traced is corrected by using the ultraviolet-visible absorption spectrum of the water body sample to be traced by internal filtration to obtain the fluorescence fingerprint of the water body sample to be traced by internal filtration correction.

[0043] According to a preset replacement strategy, the first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the pollution source sample after internal filtration correction are replaced. According to a first data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the pollution source sample after scattering correction. The first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the water body sample to be traced after internal filtration correction are replaced. According to a second data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the water body sample to be traced after scattering correction.

[0044] Based on the integral of the preset Raman scattering intensity of ultrapure water, the data units of the fluorescence fingerprints of the pollution source sample after scattering correction and the fluorescence fingerprints of the water sample to be traced after scattering correction are converted from dimensionless to Raman units.

[0045] According to one embodiment of this application, the blank correction of the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced includes:

[0046] The fluorescence fingerprint of the pollution source sample is obtained by subtracting the fluorescence fingerprint of the ultrapure water sample from the fluorescence fingerprint numerical matrix of the pollution source sample;

[0047] The fluorescent fingerprint of the clean water sample is obtained by subtracting the fluorescent fingerprint value matrix of the water sample to be traced from the fluorescent fingerprint value matrix of the water sample to be traced. The blank-corrected fluorescent fingerprint of the water sample to be traced is then obtained.

[0048] According to one embodiment of this application, the preset correction formula is:

[0049]

[0050] Among them, F corr F represents the corrected fluorescence intensity. obs The fluorescence intensity before correction, A ex To determine the absorbance at the excitation wavelength, A em ν is the absorbance at the emission wavelength.

[0051] According to one embodiment of this application, the preset parallel factor analysis method is as follows:

[0052]

[0053] Where, x ijk The fluorescence intensity of the i-th sample at emission wavelength j and excitation wavelength k; F represents the number of components; a if b jf c kfThese are elements in load matrices A, B, and C, respectively; ε ijk This represents the model residuals.

[0054] According to one embodiment of this application, a preset clustering algorithm is used for cluster analysis, wherein the preset clustering algorithm includes at least one of K-means clustering, density-based noise applied spatial clustering, hierarchical clustering, and self-organizing neural networks.

[0055] According to one embodiment of this application, a preset similarity measurement method is used to calculate similarity, wherein the preset similarity measurement method includes at least one of Kofsky distance, Manhattan distance, Euclidean distance, Chebyshev distance, cosine similarity, Pearson correlation coefficient, Mahalanobis distance, Hamming distance and Jaccard similarity coefficient.

[0056] According to the multi-dimensional comparison and identification method for pollution sources in this application, the fluorescent fingerprints of the water body to be traced are clustered with fingerprints in a typical fluorescent fingerprint library in a preset multi-dimensional fluorescent fingerprint database to obtain clustering analysis results. If no suspected pollution sources are found in the clustering analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprints of the water body to be traced, thereby determining the suspected pollution sources of the water body to be traced. This solves the problems in related technologies, such as the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods. By standardizing samples and fluorescent fingerprints and adopting online multi-pollution tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing methods, and the reliability and stability of the tracing results are improved.

[0057] A second aspect of this application provides a multi-dimensional pollution source tracing device, comprising:

[0058] Metering pump;

[0059] The filtration module is used to filter the water sample delivered by the metering pump to obtain a filtered water sample.

[0060] A water quality adjustment module is used to adjust the pH value of the filtered water sample to obtain a test water sample;

[0061] A fingerprint acquisition module is used to acquire fluorescent fingerprints of the test water sample;

[0062] The control system module stores a computer program. When the computer program is executed by the control system module, it realizes the multi-dimensional comparison and identification method of pollution sources as described in any one of claims 1-11 based on the fluorescent fingerprint of the test water sample.

[0063] According to one embodiment of this application, the above-mentioned multi-dimensional pollution source tracing device further includes:

[0064] The display module is used to display the test results and pollution source comparison results output by the control system module.

[0065] According to the pollution source multi-dimensional tracing device of this application embodiment, the fluorescent fingerprint of the water body to be traced is clustered with fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the clustering analysis results. If no suspected pollution source is found in the clustering analysis results, the optimal pollution source determination strategy is matched according to the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprint of the water body to be traced, thereby determining the suspected pollution source of the water body to be traced. This solves the problems of inability to simultaneously trace multiple pollution sources, lack of standardized procedures for sample and fluorescent fingerprint collection, low efficiency and poor timeliness in related technologies. By standardizing samples and fluorescent fingerprints and adopting multi-pollution online tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing method, and the reliability and stability of the tracing results are improved.

[0066] A third aspect of this application provides a multi-dimensional comparison and identification device for pollution sources, comprising:

[0067] The acquisition module is used to acquire the water body to be traced and perform cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprint in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results.

[0068] The matching module is used to match the optimal pollution source determination strategy based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced if no suspected pollution source is found in the cluster analysis results; and

[0069] The identification module is used to identify the suspected pollution source of the water body to be traced based on the optimal pollution source determination strategy.

[0070] According to one embodiment of this application, the acquisition module is specifically used for:

[0071] If only one typical fluorescent fingerprint of a pollution source exists in the typical fluorescent fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is regarded as the suspected pollution source.

[0072] If the typical fluorescent fingerprints of multiple pollution sources are clustered with the fluorescent fingerprints of the water body to be traced in the typical fluorescent fingerprint database, then the similarity between the typical fluorescent fingerprints of the multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0073] If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0074] According to one embodiment of this application, the matching module is specifically used to: if there is one fluorescence peak or multiple fluorescence peaks with the same emission wavelength in the fluorescent fingerprint of the water body to be traced, it indicates that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database;

[0075] If there are multiple fluorescence peaks with different emission wavelengths in the fluorescent fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is less than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database.

[0076] If the fluorescent fingerprint of the water body to be traced contains multiple fluorescence peaks with different emission wavelengths, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is larger than the preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the fluorescence components of the fluorescent fingerprint of the water body to be traced and the fluorescence component library in the preset multi-dimensional fluorescent fingerprint database.

[0077] According to one embodiment of this application, the identification module is specifically used for:

[0078] If only one pollution source's sub-fingerprint exists in the dynamic sub-fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is taken as the suspected pollution source.

[0079] If multiple pollution source sub-fingerprints are clustered with the fluorescent fingerprints of the water body to be traced in the dynamic sub-fingerprint database, then the similarity between the multiple pollution source sub-fingerprints and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0080] If the suspected pollution sources of all sub-fluorescent fingerprints of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources.

[0081] If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0082] According to one embodiment of this application, the identification module is specifically used for:

[0083] If the fluorescent component library contains only one pollution source whose fluorescent component clusters with the fluorescent fingerprint of the water body to be traced, then the only pollution source is considered the suspected pollution source.

[0084] If the fluorescent components stored in the fluorescent components of multiple pollution sources are clustered with the fluorescent components of the fluorescent fingerprint of the water body to be traced, then the similarity between the fluorescent components of the multiple pollution sources and the fluorescent components of the fluorescent fingerprint of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0085] If the suspected pollution sources of all fluorescent components of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources.

[0086] If the fluorescent components of the fluorescent fingerprint of the water body to be traced do not cluster with the fluorescent components of any pollution source in the fluorescent component database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

[0087] According to one embodiment of this application, before performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database, the acquisition module is further configured to:

[0088] Collect pollution source samples, water body samples to be traced, clean water body samples, and ultrapure water samples, and perform standardized processing on the pollution source samples, the water body samples to be traced, and the clean water body samples;

[0089] Fluorescent fingerprints of the ultrapure water sample and the standardized pollutant source sample, the water body sample to be traced, and the clean water sample were collected respectively, as well as ultraviolet-visible absorption spectra of the standardized pollutant source sample, the water body sample to be traced, and the clean water sample.

[0090] Based on a preset standardized processing strategy, the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water body sample to be traced are standardized to obtain the fluorescent fingerprints of the target pollution source sample and the target water body sample to be traced.

[0091] Based on a pre-defined parallel factor analysis method, the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample are obtained, and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced are also obtained, so as to construct a fluorescence component library based on the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced.

[0092] At least one typical fluorescent fingerprint of a pollution source sample is selected from the fluorescent fingerprints of the target pollution source sample, and at least one typical fluorescent fingerprint of a water body sample to be traced is selected from the fluorescent fingerprints of the target water body sample to be traced, so as to construct a typical fluorescent fingerprint library based on the typical fluorescent fingerprints of the at least one pollution source sample and the at least one water body sample to be traced.

[0093] Based on the composition characteristics of the fluorescence peaks of the fluorescence fingerprints of the water samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library, wherein each sub-fluorescent fingerprint contains at least one fluorescence peak.

[0094] The multi-dimensional fluorescent fingerprint database is constructed based on the fluorescent component library, the typical fluorescent fingerprint library, and the dynamic sub-fingerprint library.

[0095] According to one embodiment of this application, the acquisition module is specifically used for:

[0096] Blank corrections were performed on the fluorescent fingerprints of the pollution source sample and the water sample to be traced, respectively, to obtain the fluorescent fingerprints of the pollution source sample after blank correction and the fluorescent fingerprints of the water sample to be traced after blank correction.

[0097] Based on a preset correction formula, the fluorescence fingerprint of the pollution source sample after blank correction is corrected by using the ultraviolet-visible absorption spectrum of the pollution source sample to obtain the fluorescence fingerprint of the pollution source sample after internal filtration correction. The fluorescence fingerprint of the water body sample to be traced is corrected by using the ultraviolet-visible absorption spectrum of the water body sample to be traced by internal filtration to obtain the fluorescence fingerprint of the water body sample to be traced by internal filtration correction.

[0098] According to a preset replacement strategy, the first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the pollution source sample after internal filtration correction are replaced. According to a first data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the pollution source sample after scattering correction. The first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the water body sample to be traced after internal filtration correction are replaced. According to a second data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the water body sample to be traced after scattering correction.

[0099] Based on the integral of the preset Raman scattering intensity of ultrapure water, the data units of the fluorescence fingerprints of the pollution source sample after scattering correction and the fluorescence fingerprints of the water sample to be traced after scattering correction are converted from dimensionless to Raman units.

[0100] According to one embodiment of this application, the acquisition module is specifically used for:

[0101] The fluorescence fingerprint of the pollution source sample is obtained by subtracting the fluorescence fingerprint of the ultrapure water sample from the fluorescence fingerprint numerical matrix of the pollution source sample;

[0102] The fluorescent fingerprint of the clean water sample is obtained by subtracting the fluorescent fingerprint value matrix of the water sample to be traced from the fluorescent fingerprint value matrix of the water sample to be traced. The blank-corrected fluorescent fingerprint of the water sample to be traced is then obtained.

[0103] According to one embodiment of this application, the preset correction formula is:

[0104]

[0105] Among them, F corr F represents the corrected fluorescence intensity. obs A represents the fluorescence intensity before correction. ex To determine the absorbance at the excitation wavelength, A em ν is the absorbance at the emission wavelength.

[0106] According to one embodiment of this application, the preset parallel factor analysis method is as follows:

[0107]

[0108] Where, x ijk The fluorescence intensity of the i-th sample at emission wavelength j and excitation wavelength k; F represents the number of components; a if b jf c kf These are elements in load matrices A, B, and C, respectively; ε ijk This represents the model residuals.

[0109] According to one embodiment of this application, a preset clustering algorithm is used for cluster analysis, wherein the preset clustering algorithm includes at least one of K-means clustering, density-based noise applied spatial clustering, hierarchical clustering, and self-organizing neural networks.

[0110] According to one embodiment of this application, a preset similarity measurement method is used to calculate similarity, wherein the preset similarity measurement method includes at least one of Kofsky distance, Manhattan distance, Euclidean distance, Chebyshev distance, cosine similarity, Pearson correlation coefficient, Mahalanobis distance, Hamming distance and Jaccard similarity coefficient.

[0111] According to the multi-dimensional comparison and identification device for pollution sources in this application embodiment, the fluorescent fingerprint of the water body to be traced is clustered with fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain clustering analysis results. If no suspected pollution source is found in the clustering analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprint of the water body to be traced, thereby determining the suspected pollution source of the water body to be traced. This solves the problems in related technologies such as the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods. By standardizing samples and fluorescent fingerprints and adopting multi-pollution online tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing method, and the reliability and stability of the tracing results are improved.

[0112] A fourth aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the multi-dimensional comparison and identification method for pollution sources as described in the above embodiments.

[0113] A fifth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the multi-dimensional comparison and identification method for pollution sources as described in the above embodiments.

[0114] 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

[0115] 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:

[0116] Figure 1 This is a flowchart of a multi-dimensional comparison and identification method for pollution sources according to an embodiment of this application;

[0117] Figure 2 This is a flowchart of a water pollution multi-dimensional source tracing method according to an embodiment of this application;

[0118] Figure 3 This is a schematic diagram of a typical fluorescent fingerprint I of a water body to be traced according to an embodiment of this application;

[0119] Figure 4 This is a typical fluorescent fingerprint II of a water body to be traced according to an embodiment of this application and its decomposition diagram;

[0120] Figure 5 This is a schematic diagram of a multi-dimensional pollution source tracing device according to an embodiment of this application;

[0121] Figure 6 This is a block diagram of a multi-dimensional comparison and identification device for pollution sources according to an embodiment of this application;

[0122] Figure 7 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0123] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0124] The following description, with reference to the accompanying drawings, describes a multi-dimensional source tracing method and apparatus for water pollution according to embodiments of the present invention. Addressing the problems mentioned in the background art regarding the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods, this application provides a multi-dimensional comparison and identification method for pollution sources. In this method, the fluorescent fingerprint of the water body to be traced is clustered with fingerprints in a typical fluorescent fingerprint library of a preset multi-dimensional fluorescent fingerprint database to obtain clustering analysis results. If no suspected pollution source is found in the clustering analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprint of the water body to be traced, thereby determining the suspected pollution source of the water body to be traced. This solves the problems of the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods in related technologies. By standardizing samples and fluorescent fingerprints and adopting online multi-pollution source tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing method, and the reliability and stability of the tracing results are improved.

[0125] Specifically, Figure 1 This is a flowchart of the multi-dimensional comparison and identification method for pollution sources provided in the embodiments of this application.

[0126] like Figure 1 As shown, the multi-dimensional comparison and identification method for this pollution source includes the following steps:

[0127] In step S101, the water body to be traced is collected, and the fluorescent fingerprint of the water body to be traced is clustered with the fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the clustering analysis results.

[0128] Furthermore, in some embodiments, the fluorescent fingerprint of the water body to be traced is clustered with the fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the clustering analysis results, including: if only one pollution source's typical fluorescent fingerprint exists in the typical fluorescent fingerprint library and clusters with the fluorescent fingerprint of the water body to be traced, then the only pollution source is regarded as a suspected pollution source; if multiple pollution sources' typical fluorescent fingerprints exist in the typical fluorescent fingerprint library and cluster with the fluorescent fingerprint of the water body to be traced, then the similarity between the typical fluorescent fingerprints of multiple pollution sources and the fluorescent fingerprint of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is regarded as a suspected pollution source; if the sub-fingerprints of the fluorescent fingerprint of the water body to be traced do not cluster with the sub-fingerprints of any pollution source in the dynamic sub-fingerprint library, then it is determined that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database.

[0129] Specifically, in this embodiment of the application, the fluorescent fingerprint of the water body to be traced is clustered with the fingerprints in the typical fluorescent fingerprint database of pollution sources. The steps are as follows:

[0130] (1) If the fluorescent fingerprint of the water body to be traced clusters with the typical fluorescent fingerprint of a certain pollution source, then the pollution source is a suspected pollution source.

[0131] (2) If the fluorescent fingerprint of the water body to be traced is clustered with the typical fluorescent fingerprints of multiple (≥2) pollution sources, then calculate the similarity between the fluorescent fingerprint of the water body to be traced and the typical fluorescent fingerprints of each pollution source, and the one with the highest similarity is the suspected pollution source.

[0132] (3) If the fluorescent fingerprint of the water body to be traced does not cluster with the typical fluorescent fingerprint of any pollution source, proceed to the next step.

[0133] Furthermore, in some embodiments, before performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint library of a preset multi-dimensional fluorescent fingerprint database, the method further includes: collecting pollution source samples, water body samples to be traced, clean water samples, and ultrapure water samples, and standardizing the pollution source samples, water body samples to be traced, and clean water samples; collecting the fluorescent fingerprints of the ultrapure water samples and the standardized pollution source samples, water body samples to be traced, and clean water samples, as well as the ultraviolet-visible absorption spectra of the standardized pollution source samples, water body samples to be traced, and clean water samples; standardizing the fluorescent fingerprints of the pollution source samples and the water body samples to be traced based on a preset standardization strategy to obtain the fluorescent fingerprints of the target pollution source samples and the target water body samples to be traced; and obtaining the fluorescent components and fluorescence intensity of the fluorescent fingerprints of the target pollution source samples based on a preset parallel factor analysis method. The fluorescence components and intensities of the fluorescence fingerprints of the target water body samples to be traced are collected to construct a fluorescence component library based on the fluorescence components and intensities of the fluorescence fingerprints of the target pollution source samples and the target water body samples to be traced. At least one typical fluorescence fingerprint of a pollution source sample is selected from the fluorescence fingerprints of the target pollution source samples, and at least one typical fluorescence fingerprint of a water body sample to be traced is selected from the fluorescence fingerprints of the target water body samples to be traced, to construct a typical fluorescence fingerprint library based on the typical fluorescence fingerprints of at least one pollution source sample and at least one water body sample to be traced. Based on the compositional characteristics of the fluorescence peaks of the fluorescence fingerprints of the water body samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library, where each sub-fluorescent fingerprint contains at least one fluorescence peak. A multi-dimensional fluorescence fingerprint database is constructed based on the fluorescence component library, the typical fluorescence fingerprint library, and the dynamic sub-fingerprint library.

[0134] Furthermore, in some embodiments, based on a preset standardization processing strategy, the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water body sample to be traced are standardized, including: performing blank correction on the fluorescent fingerprints of the pollution source sample and the water body sample to be traced, respectively, to obtain blank-corrected fluorescent fingerprints of the pollution source sample and blank-corrected fluorescent fingerprints of the water body sample to be traced; based on a preset correction formula, using the UV-Vis absorption spectrum of the pollution source sample, performing internal filtration correction on the blank-corrected fluorescent fingerprint of the pollution source sample to obtain an internally filtration corrected fluorescent fingerprint of the pollution source sample, and using the UV-Vis absorption spectrum of the water body sample to be traced, performing internal filtration correction on the fluorescent fingerprint of the water body sample to be traced to obtain an internally filtration corrected fluorescent fingerprint of the water body sample to be traced; and following a preset replacement strategy, ... The first-order and second-order Rayleigh scattering data of the fluorescence fingerprint of the pollution source sample after internal filtration correction are replaced, and the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled according to the first data processing strategy to obtain the fluorescence fingerprint of the pollution source sample after scattering correction. The first-order and second-order Rayleigh scattering data of the fluorescence fingerprint of the water body sample to be traced after internal filtration correction are replaced, and the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled according to the second data processing strategy to obtain the fluorescence fingerprint of the water body sample to be traced after scattering correction. Based on the preset integral of the Raman scattering intensity of ultrapure water, the data units of the fluorescence fingerprint of the pollution source sample after scattering correction and the fluorescence fingerprint of the water body sample to be traced after scattering correction are converted from dimensionless to Raman units.

[0135] Furthermore, in some embodiments, blank correction is performed on the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced, including: subtracting the fluorescent fingerprint value matrix of the ultrapure water sample from the fluorescent fingerprint value matrix of the pollution source sample to obtain the blank-corrected fluorescent fingerprint of the pollution source sample; and subtracting the fluorescent fingerprint value matrix of the clean water sample from the fluorescent fingerprint value matrix of the water sample to be traced to obtain the blank-corrected fluorescent fingerprint of the water sample to be traced.

[0136] Specifically, such as Figure 2 As shown in the embodiment of this application, before performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database, it is necessary to collect 20 samples from each pollution source discharger, collect 20 samples from the L1 river section, and collect 2 samples from the upstream of the L1 river section as clean water samples, and standardize them, for example, by filtering all samples with a 0.45μm microporous membrane, and then diluting the samples of the pollution source and the water body to be traced with ultrapure water until the absorbance A254 ≤ 0.75 at 254nm. The dilution factor of the clean water samples is consistent with that of the water body samples to be traced. Finally, the pH value of all samples is adjusted to between 6 and 8 with dilute hydrochloric acid or sodium hydroxide solution.

[0137] Furthermore, in this embodiment, three-dimensional fluorescence spectra and UV-Vis absorption spectra of ultrapure water samples and standardized treated pollution sources, water bodies to be traced, clean water bodies, and ultrapure water samples were collected. The absorption spectrum scanning range was 220-700 nm, with a scanning interval of 1 nm. The excitation wavelength scanning range of the three-dimensional fluorescence spectrum was 220-450 nm, the emission wavelength scanning range was 230-550 nm, the scanning interval was 5 nm, and the scanning speed was 12000 nm / min. Each fluorescence fingerprint was a 47×65 matrix, with rows representing the excitation wavelength and columns representing the emission wavelength. Taking a certain sample as an example, the original fluorescence fingerprint data matrix of this sample is as follows:

[0138]

[0139] Furthermore, based on a preset standardization processing strategy, this application embodiment first standardizes the fluorescent fingerprints of the pollution source samples and the fluorescent fingerprints of the water samples to be traced. Specifically, it subtracts the fluorescent fingerprint data matrix of ultrapure water from the fluorescent fingerprint data matrix of all pollution source samples, and subtracts the mean of the fluorescent fingerprint data matrix of clean water samples from the fluorescent fingerprint data matrix of the water samples to be traced. Then, it performs internal filtration correction using ultraviolet-visible absorption spectroscopy. The correction formula is as follows:

[0140]

[0141] Among them, F corr F represents the corrected fluorescence intensity. obs A represents the fluorescence intensity before correction. ex To determine the absorbance at the excitation wavelength, A em Let be the absorbance at the emission wavelength. Therefore, after fluorescence fingerprint correction for the above samples, the data matrix becomes:

[0142]

[0143] Secondly, in this embodiment, the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced are internally filtered and then subjected to scattering correction. Specifically, the data for ±20nm of the first-order and second-order Rayleigh scattering lines are replaced with 0, the data on the first-order Raman scattering line are subtracted, and then interpolation is used to fill in the subtracted data. After scattering correction, the fluorescent fingerprint data matrix of the above samples becomes:

[0144]

[0145] Finally, in this embodiment, the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced are Raman corrected. The intensity of the scattering-corrected fluorescent fingerprint is converted into Raman units (RU) by integrating the Raman scattering intensity of ultrapure water between 350 nm excitation wavelength and 365–435 nm emission wavelength. After Raman correction, the fluorescent fingerprint data matrix of the above samples becomes:

[0146]

[0147] In summary, the embodiments of this application perform blank correction, internal filtration correction, scattering correction, and Raman correction on the fluorescent fingerprints of the pollution source sample and the water sample to be traced, respectively, to obtain the fluorescent fingerprints of the target pollution source sample and the target water sample to be traced.

[0148] Furthermore, in this embodiment of the application, parallel factor analysis is used to extract the main fluorescent components and their intensities (F) of the fluorescent fingerprints of each pollution source and the water body to be traced. max Its basic principle is:

[0149]

[0150] Where, x ijk The fluorescence intensity of the i-th sample at emission wavelength j and excitation wavelength k; F represents the number of components; a if b jf c kf These are elements in load matrices A, B, and C, respectively; ε ijk The model residuals represent the portion that cannot be explained by the model. Parallel factor analysis revealed 2, 3, 4, 4, 2, 2, 3, and 2 major fluorescent components in the fluorescent fingerprint of pollution source AH, while 3 major fluorescent components were extracted from the fluorescent fingerprint of the water sample to be traced.

[0151] Furthermore, in this embodiment of the application, a self-organizing neural network algorithm is used to perform cluster analysis on the fluorescent fingerprints of the pollution source and the water body to be traced. The fluorescent fingerprints of pollution sources A to H are clustered into categories 1, 2, 2, 2, 1, 1, 2, 2, respectively, and the fluorescent fingerprints of the water body to be traced are clustered into 2 categories. One fingerprint from each category is selected as the typical fluorescent fingerprint of each category.

[0152] Furthermore, in this embodiment of the application, based on the compositional characteristics of the fluorescence peaks of the fluorescence fingerprints of the water samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library.

[0153] Specifically, such as Figure 3As shown, cluster analysis revealed two typical fluorescent fingerprints in the water body to be traced. Firstly, typical fingerprint I has two fluorescence peaks with different emission wavelengths. This fingerprint is simple in composition and can be directly compared with a typical fluorescent fingerprint library without needing to compare with a fluorescent component library or establish a dynamic sub-fingerprint library.

[0154] Secondly, such as Figure 4 As shown, typical fluorescent fingerprint II, in addition to the two fluorescence peaks of fingerprint I, also has two additional fluorescence peaks near the excitation / emission wavelengths of 270nm / 415nm and 325nm / 415nm. This type of fluorescent fingerprint, with multiple fluorescence peaks, is likely a composite fingerprint from multiple pollution sources with minimal overlap. Direct comparison with a typical fluorescent fingerprint library may not match any single pollution source. Due to the limited overlap of the fluorescence peaks, a comparison method using a dynamic sub-fingerprint library can be prioritized. For example, using the emission wavelength of 375nm as a boundary, typical fluorescent fingerprint II can be decomposed into sub-fingerprint 1 and sub-fingerprint 2. Then, all fingerprints in the typical fluorescent fingerprint library can be decomposed using the emission wavelength of 375nm as a boundary, thereby constructing dynamic sub-fingerprint library 1 and dynamic sub-fingerprint library 2.

[0155] Furthermore, in this embodiment of the application, a fluorescent component library is constructed by summarizing the main fluorescent components and their intensities of the extracted pollution source samples, a typical fluorescent fingerprint library is constructed by summarizing the pre-screened typical fluorescent fingerprints, a dynamic sub-fingerprint library is constructed by summarizing the decomposed sub-fingerprints, and a multi-dimensional fluorescent fingerprint database is constructed by summarizing the fluorescent component library, the typical fluorescent fingerprint library, and the dynamic sub-fingerprint library.

[0156] In step S102, if no suspected pollution source is found in the cluster analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced.

[0157] Furthermore, in some embodiments, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the fluorescence peak wavelengths in the fluorescence fingerprint of the water body to be traced. This includes: if there is one fluorescence peak or multiple fluorescence peaks with the same emission wavelength in the fluorescence fingerprint of the water body to be traced, it indicates that there is no suspected pollution source in the preset multi-dimensional fluorescence fingerprint database; if there are multiple fluorescence peaks with different emission wavelengths in the fluorescence fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is less than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the sub-fingerprints of the fluorescence fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescence fingerprint database; if there are multiple fluorescence peaks with different emission wavelengths in the fluorescence fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is greater than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the fluorescence components of the fluorescence fingerprint of the water body to be traced and the fluorescence component library in the preset multi-dimensional fluorescence fingerprint database.

[0158] Specifically, in this embodiment of the application, the next step of comparing the database is determined based on the compositional characteristics of the fluorescence peaks of the fluorescence fingerprint of the water body to be traced. The steps are as follows:

[0159] (1) If the fluorescent fingerprint of the water body to be traced contains only one fluorescent peak or multiple fluorescent peaks with the same emission wavelength, it indicates that the existing database does not contain potential pollution sources. The fingerprint is added to the typical fluorescent fingerprint database to further investigate other pollution sources.

[0160] (2) If the fluorescent fingerprint of the water body to be traced contains multiple fluorescent peaks with different emission wavelengths and the overlapping area between each fluorescent peak is small, it indicates that there may be multiple (≥2) potential pollution sources. Then, the fluorescent fingerprint of the water body to be traced is decomposed into sub-fingerprints containing fluorescent peaks with different emission wavelengths. The boundary of the dynamic sub-fingerprint of the pollution source is consistent with the sub-fingerprint of the water body to be traced. Then, the sub-fingerprint of the fluorescent fingerprint of the water body to be traced is entered into the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database for cluster analysis.

[0161] (3) If the fluorescent fingerprint of the water body to be traced contains multiple fluorescent peaks with different emission wavelengths and the overlapping area between each fluorescent peak is large, or otherwise, it indicates that there may be multiple (≥2) potential pollution sources. After extracting the main fluorescent components of the water body to be traced, cluster analysis is performed on each fluorescent component of the water body to be traced and the fluorescent components of the pollution sources.

[0162] In step S103, the suspected pollution sources of the water body to be traced are determined according to the optimal pollution source determination strategy.

[0163] Furthermore, in some embodiments, the optimal pollution source determination strategy involves performing cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint database in a preset multi-dimensional fluorescent fingerprint database. The strategy then determines the suspected pollution source of the water body to be traced, including: if only one pollution source's sub-fingerprint exists in the dynamic sub-fingerprint database and clusters with the sub-fingerprints of the fluorescent fingerprint of the water body to be traced, then that single pollution source is considered a suspected pollution source; if multiple pollution sources' sub-fingerprints exist in the dynamic sub-fingerprint database and cluster with the sub-fingerprints of the fluorescent fingerprint of the water body to be traced, then the similarity between the sub-fingerprints of the multiple pollution sources and the sub-fingerprints of the fluorescent fingerprint of the water body to be traced is calculated, and the pollution source with the highest similarity is considered a suspected pollution source; if the suspected pollution sources of all the sub-fluorescent fingerprints of the fluorescent fingerprint of the water body to be traced are not completely identical, it indicates the existence of multiple pollution sources; if the sub-fingerprints of the fluorescent fingerprint of the water body to be traced do not cluster with any pollution source's sub-fingerprint in the dynamic sub-fingerprint database, then it is determined that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database.

[0164] Specifically, in this embodiment of the application, each sub-fingerprint of the water body to be traced is clustered with the dynamic sub-fingerprint of the pollution source, and the steps are as follows:

[0165] (1) If a sub-fingerprint of the water body to be traced clusters with a sub-fingerprint of a pollution source, then the pollution source is a suspected pollution source.

[0166] (2) If a certain sub-fluorescent fingerprint of the water body to be traced clusters with the sub-fingerprints of multiple (≥2) pollution sources, then calculate the similarity between the sub-fingerprint of the water body to be traced and the sub-fingerprints of each pollution source, and the one with the highest similarity is the suspected pollution source.

[0167] (3) If the suspected pollution sources of all the sub-fluorescent fingerprints of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources;

[0168] (4) If all the sub-fluorescent fingerprints of the water body to be traced do not cluster with any of the sub-fingerprints of any pollution source, it means that the database does not contain any suspected pollution sources and other pollution sources need to be further investigated.

[0169] Furthermore, in some embodiments, the optimal pollution source determination strategy involves performing cluster analysis based on the fluorescent components of the fluorescent fingerprint of the water body to be traced and the fluorescent component library in a preset multi-dimensional fluorescent fingerprint database. The strategy then determines suspected pollution sources for the water body to be traced, including: if only one pollution source's fluorescent component clusters with the fluorescent components of the fluorescent fingerprint of the water body to be traced in the fluorescent component library, then that single pollution source is considered a suspected pollution source; if multiple pollution sources' fluorescent components cluster with the fluorescent components of the fluorescent fingerprint of the water body to be traced in the fluorescent component library, then the similarity between the fluorescent components of the multiple pollution sources and the fluorescent components of the fluorescent fingerprint of the water body to be traced is calculated, and the pollution source with the highest similarity is considered a suspected pollution source; if the suspected pollution sources for all fluorescent components of the fluorescent fingerprint of the water body to be traced are not completely identical, it indicates the existence of multiple pollution sources; if the fluorescent components of the fluorescent fingerprint of the water body to be traced do not cluster with the fluorescent components of any pollution source in the fluorescent component library, then it is determined that no suspected pollution source exists in the preset multi-dimensional fluorescent fingerprint database.

[0170] Specifically, in this application, the embodiments perform cluster analysis on each fluorescent component of the water body to be traced and the fluorescent components of the pollution source, and the steps are as follows:

[0171] (1) If a fluorescent component of a water body to be traced clusters with a fluorescent component of a pollution source, then the pollution source is a suspected pollution source.

[0172] (2) If a fluorescent component of the water body to be traced is clustered with the fluorescent components of multiple (≥2) pollution sources, the similarity between the fluorescent component of the water body to be traced and the fluorescent components of each pollution source in the same category shall be calculated respectively, and the one with the highest similarity shall be the suspected pollution source.

[0173] (3) If the suspected sources of pollution for all fluorescent components in the water body to be traced are not completely the same, it indicates that there are multiple sources of pollution.

[0174] (4) If all the fluorescent components of the water body to be traced do not cluster with the fluorescent components of any pollution source, it means that the database does not contain any suspected pollution sources and other pollution sources need to be further investigated.

[0175] For example, through the multi-dimensional comparison and identification of pollution sources mentioned above, this application embodiment first uses a self-organizing neural network algorithm to perform cluster analysis on the typical fingerprint I of the water body to be traced and all fingerprints in the typical fluorescent fingerprint database. If the result shows that the fingerprint of pollution source A is clustered with the typical fingerprint I of the water body to be traced, it indicates that pollution source A is a suspected pollution source.

[0176] Secondly, the self-organizing neural network algorithm is used to perform cluster analysis on the sub-fingerprints of the typical fingerprint II of the water body to be traced and the corresponding sub-fingerprints in the dynamic sub-fingerprint database. If sub-fingerprint 1 clusters with the sub-fingerprint of pollution source A, and sub-fingerprint 2 clusters with the sub-fingerprints of pollution sources D and E, it indicates that pollution source A is a suspected pollution source, and pollution sources D and E need to be further judged.

[0177] Specifically, the cosine similarity between sub-fingerprint 2 and the sub-fingerprints of pollution sources D and E is calculated. If the cosine similarity between sub-fingerprint 2 and the sub-fingerprint of pollution source D is 0.89 and the cosine similarity between sub-fingerprint 2 and the sub-fingerprint of pollution source E is 0.93, it indicates that pollution source E is more likely to be a potential pollution source. Therefore, the water body to be traced may be polluted by pollution sources A and E at the same time.

[0178] Next, the self-organizing neural network algorithm is used to perform cluster analysis on the typical fingerprint II of the water body to be traced and all fingerprints in the typical fluorescent fingerprint database. If the typical fingerprint II does not cluster with any fingerprint in the typical fingerprint database, it indicates that there may be multiple pollution sources or other pollution sources.

[0179] It should be noted that the clustering methods in the above steps include, but are not limited to, K-means clustering, density-based spatial clustering of applications with noise (DBSCAN (Density-Based Spatial Clustering of Applications with Noise), hierarchical clustering, and self-organizing neural networks. Preferably, the embodiments of this application use self-organizing neural networks. The similarity measurement methods in the above steps include, but are not limited to, Minkowski distance, Manhattan distance, Euclidean distance, Chebyshev distance, cosine similarity, Pearson correlation coefficient, Mahalanobis distance, Hamming distance, and Jaccard similarity coefficient. Preferably, the embodiments of this application use the cosine similarity measurement method, the formula of which is as follows:

[0180]

[0181] Where cos(θ) is the cosine similarity, and x and y are two one-dimensional row vectors, x i Let y be the i-th component of the x-vector. i Let y be the i-th component of the vector y, and n be the dimension of the vector.

[0182] According to the multi-dimensional comparison and identification method for pollution sources in this application, the fluorescent fingerprints of the water body to be traced are clustered with fingerprints in a typical fluorescent fingerprint library in a preset multi-dimensional fluorescent fingerprint database to obtain clustering analysis results. If no suspected pollution sources are found in the clustering analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprints of the water body to be traced, thereby determining the suspected pollution sources of the water body to be traced. This solves the problems in related technologies, such as the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods. By standardizing samples and fluorescent fingerprints and adopting online multi-pollution tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing methods, and the reliability and stability of the tracing results are improved.

[0183] Next, the multi-dimensional source tracing device for pollution sources proposed according to the embodiments of this application is described with reference to the accompanying drawings.

[0184] Figure 5 This is a schematic diagram of a pollution source multi-dimensional tracing device according to an embodiment of this application.

[0185] like Figure 5 As shown, the multi-dimensional pollution source tracing device 10 includes: a metering pump 100, a filtration module 200, a water quality adjustment module 300, a fingerprint acquisition module 400, and a control system module 500.

[0186] The metering pump 100 is used for the delivery and metering of water samples or reagents between the filtration module, water quality condition module, and fingerprint acquisition module.

[0187] The filter module 200 is used to filter the water sample delivered by the metering pump to obtain a filtered water sample;

[0188] The water quality conditioning module 300 is used to adjust the pH value of the filtered water sample. It includes a conditioning water tank, an ultrapure water tank, an alkaline solution tank, an acid solution tank, a metering pump, a pH probe, and an ultraviolet-visible absorption spectroscopy probe to obtain the test water sample.

[0189] The fingerprint acquisition module 400 is used to collect fluorescent fingerprints from test water samples;

[0190] The control system module 500 stores a computer program. When the computer program is executed by the control system module, it realizes the multi-dimensional comparison and identification method of pollution sources as described above based on the fluorescent fingerprint of the test water sample.

[0191] Furthermore, in some embodiments, the above-mentioned multi-dimensional pollution source tracing device 10 further includes:

[0192] The display module is used to display the test results and pollution source comparison results output by the control system module.

[0193] Specifically, the operation of the multi-dimensional pollution source tracing device in this application embodiment includes the following steps:

[0194] (1) The control system module starts the sampling program, and the water sample is delivered to the filtration module via a metering pump;

[0195] (2) The water sample is processed by the filtration module and then enters the water quality conditioning tank;

[0196] (3) The control system starts the ultraviolet-visible absorption spectroscopy probe in the water quality conditioning tank to collect the ultraviolet-visible absorption spectrum of the water sample, and the result is displayed in real time in the display module. If the absorbance at the lowest excitation wavelength of the water sample is greater than 0.75, the metering pump is started to add ultrapure water for appropriate dilution. If it is less than or equal to 0.75, the dilution step is skipped and the process proceeds to step (4).

[0197] (4) The control system starts the pH probe in the water quality conditioning tank to test the pH value of the water sample, and the result is displayed in real time in the display module. If the pH value of the water sample is less than 6, the metering pump is started to add an appropriate amount of alkaline solution to make the pH value of the water sample between 6 and 8; if the pH value of the water sample is greater than 8, the metering pump is started to add an appropriate amount of acid solution to make the pH value of the water sample between 6 and 9; if the pH value of the water sample is between 6 and 8, the pH adjustment step is skipped and the process proceeds to step (5).

[0198] (5) The control system starts the metering pump to transport the water sample in the water quality conditioning tank to the fingerprint acquisition module, and collects the fluorescent fingerprint of the water sample in the fingerprint acquisition module. The result is displayed in real time in the display module.

[0199] (6) After fingerprint acquisition is completed, the control system starts the metering pump and solenoid valve to empty the sample in the regulating water tank and fingerprint acquisition module;

[0200] (7) After the sample is emptied, the control system starts the metering pump and solenoid valve to add an appropriate amount of ultrapure water to clean the water tank and fingerprint acquisition module, and then empties the cleaning waste liquid.

[0201] (8) The fluorescent fingerprint is collected and transmitted to the control system module. The control system module uses the programmed program to correct the three-dimensional fluorescent fingerprint.

[0202] (9) The control system starts the comparison program, compares the fluorescent fingerprint of the sample with the pollution source fingerprint database that has been embedded in the control system, and displays the comparison results in real time on the display module.

[0203] (10) Repeat the above steps continuously to achieve online early warning and source tracing of water pollution.

[0204] According to the pollution source multi-dimensional tracing device of this application embodiment, the fluorescent fingerprint of the water body to be traced is clustered with fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the clustering analysis results. If no suspected pollution source is found in the clustering analysis results, the optimal pollution source determination strategy is matched according to the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprint of the water body to be traced, thereby determining the suspected pollution source of the water body to be traced. This solves the problems of inability to simultaneously trace multiple pollution sources, lack of standardized procedures for sample and fluorescent fingerprint collection, low efficiency and poor timeliness in related technologies. By standardizing samples and fluorescent fingerprints and adopting multi-pollution online tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing method, and the reliability and stability of the tracing results are improved.

[0205] The multi-dimensional comparison and identification device for pollution sources according to an embodiment of the present invention will be described again with reference to the accompanying drawings.

[0206] Figure 6 This is a block diagram of a multi-dimensional comparison and identification device for pollution sources according to an embodiment of this application.

[0207] like Figure 6 As shown, the multi-dimensional comparison and identification device 20 for the pollution source includes: a data acquisition module 600, a matching module 700, and an identification module 800.

[0208] Among them, the acquisition module 600 is used to acquire the water body to be traced, and perform cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprint in the typical fluorescent fingerprint library in the preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results;

[0209] The matching module 700 is used to match the optimal pollution source determination strategy based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced if no suspected pollution source is found in the cluster analysis results; and

[0210] The identification module 800 is used to identify suspected pollution sources in the water body to be traced based on the best pollution source identification strategy.

[0211] Furthermore, in some embodiments, the acquisition module 600 is specifically used for:

[0212] If the typical fluorescent fingerprint of only one pollution source in the typical fluorescent fingerprint database is clustered with the fluorescent fingerprint of the water body to be traced, then the only pollution source is regarded as a suspected pollution source.

[0213] If there are clusters of typical fluorescent fingerprints of multiple pollution sources and fluorescent fingerprints of the water body to be traced in the typical fluorescent fingerprint database, then the similarity between the typical fluorescent fingerprints of multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0214] If the sub-fingerprints of the fluorescent fingerprint of the water body to be traced do not cluster with the sub-fingerprints of any pollution source in the dynamic sub-fingerprint database, then it is determined that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database.

[0215] Furthermore, in some embodiments, the matching module 700 is specifically used to: if there is one fluorescence peak or multiple fluorescence peaks with the same emission wavelength in the fluorescent fingerprint of the water body to be traced, it indicates that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database;

[0216] If there are multiple fluorescence peaks with different emission wavelengths in the fluorescent fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is less than the preset value, then the best pollution source determination strategy is to perform cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database.

[0217] If the fluorescent fingerprint of the water body to be traced contains multiple fluorescence peaks with different emission wavelengths, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is greater than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the fluorescence components of the fluorescent fingerprint of the water body to be traced and the fluorescence component library in the preset multi-dimensional fluorescent fingerprint database.

[0218] Furthermore, in some embodiments, the identification module 800 is specifically used for:

[0219] If only one pollution source's sub-fingerprint exists in the dynamic sub-fingerprint database and its sub-fingerprint is clustered with the fluorescent fingerprint of the water body to be traced, then the only pollution source that exists will be considered a suspected pollution source.

[0220] If multiple pollution source sub-fingerprints are clustered with the fluorescent fingerprints of the water body to be traced in the dynamic sub-fingerprint database, then the similarity between the sub-fingerprints of multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0221] If the suspected pollution sources of all sub-fluorescent fingerprints of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources.

[0222] If the sub-fingerprints of the fluorescent fingerprint of the water body to be traced do not cluster with the sub-fingerprints of any pollution source in the dynamic sub-fingerprint database, then it is determined that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database.

[0223] Furthermore, in some embodiments, the identification module 800 is specifically used for:

[0224] If the fluorescent component of only one pollution source in the fluorescent component library is clustered with the fluorescent component of the fluorescent fingerprint of the water body to be traced, then the only pollution source is regarded as a suspected pollution source.

[0225] If the fluorescent components stored in multiple pollution sources are clustered with the fluorescent fingerprints of the water body to be traced, then the similarity between the fluorescent components of multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated separately, and the pollution source with the highest similarity is taken as the suspected pollution source.

[0226] If the suspected sources of pollution for all fluorescent components in the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple sources of pollution.

[0227] If the fluorescent components of the fluorescent fingerprint of the water body to be traced do not cluster with the fluorescent components of any pollution source in the fluorescent component library, then it is determined that there is no suspected pollution source in the preset multi-dimensional fluorescent fingerprint database.

[0228] Furthermore, in some embodiments, before performing cluster analysis on the fluorescent fingerprints of the water body to be traced and the fingerprints in the typical fluorescent fingerprint database of the preset multi-dimensional fluorescent fingerprint database, the acquisition module 600 is also used for:

[0229] Collect pollution source samples, water samples to be traced, clean water samples, and ultrapure water samples, and standardize the pollution source samples, water samples to be traced, and clean water samples.

[0230] Fluorescent fingerprints of ultrapure water samples and standardized pollutant source samples, water samples to be traced, and clean water samples were collected, as well as ultraviolet-visible absorption spectra of standardized pollutant source samples, water samples to be traced, and clean water samples.

[0231] Based on a pre-defined standardized processing strategy, the fluorescent fingerprints of the pollution source samples and the fluorescent fingerprints of the water samples to be traced are standardized to obtain the fluorescent fingerprints of the target pollution source samples and the target water samples to be traced.

[0232] Based on the pre-defined parallel factor analysis method, the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample are obtained, and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced are also obtained, so as to construct a fluorescence component library based on the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced.

[0233] At least one typical fluorescent fingerprint of a pollution source sample is selected from the fluorescent fingerprints of the target pollution source sample, and at least one typical fluorescent fingerprint of a water body sample to be traced is selected from the fluorescent fingerprints of the target water body sample to be traced, so as to construct a typical fluorescent fingerprint library based on the typical fluorescent fingerprints of at least one pollution source sample and at least one water body sample to be traced.

[0234] Based on the composition characteristics of the fluorescence peaks of the fluorescence fingerprints of the water samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library, wherein each sub-fluorescent fingerprint contains at least one fluorescence peak.

[0235] A multi-dimensional fluorescent fingerprint database was constructed based on the fluorescent component library, the typical fluorescent fingerprint library, and the dynamic sub-fingerprint library.

[0236] Furthermore, in some embodiments, the acquisition module 600 is specifically used for:

[0237] Blank correction was performed on the fluorescent fingerprints of the pollution source samples and the fluorescent fingerprints of the water samples to be traced, respectively, to obtain the fluorescent fingerprints of the pollution source samples after blank correction and the fluorescent fingerprints of the water samples to be traced after blank correction.

[0238] Based on the preset correction formula, the fluorescence fingerprint of the pollution source sample after blank correction is corrected by internal filtration using the ultraviolet-visible absorption spectrum of the pollution source sample, and the fluorescence fingerprint of the pollution source sample after internal filtration correction is obtained by using the ultraviolet-visible absorption spectrum of the water body sample to be traced by internal filtration, and the fluorescence fingerprint of the water body sample to be traced by internal filtration correction is obtained.

[0239] According to the preset replacement strategy, the first-order Rayleigh scattering data and second-order Rayleigh scattering data of the fluorescence fingerprint of the pollution source sample after internal filtration correction are replaced. According to the first data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the pollution source sample after scattering correction. The first-order Rayleigh scattering data and second-order Rayleigh scattering data of the fluorescence fingerprint of the water body sample to be traced after internal filtration correction are replaced. According to the second data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the water body sample to be traced after scattering correction.

[0240] Based on the integral of the preset Raman scattering intensity of ultrapure water, the data units of the fluorescence fingerprints of the pollution source sample after scattering correction and the fluorescence fingerprints of the water body sample to be traced after scattering correction are converted from dimensionless to Raman units.

[0241] Furthermore, in some embodiments, the acquisition module 600 is specifically used for:

[0242] The fluorescence fingerprint of the pollution source sample is obtained by subtracting the fluorescence fingerprint of the ultrapure water sample from the fluorescence fingerprint numerical matrix of the pollution source sample;

[0243] The fluorescent fingerprint of the water sample to be traced is obtained by subtracting the fluorescent fingerprint of the clean water sample from the fluorescent fingerprint numerical matrix of the water sample to be traced. The fluorescent fingerprint of the water sample to be traced is then corrected for blank.

[0244] Furthermore, in some embodiments, the preset correction formula is:

[0245]

[0246] Among them, F corr F represents the corrected fluorescence intensity. obs A represents the fluorescence intensity before correction. ex To determine the absorbance at the excitation wavelength, A em ν is the absorbance at the emission wavelength.

[0247] Furthermore, in some embodiments, the preset parallel factor analysis method is as follows:

[0248]

[0249] Where, x ijk The fluorescence intensity of the i-th sample at emission wavelength j and excitation wavelength k; F represents the number of components; a if b jf c kf These are elements in load matrices A, B, and C, respectively; ε ijk This represents the model residuals.

[0250] Furthermore, in some embodiments, a preset clustering algorithm is used for clustering analysis, wherein the preset clustering algorithm includes at least one of K-means clustering, density-based noise applied spatial clustering, hierarchical clustering, and self-organizing neural networks.

[0251] Furthermore, in some embodiments, a preset similarity measurement method is used to calculate similarity, wherein the preset similarity measurement method includes at least one of Kofsky distance, Manhattan distance, Euclidean distance, Chebyshev distance, cosine similarity, Pearson correlation coefficient, Mahalanobis distance, Hamming distance and Jaccard similarity coefficient.

[0252] According to the multi-dimensional comparison and identification device for pollution sources in this application embodiment, the fluorescent fingerprint of the water body to be traced is clustered with fingerprints in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain clustering analysis results. If no suspected pollution source is found in the clustering analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescent fingerprint of the water body to be traced, thereby determining the suspected pollution source of the water body to be traced. This solves the problems in related technologies such as the inability to simultaneously trace multiple pollution sources, the lack of standardized procedures for sample and fluorescent fingerprint collection, and the low efficiency and poor timeliness of the tracing methods. By standardizing samples and fluorescent fingerprints and adopting multi-pollution online tracing technology, the accuracy of fluorescent fingerprints, the timeliness of the tracing method, and the reliability and stability of the tracing results are improved.

[0253] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:

[0254] The memory 701, the processor 702, and the computer program stored on the memory 701 and executable on the processor 702.

[0255] When the processor 702 executes the program, it implements the multi-dimensional comparison and identification method for pollution sources provided in the above embodiments.

[0256] Furthermore, electronic devices also include:

[0257] Communication interface 707 is used for communication between memory 701 and processor 702.

[0258] The memory 701 is used to store computer programs that can run on the processor 702.

[0259] The memory 701 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0260] If the memory 701, processor 702, and communication interface 707 are implemented independently, then the communication interface 707, memory 701, and processor 702 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0261] Optionally, in a specific implementation, if the memory 701, processor 702, and communication interface 707 are integrated on a single chip, then the memory 701, processor 702, and communication interface 707 can communicate with each other through an internal interface.

[0262] The processor 702 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0263] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described multi-dimensional comparison and identification method for pollution sources.

[0264] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. 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.

[0265] 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.

[0266] 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 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.

[0267] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A multi-dimensional comparison and identification method for pollution sources, characterized in that, Includes the following steps: Collect pollution source samples, water body samples to be traced, clean water body samples, and ultrapure water samples, and perform standardized processing on the pollution source samples, the water body samples to be traced, and the clean water body samples; Fluorescent fingerprints of the ultrapure water sample and the standardized pollutant source sample, the water body sample to be traced, and the clean water sample were collected respectively, as well as ultraviolet-visible absorption spectra of the standardized pollutant source sample, the water body sample to be traced, and the clean water sample. Based on a pre-defined standardization strategy, the fluorescent fingerprints of the pollution source sample and the water sample to be traced are standardized to obtain the fluorescent fingerprints of the target pollution source sample and the target water sample to be traced. This includes: performing blank correction on the fluorescent fingerprints of the pollution source sample and the water sample to be traced, respectively, to obtain blank-corrected fluorescent fingerprints of the pollution source sample and blank-corrected fluorescent fingerprints of the water sample to be traced; and based on a pre-defined correction formula, performing internal filtration correction on the blank-corrected fluorescent fingerprint of the pollution source sample using the UV-Vis absorption spectrum of the pollution source sample, to obtain the internally filtration corrected fluorescent fingerprint of the pollution source sample, and then using the UV-Vis absorption spectrum of the water sample to be traced. The fluorescence fingerprint of the water sample to be traced is subjected to internal filtration correction to obtain the fluorescence fingerprint of the water sample to be traced after internal filtration correction. According to a preset replacement strategy, the first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the pollution source sample after internal filtration correction are replaced. According to a first data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the pollution source sample after scattering correction. The first-order Rayleigh scattering data and the second-order Rayleigh scattering data of the fluorescence fingerprint of the water sample to be traced after internal filtration correction are replaced. According to a second data processing strategy, the data on the first-order Raman scattering line of the pollution source sample after internal filtration correction are subtracted and filled to obtain the fluorescence fingerprint of the water sample to be traced after scattering correction. Based on a pre-defined parallel factor analysis method, the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample are obtained, and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced are also obtained, so as to construct a fluorescence component library based on the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target pollution source sample and the fluorescence components and fluorescence intensity of the fluorescence fingerprint of the target water body sample to be traced. At least one typical fluorescent fingerprint of a pollution source sample is selected from the fluorescent fingerprints of the target pollution source sample, and at least one typical fluorescent fingerprint of a water body sample to be traced is selected from the fluorescent fingerprints of the target water body sample to be traced, so as to construct a typical fluorescent fingerprint library based on the typical fluorescent fingerprints of the at least one pollution source sample and the at least one water body sample to be traced. Based on the composition characteristics of the fluorescence peaks of the fluorescence fingerprints of the water samples to be traced, the typical fluorescence fingerprint of each pollution source sample is decomposed into multiple sub-fluorescent fingerprints to obtain a dynamic sub-fingerprint library, wherein each sub-fluorescent fingerprint contains at least one fluorescence peak. A multi-dimensional fluorescent fingerprint database is constructed based on the fluorescent component library, the typical fluorescent fingerprint library, and the dynamic sub-fingerprint library. Collect the water body to be traced, and perform cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprint in the typical fluorescent fingerprint library in the preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results; If no suspected pollution source is found in the cluster analysis results, the optimal pollution source determination strategy is matched based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced; and The suspected pollution sources of the water body to be traced are determined based on the optimal pollution source determination strategy.

2. The method according to claim 1, characterized in that, The step involves performing cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprints in the typical fluorescent fingerprint database of a preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results, including: If only one typical fluorescent fingerprint of a pollution source exists in the typical fluorescent fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is regarded as the suspected pollution source. If the typical fluorescent fingerprints of multiple pollution sources are clustered with the fluorescent fingerprints of the water body to be traced in the typical fluorescent fingerprint database, then the similarity between the typical fluorescent fingerprints of the multiple pollution sources and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source. If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

3. The method according to claim 1, characterized in that, The optimal pollution source determination strategy based on matching the number of fluorescence peaks and the wavelength of fluorescence peaks in the fluorescence fingerprint of the water body to be traced includes: If there is one fluorescence peak or multiple fluorescence peaks with the same emission wavelength in the fluorescent fingerprint of the water body to be traced, it indicates that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database. If there are multiple fluorescence peaks with different emission wavelengths in the fluorescent fingerprint of the water body to be traced, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is less than a preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database. If the fluorescent fingerprint of the water body to be traced contains multiple fluorescence peaks with different emission wavelengths, and the overlapping area of ​​the multiple fluorescence peaks with different emission wavelengths is larger than the preset value, then the optimal pollution source determination strategy is to perform cluster analysis based on the fluorescence components of the fluorescent fingerprint of the water body to be traced and the fluorescence component library in the preset multi-dimensional fluorescent fingerprint database.

4. The method according to claim 3, characterized in that, The optimal pollution source identification strategy involves clustering analysis of the sub-fingerprints of the fluorescent fingerprint of the water body to be traced and the dynamic sub-fingerprint library in the preset multi-dimensional fluorescent fingerprint database. Identifying the suspected pollution source of the water body to be traced based on the optimal pollution source identification strategy includes: If only one pollution source's sub-fingerprint exists in the dynamic sub-fingerprint database and clusters with the fluorescent fingerprint of the water body to be traced, then the only existing pollution source is taken as the suspected pollution source. If multiple pollution source sub-fingerprints are clustered with the fluorescent fingerprints of the water body to be traced in the dynamic sub-fingerprint database, then the similarity between the multiple pollution source sub-fingerprints and the fluorescent fingerprints of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source. If the suspected pollution sources of all sub-fluorescent fingerprints of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources. If the sub-fingerprint of the fluorescent fingerprint of the water body to be traced does not cluster with the sub-fingerprint of any pollution source in the dynamic sub-fingerprint database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

5. The method according to claim 3, characterized in that, The optimal pollution source identification strategy involves clustering analysis of the fluorescent components of the fluorescent fingerprint of the water body to be traced against the fluorescent component library in the preset multi-dimensional fluorescent fingerprint database. Identifying suspected pollution sources of the water body to be traced based on the optimal pollution source identification strategy includes: If the fluorescent component library contains only one pollution source whose fluorescent component clusters with the fluorescent fingerprint of the water body to be traced, then the only pollution source is considered the suspected pollution source. If the fluorescent components stored in the fluorescent components of multiple pollution sources are clustered with the fluorescent components of the fluorescent fingerprint of the water body to be traced, then the similarity between the fluorescent components of the multiple pollution sources and the fluorescent components of the fluorescent fingerprint of the water body to be traced is calculated respectively, and the pollution source with the highest similarity is taken as the suspected pollution source. If the suspected pollution sources of all fluorescent components of the fluorescent fingerprint of the water body to be traced are not completely the same, it indicates that there are multiple pollution sources. If the fluorescent components of the fluorescent fingerprint of the water body to be traced do not cluster with the fluorescent components of any pollution source in the fluorescent component database, then it is determined that the suspected pollution source does not exist in the preset multi-dimensional fluorescent fingerprint database.

6. The method according to claim 1, characterized in that, The blank correction of the fluorescent fingerprints of the pollution source sample and the fluorescent fingerprints of the water sample to be traced includes: The fluorescence fingerprint of the pollution source sample is obtained by subtracting the fluorescence fingerprint of the ultrapure water sample from the fluorescence fingerprint numerical matrix of the pollution source sample; The fluorescent fingerprint of the clean water sample is obtained by subtracting the fluorescent fingerprint value matrix of the water sample to be traced from the fluorescent fingerprint value matrix of the water sample to be traced. The blank-corrected fluorescent fingerprint of the water sample to be traced is then obtained.

7. The method according to claim 1, characterized in that, The preset correction formula is: ； in, F corr The corrected fluorescence intensity, F obs The fluorescence intensity before correction. A ex To determine the absorbance at the excitation wavelength, A em ν is the absorbance at the emission wavelength.

8. The method according to claim 1, characterized in that, The preset parallel factor analysis method is as follows: ； in, The fluorescence intensity of the i-th sample at emission wavelength j and excitation wavelength k; F represents the number of components; , , These are the elements in load matrices A, B, and C, respectively; This represents the model residuals.

9. The method according to any one of claims 1-5, characterized in that, Cluster analysis is performed using a preset clustering algorithm, wherein the preset clustering algorithm includes at least one of K-means clustering, density-based noise applied spatial clustering, hierarchical clustering, and self-organizing neural networks.

10. The method according to claim 4 or 5, characterized in that, Similarity is calculated using a preset similarity measurement method, wherein the preset similarity measurement method includes at least one of Kofsky distance, Manhattan distance, Euclidean distance, Chebyshev distance, cosine similarity, Pearson correlation coefficient, Mahalanobis distance, Hamming distance and Jaccard similarity coefficient.

11. A multi-dimensional pollution source tracing device, characterized in that, include: Metering pump; The filtration module is used to filter the water sample delivered by the metering pump to obtain a filtered water sample. A water quality adjustment module is used to adjust the pH value of the filtered water sample to obtain a test water sample; A fingerprint acquisition module is used to acquire fluorescent fingerprints of the test water sample; The control system module stores a computer program. When the computer program is executed by the control system module, it realizes the multi-dimensional comparison and identification method of pollution sources as described in any one of claims 1-10 based on the fluorescent fingerprint of the test water sample.

12. The apparatus according to claim 11, characterized in that, Also includes: The display module is used to display the test results and pollution source comparison results output by the control system module.

13. A multi-dimensional comparison and identification device for pollution sources, characterized in that, To implement the multi-dimensional comparison and identification method for pollution sources as described in any one of claims 1-10, the multi-dimensional comparison and identification device for pollution sources includes: The acquisition module is used to acquire the water body to be traced and perform cluster analysis on the fluorescent fingerprint of the water body to be traced and the fingerprint in the typical fluorescent fingerprint library of the preset multi-dimensional fluorescent fingerprint database to obtain the cluster analysis results. The matching module is used to match the optimal pollution source determination strategy based on the number of fluorescence peaks and the emission wavelength of the fluorescence peaks in the fluorescence fingerprint of the water body to be traced if no suspected pollution source is found in the cluster analysis results; and The identification module is used to identify the suspected pollution source of the water body to be traced based on the optimal pollution source determination strategy.

14. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the multi-dimensional comparison and identification method for pollution sources as described in any one of claims 1-10.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the multi-dimensional comparison and identification method for pollution sources as described in any one of claims 1-10.

Images