CN117309831A - Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM - Google Patents
Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM Download PDFInfo
- Publication number
- CN117309831A CN117309831A CN202311245608.6A CN202311245608A CN117309831A CN 117309831 A CN117309831 A CN 117309831A CN 202311245608 A CN202311245608 A CN 202311245608A CN 117309831 A CN117309831 A CN 117309831A
- Authority
- CN
- China
- Prior art keywords
- data
- dimensional
- sample
- pollution
- samples
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 77
- 238000002189 fluorescence spectrum Methods 0.000 claims abstract description 57
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 38
- 238000012545 processing Methods 0.000 claims abstract description 37
- 230000002159 abnormal effect Effects 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 29
- 239000013598 vector Substances 0.000 claims abstract description 24
- 239000010865 sewage Substances 0.000 claims abstract description 23
- 238000001506 fluorescence spectroscopy Methods 0.000 claims abstract description 15
- 239000011159 matrix material Substances 0.000 claims description 45
- 238000001069 Raman spectroscopy Methods 0.000 claims description 37
- 239000000126 substance Substances 0.000 claims description 29
- 230000005284 excitation Effects 0.000 claims description 27
- 230000009467 reduction Effects 0.000 claims description 23
- 238000012937 correction Methods 0.000 claims description 20
- 238000010606 normalization Methods 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 18
- 238000004458 analytical method Methods 0.000 claims description 17
- 229910021642 ultra pure water Inorganic materials 0.000 claims description 15
- 239000012498 ultrapure water Substances 0.000 claims description 15
- 238000000556 factor analysis Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 13
- 238000012216 screening Methods 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 7
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 230000010354 integration Effects 0.000 claims description 6
- 239000005416 organic matter Substances 0.000 claims description 6
- 238000001237 Raman spectrum Methods 0.000 claims description 3
- 125000003275 alpha amino acid group Chemical group 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 3
- 239000003344 environmental pollutant Substances 0.000 claims description 3
- 231100000719 pollutant Toxicity 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 2
- 239000002957 persistent organic pollutant Substances 0.000 claims 4
- 238000007405 data analysis Methods 0.000 claims 1
- 230000007547 defect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000001514 detection method Methods 0.000 description 4
- 238000010790 dilution Methods 0.000 description 4
- 239000012895 dilution Substances 0.000 description 4
- 239000000463 material Substances 0.000 description 3
- 239000012528 membrane Substances 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000004043 dyeing Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000002071 nanotube Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000004321 preservation Methods 0.000 description 2
- 239000002351 wastewater Substances 0.000 description 2
- 238000003911 water pollution Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000012851 eutrophication Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001917 fluorescence detection Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000002574 poison Substances 0.000 description 1
- 231100000614 poison Toxicity 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/65—Raman scattering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2132—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on discrimination criteria, e.g. discriminant analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/63—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light optically excited
- G01N21/64—Fluorescence; Phosphorescence
- G01N2021/6417—Spectrofluorimetric devices
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Health & Medical Sciences (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Immunology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention provides a pollution tracing method of river channel organic matters based on a three-dimensional fluorescent LPP-SVM. The method comprises the following steps: the method comprises the steps of collecting sewage samples in a region to be traced, obtaining corresponding three-dimensional fluorescence data, processing the obtained three-dimensional fluorescence data, inputting the processed data into a vector machine for training, constructing a pollution source three-dimensional fluorescence identification model, collecting a plurality of point location water samples of a river channel, obtaining three-dimensional fluorescence data, distinguishing abnormal point locations, processing the three-dimensional fluorescence data of the abnormal point locations, inputting the three-dimensional fluorescence data of the abnormal point locations into the identification model for identification, and thus locking the pollution source, and obtaining tracing results. According to the invention, through the pre-constructed LPP-SVM three-dimensional fluorescence spectrum identification model, after the abnormal point position of the organic matters in the river channel is found, the pollution source can be accurately locked, the defects of long consumed time, low information utilization rate and the like of the traditional fluorescence tracing are avoided, and the accuracy and the scientificity of the identification are improved.
Description
Technical Field
The invention relates to the technical field of water pollution tracing, in particular to a method for tracing pollution of river channel organic matters based on a three-dimensional fluorescent LPP-SVM.
Background
With the continuous promotion of industrialization, organic matter pollution events of a river channel caused by drainage leakage, drainage stealing, overflow and sudden pollution events occur, and organic wastewater with higher concentration continuously enters the river channel, so that serious consequences such as water eutrophication, poison organisms and the like are brought. Conventional traceability investigation often requires a lot of manpower, and is long in time consumption and poor in timeliness and accuracy. Therefore, a method for rapidly and efficiently inspecting and tracing organic pollution sources of river channels is urgently needed.
Compared with other organic matter detection methods, the three-dimensional fluorescence detection method has the advantages of simple pretreatment, green and high efficiency, high sensitivity, rich information content and the like. When exogenous organic matters are not input into the river channel, the three-dimensional fluorescence spectrum of each monitoring point position of the river channel has little change. When pollution occurs, abnormal point screening is mainly carried out by judging the positions of characteristic fluorescent peaks, peak intensities, the number of peaks, fluorescent parameters and the like of a map, and when the input quantity is small and the fluorescent peaks of substances overlap, a parallel factor analysis method is introduced to screen abnormal points. Greatly reduces the time for screening the organic pollution points and the use of chemical reagents in detection.
At present, the commonly used three-dimensional fluorescence spectrum identification method is to analyze the fluorescence spectrum by using a parallel factor analysis method, a partial least square method, an alternate three-linear decomposition method, a non-negative matrix factorization method and the like to obtain multi-component fluorescence information, and compare the multi-component fluorescence information with the known pollution sources, wherein the pollution points are usually required to be continuously sampled, the manual comparison time is long, and part of information is lost in the analysis process, so that misjudgment is caused.
In addition, the traditional method for identifying the pollution point location by utilizing characteristic parameters such as the shape characteristic, peak position, fluorescence intensity and the like of the fluorescent spectrum has great human randomness, and the utilization rate of information is not high, so that the identification or comparison accuracy is low. Literature research and experiments show that the three-dimensional fluorescence spectrum of the same type of industry has similarity under the condition that the production process is approximately the same. The three-dimensional fluorescence identification method is used as a learning material, an identification model is trained, the three-dimensional fluorescence identification analysis step is greatly simplified, the information utilization rate is improved, and the scientificity of a tracing result is improved.
Disclosure of Invention
The invention provides a method for tracing pollution of river channel organic matters based on a three-dimensional fluorescent LPP-SVM, which aims at judging the abnormality of a river channel water sample by a three-dimensional fluorescent analysis means from the existing laboratory substance analysis level, screens the pollution points of the river channel organic matters and improves the organic pollution screening efficiency.
The specific technical scheme is as follows:
a method for tracing pollution of river channel organic matters based on three-dimensional fluorescent LPP-SVM comprises the following steps:
(1) Collecting sewage samples of water-wading polluted enterprise water outlets in the area to be traced, and carrying out three-dimensional fluorescence spectrum scanning on the sewage samples to obtain three-dimensional fluorescence spectrum data corresponding to the samples;
(2) Sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the acquired three-dimensional fluorescence spectrum data, and classifying the processed data according to the pollution industry category to obtain a classified training set;
(3) Inputting the training set into a support vector machine model for distinguishing the types of the pollution industry for training to obtain a pollution source three-dimensional fluorescence identification model for distinguishing the types of the pollution industry;
(4) Collecting river water samples at different sampling points of a pollutant discharge river channel in a tracing area, and performing three-dimensional fluorescence spectrum scanning on the river water samples to obtain three-dimensional fluorescence spectrum data of each river water sample;
(5) Screening out river water samples with abnormal three-dimensional fluorescence spectrum data according to the three-dimensional fluorescence spectrum data obtained in the step (4);
(6) Sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the three-dimensional fluorescence spectrum data of the abnormal river sample screened in the step (5) to obtain sample data to be traced;
(7) Inputting the sample data to be traced in the step (6) into the pollution source three-dimensional fluorescent identification model in the step (3) to obtain pollution industry types corresponding to the sample to be traced;
(8) And (3) locking a pollution enterprise from the area to be traced according to the pollution category in the step (7).
Further, the enterprise pollution source is sewage before the nano tube after passing through the enterprise sewage treatment facility.
Further, the pollution source water sample is filtered by a Millipore filter membrane with the pore diameter of 0.22 mu m after being collected, scanned by a machine at the room temperature of 25 ℃, and the sample to be measured is put into a refrigerator at the temperature of 4 ℃ for preservation.
Further, in the step (1), the instrument parameters of the three-dimensional fluorescence scanning are set to be in an Ex scanning range of 220-450 nm, an Em scanning range of 260-600 nm, the scanning bandwidths are all 5nm, the Ex step length is set to be 5nm, the Em step length is set to be 1nm, the slit width is set to be 5nm, and the scanning speed is set to 2400nm/min.
Further, in step (2), the raman correction process includes the steps of:
step (2-1), carrying out three-dimensional fluorescence spectrum scanning on ultrapure water to obtain three-dimensional fluorescence spectrum data of the ultrapure water, carrying out dilution treatment on a sample with higher concentration exceeding the upper limit of a three-dimensional fluorescence detector, and carrying out multiple dilutions with 5-time gradients until the fluorescence intensity falls into a detection limit;
step (2-2) of calculating a Raman peak integrated value A of the ultrapure water by using the formula (1) rp The calculation formula is as follows:
(1),for a specific lambda ex Lower corresponding lambda em I.e. the integral of the fluorescence intensity, i.e. the raman integral; lambda (lambda) ex Represents the excitation wavelength; lambda (lambda) em Representing the emission wavelength; d represents the integral symbol, < >>Is a certain lambda ex Lower lambda em The measured fluorescence intensity of the Raman spectrum;and->The start point and the end point of the integration interval are the common interval is (371, 428) nm;
step (2-3) of dividing all the three-dimensional fluorescence spectrum data obtained in step (1) by A of ultrapure water rp The method comprises the steps of obtaining spectral data of a sewage sample in units of (R.U.), wherein the calculation formula is as follows:
in the formula (2), the amino acid sequence of the compound,is of arbitrary lambda ex 、λ em The corresponding corrected data, i.e. fluorescence intensity in raman (r.u.); />To correct for the former arbitrary lambda ex 、λ em The corresponding fluorescence intensity, in (a.u.); arp is the integral value of the Raman peak of ultrapure water.
Further, the raman correction process and the data normalization process sequentially include: removal of raman rayleigh scattering and tiling of data;
step (2-4), the method for removing the raman rayleigh scattering comprises the following steps:
removing the regions of Em < Ex+/-20 nm and Em >2 Ex+/-10 nm, inserting 0 value into the removed region for replacement, and retaining the most obvious region of fluorescence characteristics;
step (2-5), the tiling method of the data comprises the following steps:
expanding the corrected data along the direction of the excitation wavelength i, and connecting the data points between adjacent rows end to end; the method comprises the steps of converting a matrix of a×b into a vector form of 1×m, merging N samples into a matrix form of n×m, wherein N is the number of samples, M=a×b, M is a characteristic quantity, a is the number of rows for acquiring three-dimensional fluorescent matrix data, and b is the number of columns for acquiring the three-dimensional fluorescent matrix data;
step (2-6), the normalization processing mode comprises:
and normalizing the matrix by using a mapmin max function to convert the sample into a row vector, wherein the normalization formula is as follows:
in the formula (3), x' represents the value of the single data of each sample characteristic, min is the minimum value of the sample characteristic data, and max is the maximum value of the sample characteristic data.
Further, in the step (2), the dimension reduction method includes inputting the normalized nxm data set into a LPP (Locality Preserving Projections) dimension reduction model to obtain dimension reduced data nxm, M < 16027, where N is the number of samples, M is a feature quantity, and M is a feature quantity after dimension reduction.
Further, in the step (2), labeling is carried out on the data subjected to dimension reduction according to the pollution source type.
Further, the method for training the model in the step (3) comprises the following steps:
step (3-1), training the training set classified in the step (2) by using a Libsvm tool box, wherein the number of samples of each training set is more than 20;
step (3-2), optimizing penalty parameters c and kernel function parameters g by adopting an SVMcgForclass function in the training process;
and (3-3) randomly selecting a plurality of unmodeled samples from the sewage samples of each pollution industry category collected in the step (1) as a prediction set, and checking the recognition performance of the model.
Further, in the step (5), the method adopted in the screening of the river water sample with abnormal three-dimensional fluorescence spectrum data is a parallel factor analysis method or a fluorescence data analysis method;
the parallel factor analysis method specifically comprises the following steps:
step (5-1), inputting the processed three-dimensional fluorescence spectrum data of the river point positions into a DOMFluor tool box in Matlab software for parallel factor analysis, and obtaining a fluorescence intensity load matrix A;
constructing a matrix B by using a fluorescence intensity load matrix A according to a formula (4), and calculating a leverage ratio L according to a formula (5);
L i ＝b ii i＝1,2,…,I (5)
in the formulas (4) and (5), L i The lever rate of the ith sample is calculated, bii is the main diagonal element of the matrix B, and I is the number of samples; matrix A is the fluorescence intensity load matrix of each component, A H Is a conjugate matrix of matrix A, (A) H A) + Is A H A pseudo-inverse matrix of A;
step (5-3), the leverage ratio is influenced by the selection of the factor number F, the factor number is required to be adjusted to observe a residual error matrix diagram formed by the factor number to determine the optimal factor number, after the factor number is increased, the component residual error value is not greatly changed and the residual error diagram is randomly distributed, if no special structure exists, the factor number is confirmed to be optimal before the factor number is not increased, the general factor number is 2-6, when the ith sample leverage ratio is more than 0.5, the point is used as a river channel abnormal point, and the point is further input into the model trained in the step (3) for recognition;
the fluorescence data analysis method specifically comprises the following steps:
step (5-I), finding out characteristic substances from the three-dimensional fluorescence spectrum data obtained in the step (4); the characteristic substance is tyrosine-like peak B, em=305 nm; ex=275 nm; tryptophan-like peak T, em=340 nm; ex=275 nm; humic substances a, em=400-460 nm, ex=260 nm; humic-like substances M, em=370-410 nm, ex=290-310 nm; humic substances C, em=420-460 nm, ex=320-360 nm; wherein Em is the emission wavelength of the substance to be measured, and Ex is the excitation wavelength of the substance to be measured;
step (5-II), calculating the numerical value of the characteristic index corresponding to the characteristic substance in the step (5-I); the characteristic indexes are fluorescence area integral, fluorescence index, humification index, biological index and freshness fluorescence parameters; determining the position of a fluorescence peak value according to the different positions of different fluorescent groups at excitation wavelength and emission wavelength, extracting the peak value and a fluorescence integration area, simultaneously calculating fluorescence parameters,
the calculation of the fluorescence area integral is as follows:
the fluorescent region, region I (lambda) Ex <250nm，λ Em <350 nm), region II (lambda Ex <250nm，350<λ Em <380 nm), III region (lambda Ex <250nm，λ Em >380 nm), intermediate excitation wavelength in IV region (250 nm)<λ Ex <280nm，λ Em <380 nm), V (region) (lambda Ex >280nm，λ Em >380 nm), the divided areas are shown in figure 9, and can be adjusted according to the three-dimensional fluorescence characteristics of the actual water sample, and the integral formula of the fluorescence area corresponding to the ith area is as follows:
in formula (6), phi i For the volume integral of the i-th region,for the corresponding fluorescence intensity under a certain excitation emission wavelength dlambda em And dλ ex Is the differentiation of the excitation wavelength and the emission wavelength.
The fluorescent region is unevenly divided in the following manner: using multiplication factor MF i Correction is performed as follows:
s in (7) i For the area of the i-th block region, MF i For multiplication factor, the volume fraction of the modified i region is phi' i ＝φ i ×MF i 。
The fluorescence index FI is calculated as follows:
I Em470 /I Em520 ，Ex＝370nm (8)
in formula (8) I Em470 And I Em520 For fluorescence intensities at emission wavelengths 480nm and 520nm at ex=370 nm excitation wavelength.
The humification index HIX is calculated as follows:
HIX＝∑I Em435-480 /∑I Em300-345 ,Ex＝245nm) (9)
sigma I in formula (9) Em435-480 And Sigma I Em300 345 is the integral of emission wavelengths 435-489nm and 300-345nm at excitation wavelengths ex=245 nm.
The self-biogenic index BIX is calculated as follows:
BIX＝I Em380 /I Ex430 ,Ex＝310nm (10)
in the formula (10), I Em380 And I Em430 For fluorescence intensities at emission wavelengths 380nm and 430nm at ex=310 nm excitation wavelength.
The freshness index calculation formula is as follows:
freshness=i Em380 /max(I Em420-435 ),Ex＝310nm (11)
In the formula (11) Em380 For a fluorescence intensity of 380nm emission wavelength at ex=310 nm excitation wavelength. max (I) Em420-435 ) Is the maximum at emission wavelengths 420-435nm at ex=310 nm excitation wavelength.
Step (5-III), carrying out statistical difference analysis on the calculated characteristic index value by using a triple standard difference method to obtain an abnormal value; will have anomalies; the method for performing outlier difference analysis is a triple standard deviation method.
Further, the three-dimensional fluorescence spectrum data processing method of the abnormal river water sample sequentially carries out Raman correction processing, data normalization processing and dimension reduction processing.
Further, in the step (8), if the locked pollution enterprise is one, determining the enterprise as a pollution source; if the locked polluted enterprises are more than two, performing the step (8-1);
step (8-1), collecting the sewage samples of a plurality of locked polluted enterprises, carrying out three-dimensional fluorescence spectrum scanning on the sewage samples to obtain three-dimensional fluorescence spectrum data, and sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the obtained three-dimensional fluorescence spectrum data;
step (8-2), performing similarity matching on a plurality of locked polluted enterprises by adopting a formula (12) to obtain an included angle cosine value; the calculation method of the cosine value of the included angle is as follows
In the formula (12), a is a vector of 1×m after the three-dimensional fluorescent sample of the abnormal river sample screened in the step (5) is spread, b i 1 Xm vector after expansion for the ith locked enterprise three-dimensional fluorescence data, |a| is the modulus of the a vector, |b i The I is a modulus of bi vectors, cos theta is cosine values of the two vectors, and m is the characteristic dimension after expansion;
the cosine value range of the included angle is [0,1], the closer the acquired value is to 1, the more similar the two groups of data are represented, and the closest to 1 is obtained by comparing the obtained cosine values of different included angles through locking a plurality of polluted enterprises, and the polluted enterprises are judged as pollution sources.
Compared with the prior art, the invention has the following beneficial effects:
(1) The invention provides a method for tracing pollution of river channel organic matters based on a three-dimensional fluorescent LPP-SVM, which aims at judging the abnormality of a river channel water sample by a three-dimensional fluorescent analysis means from the existing laboratory substance analysis level, screens the pollution points of the river channel organic matters and improves the organic pollution screening efficiency.
(2) The invention aims at providing a method for constructing an LPP-SVM three-dimensional fluorescence identification model to identify pollution sources, and improving the accuracy and scientificity of pollution tracing.
Drawings
Fig. 1 is a flowchart of a method for tracing organic matters in a river channel by using the three-dimensional fluorescent LPP-SVM of example 1.
Fig. 2 is a diagram of the result of optimizing the support vector machine by using the grid optimization parameters in embodiment 1.
Fig. 3 is a diagram of recognition accuracy of the support vector machine in embodiment 1.
FIG. 4 is a graph showing the results of the lever rate analysis of the three-dimensional fluorescence parallel factor analysis in example 1.
FIG. 5 is a graph showing the result of extraction of fluorescence peak intensities in example 1.
FIG. 6 is a fluorescent region division diagram in example 1.
FIG. 7 is a graph showing the results of the calculation of fluorescence area integration in example 1.
FIG. 8 is a graph showing the calculation results of fluorescence parameters in example 1.
Fig. 9 is a graph showing the result of identifying abnormal points in example 1.
Detailed Description
The invention will be further described with reference to the following examples, which are given by way of illustration only, but the scope of the invention is not limited thereto.
Example 1
The case provides a pollution tracing method of river channel organic matters based on a three-dimensional fluorescent LPP-SVM, which comprises the following steps:
(1) Collecting sewage samples of water-wading polluted enterprise water outlets in the area to be traced, and carrying out three-dimensional fluorescence spectrum scanning on the sewage samples to obtain three-dimensional fluorescence spectrum data corresponding to the samples;
the method comprises the steps of collecting an enterprise pollution source, namely sewage before a nano tube after the enterprise pollution source passes through an enterprise sewage treatment facility, filtering a pollution source water sample by a Millipore filter membrane with the aperture of 0.22 mu m, scanning the pollution source water sample on a machine at the room temperature of 25 ℃, placing a sample to be detected in a refrigerator with the temperature of 4 ℃ for preservation, setting the instrument parameters of three-dimensional fluorescence scanning to be Ex scanning range of 220-450 nm, em scanning range of 260-600 nm, setting the scanning bandwidth to be 5nm, setting the Ex step length to be 5nm, setting the Em step length to be 1nm, setting the slit width to be 5nm, and setting the scanning speed to be 2400nm/min;
(2) Sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the acquired three-dimensional fluorescence spectrum data, and classifying the processed data according to the pollution industry category to obtain a classified training set;
the raman correction process comprises the following steps:
step (2-1), carrying out three-dimensional fluorescence spectrum scanning on ultrapure water to obtain three-dimensional fluorescence spectrum data of the ultrapure water, carrying out dilution treatment on a sample with higher concentration exceeding the upper limit of a three-dimensional fluorescence detector, and carrying out multiple dilutions with 5-time gradients until the fluorescence intensity falls into a detection limit;
step (2-2) of calculating a Raman peak integrated value A of the ultrapure water by using the formula (1) rp The calculation formula is as follows:
(1),for a specific lambda ex Lower corresponding lambda em Raman integral value of (a); lambda (lambda) ex Represents the excitation wavelength; lambda (lambda) em Representing the emission wavelength; d represents the integral symbol, < >>Is a certain lambda ex Lower lambda em The measured fluorescence intensity of the Raman spectrum; />And->Starting and ending points of an integration interval;
step (2-3), dividing the fluorescence intensity data of all samples of the batch by A of ultrapure water rp The method comprises the steps of obtaining spectral data of a sewage sample in units of (R.U.), wherein the calculation formula is as follows:
in the formula (2), the amino acid sequence of the compound,is of arbitrary lambda ex 、λ em The corresponding corrected data, i.e. fluorescence intensity in raman (r.u.); />To correct for the former arbitrary lambda ex 、λ em The corresponding fluorescence intensity, in (a.u.); arp is the integral value of the Raman peak of ultrapure water.
The raman correction process and the data normalization process further sequentially comprise: removal of raman rayleigh scattering and tiling of data;
step (2-4), the method for removing the raman rayleigh scattering comprises the following steps:
removing the regions of Em < Ex+/-20 nm and Em >2 Ex+/-10 nm, inserting 0 value into the removed region for replacement, and retaining the most obvious region of fluorescence characteristics;
step (2-5), the tiling method of the data comprises the following steps:
expanding the corrected data along the direction of the excitation wavelength i, and connecting the data points between adjacent rows end to end; the samples are converted from a 47 x 341 matrix (to be confirmed) into a 1 x 16027 vector form, and the N samples are combined into an N x 16027 matrix form;
step (2-6), the normalization processing mode comprises:
and normalizing the matrix by using a mapmin max function to convert the sample into a row vector, wherein the normalization formula is as follows:
in the formula (3), x' represents the value of the single data of each sample characteristic, min is the minimum value of the sample characteristic data, and max is the maximum value of the sample characteristic data.
Step (2-6), in step (2), the dimension reduction method comprises the following steps:
the dimension reduction method comprises the steps of inputting a normalized N multiplied by 16027 data set into an LPP dimension reduction model to obtain dimension reduced data N multiplied by m, m being the number of samples and m being the feature quantity after dimension reduction; marking the data after dimension reduction according to the pollution source type.
(3) Inputting the training set into a support vector machine model for distinguishing the types of the pollution industry for training to obtain a pollution source three-dimensional fluorescence identification model for distinguishing the types of the pollution industry;
the method for training the model comprises the following steps:
step (3-1), training the training set classified in the step (2) by using a Libsvm tool box, wherein the number of samples of each training set is more than 20;
step (3-2), optimizing penalty parameters c and kernel function parameters g by adopting an SVMcgForclass function in the training process;
and (3-3) randomly selecting a plurality of unmodeled samples from the sewage samples of each polluted industry category collected in the step (1) as a prediction set, checking the recognition performance of the model, and supporting the recognition accuracy rate result of the vector machine to be shown in the figure 3.
(4) Collecting river water samples at different sampling points of a pollutant discharge river channel in a tracing area, and performing three-dimensional fluorescence spectrum scanning on the river water samples to obtain three-dimensional fluorescence spectrum data of each river water sample; the distance between sampling points is less than 300m.
(5) Screening out river water samples with abnormal three-dimensional fluorescence spectrum data according to the three-dimensional fluorescence spectrum data obtained in the step (4);
the method for screening the river water sample with abnormal three-dimensional fluorescence spectrum data is a parallel factor analysis method and a fluorescence data analysis method;
the parallel factor analysis method specifically comprises the following steps:
step (5-1), inputting the processed three-dimensional fluorescence spectrum data of the river point positions into a DOMFluor tool box in Matlab software for parallel factor analysis, and obtaining a fluorescence intensity load matrix A;
constructing a matrix B by using a fluorescence intensity load matrix A according to a formula (4), and calculating a leverage ratio L according to a formula (5);
L i ＝b ii i＝1,2,…,I (5)
in the formulas (4) and (5), L i The lever rate of the ith sample is calculated, bii is the main diagonal element of the matrix B, and I is the number of samples; matrix A is the fluorescence intensity load matrix of each component, A H Is a conjugate matrix of matrix A, (A) H A) + Is A H A pseudo-inverse matrix of A;
step (5-3), the leverage ratio is influenced by the selection of the factor number F, the factor number is required to be adjusted to observe a residual error matrix diagram formed by the factor number to determine the optimal factor number, after the factor number is increased, the component residual error value is not greatly changed and the residual error diagram is randomly distributed, if no special structure exists, the factor number is confirmed to be optimal before the factor number is not increased, the general factor number is 2-6, when the ith sample leverage ratio is more than 0.5, the point is used as a river channel abnormal point, and the point is further input into the model trained in the step (3) for recognition; the lever rate analysis result of the three-dimensional fluorescence parallel factor is shown in figure 4;
the fluorescence data analysis method specifically comprises the following steps:
step (5-I), finding out characteristic substances from the three-dimensional fluorescence spectrum data obtained in the step (4); the characteristic substance is tyrosine-like peak B, em=305 nm; ex=275 nm; tryptophan-like peak T, em=340 nm; ex=275 nm; humic substances a, em=400-460 nm, ex=260 nm; humic-like substances M, em=370-410 nm, ex=290-310 nm; humic substances C, em=420-460 nm, ex=320-360 nm; wherein Em is the emission wavelength of the substance to be measured, and Ex is the excitation wavelength of the substance to be measured;
step (5-II), calculating the numerical value of the characteristic index corresponding to the characteristic substance in the step (5-I); the characteristic indexes are fluorescence area integral, fluorescence index, humification index, biological index and freshness fluorescence parameters; determining the position of a fluorescence peak value according to the difference of the positions of different fluorescent groups at excitation wavelength and emission wavelength, extracting a peak value and a fluorescence integration area, and simultaneously calculating fluorescence parameters, wherein the extraction result of the fluorescence peak intensity is shown in figure 5;
the calculation of the fluorescence area integral is as follows:
the fluorescent region, region I (lambda) Ex <250nm，λ Em <350 nm), region II (lambda Ex <250nm，350<λ Em <380 nm), III region (lambda Ex <250nm，λ Em >380 nm), intermediate excitation wavelength in IV region (250 nm)<λ Ex <280nm，λ Em <380 nm), V (region) (lambda Ex >280nm，λ Em >380 nm), the divided areas are shown in figure 6, and can be adjusted according to the three-dimensional fluorescence characteristics of the actual water sample, and the integral formula of the fluorescence area corresponding to the ith area is as follows:
in formula (6), phi i For the volume integral of the i-th region,for the corresponding fluorescence intensity under a certain excitation emission wavelength dlambda em And dλ ex Is the differentiation of the excitation wavelength and the emission wavelength.
The fluorescent region is unevenly divided in the following manner: using multiplication factor MF i Correction is performed as follows:
s in (7) i For the area of the i-th block region, MF i For multiplication factor, the volume fraction of the modified i region is phi' i ＝φ i ×MF i The result of fluorescence area integral calculation is shown in figure 7;
the fluorescence index FI is calculated as follows:
I Em470 /I Em520 ，Ex＝370nm (8)
in formula (8) I Em470 And I Em520 For fluorescence intensities at emission wavelengths 480nm and 520nm at ex=370 nm excitation wavelength.
The humification index HIX is calculated as follows:
HIX＝∑I Em435-480 /∑I Em300-345 ,Ex＝245nm) (9)
sigma I in formula (9) Em435-480 And Sigma I Em300 345 is the integral of emission wavelengths 435-489nm and 300-345nm at excitation wavelengths ex=245 nm.
The self-biogenic index BIX is calculated as follows:
BIX＝I Em380 /I Ex430 ,Ex＝310nm (10)
in the formula (10), I Em380 And I Em430 For fluorescence intensities at emission wavelengths 380nm and 430nm at ex=310 nm excitation wavelength.
The freshness index calculation formula is as follows:
freshness=i Em380 /max(I Em420-435 ),Ex＝310nm (11)
In the formula (11) Em380 For a fluorescence intensity of 380nm emission wavelength at ex=310 nm excitation wavelength. max (I) Em420-435 ) For the maximum value of the emission wavelength of 420-435nm under the excitation wavelength of Ex=310 nm, the calculation result of fluorescence parameters is shown in the attached figure 8;
step (5-III), carrying out statistical difference analysis on the calculated characteristic index value by using a triple standard difference method to obtain an abnormal value; the river water sample with the abnormal value at the sampling point is regarded as an abnormal river water sample, and the abnormal point identification result is shown in figure 9;
(6) Carrying out Raman correction, data normalization, partial reservation projection algorithm dimension reduction and the like on the abnormal point position fluorescence data in the step (5), inputting the processed data into the identification model in the step (3) for identification, and locking a pollution source to obtain a final tracing result;
if the locked pollution enterprise is one, determining the enterprise as a pollution source; if the locked polluted enterprises are more than two, performing the step (8-1);
step (8-1), collecting the sewage samples of a plurality of locked polluted enterprises, carrying out three-dimensional fluorescence spectrum scanning on the sewage samples to obtain three-dimensional fluorescence spectrum data, and sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the obtained three-dimensional fluorescence spectrum data;
step (8-2), performing similarity matching on a plurality of locked polluted enterprises by adopting a formula (6) to obtain an included angle cosine value; the calculation method of the cosine value of the included angle is as follows
In the formula (6), a is a vector of 1×m after the three-dimensional fluorescent sample of the abnormal river sample screened in the step (5) is spread, b i 1 Xm vector after expansion for the ith locked enterprise three-dimensional fluorescence data, |a| is the modulus of the a vector, |b i The I is a modulus of bi vectors, cos theta is cosine values of the two vectors, and m is the characteristic dimension after expansion;
the cosine value range of the included angle is [0,1], the closer the acquired value is to 1, the more similar the two groups of data are represented, and the closest to 1 is obtained by comparing the obtained cosine values of different included angles through locking a plurality of polluted enterprises, and the polluted enterprises are judged as pollution sources.
Application example 1
A river channel organic matter pollution tracing method of a three-dimensional fluorescent LPP-SVM comprises the following specific steps:
three-dimensional fluorescence spectrum data of various water pollution sources in a certain area are collected to be used as machine learning materials, and the machine learning materials are classified according to the sources.
Pollution source information is detailed in the following table:
filtering a pollution source water sample after collection by using a Millipore filter membrane with the aperture of 0.22 mu m, scanning the sample at the room temperature of 25 ℃ in a machine, storing the sample to be detected in a refrigerator with the temperature of 4 ℃, setting the instrument parameters of three-dimensional fluorescence scanning to be in an Ex scanning range of 220-450 nm, an Em scanning range of 260-600 nm, setting the scanning bandwidth to be 5nm, setting the Ex step length to be 5nm, setting the Em step length to be 1nm, setting the slit width to be 5nm, and setting the scanning speed to be 2400nm/min;
the samples with higher concentration are diluted for multiple times with 5-fold gradient. The waste water should be collected during the normal production period of the enterprise.
And carrying out Raman correction on the obtained fluorescence spectrum data to eliminate Rayleigh and Raman scattering areas.
After scattering was removed, the fluorescence data for each sample was spread out in rows and converted from a 47×341 matrix form to a 1×16027 vector form.
N samples are combined into a matrix of n x 16027, labeled by pollution source category, labeled 1, 2, 3 …, each number representing a type of pollution source.
And carrying out normalization processing on the tidied data by using a mapmin max function.
The mapmin max function is given by
Where x' represents the value of a single data, min is the minimum value of the column in which the data is located, and max is the maximum value of the column in which the data is located.
And the LPP dimension reduction algorithm is adopted to reduce the dimension, the dimension error is 0.01, and 99% of characteristic information is reserved. After the dimension reduction, the feature number of each sample is changed from 16027 to 31.
The training process comprises the steps of training a training set by using a Libsvm tool box to realize training of a support vector machine, constructing a three-dimensional fluorescent pollution source identification model, and optimizing a punishment parameter c and a parameter g of a kernel function by using grid optimization;
and (3) arranging the points in the river channel A at intervals of 300m, wherein the number of the arranged points is 35, after the water sample at the monitoring points is collected, processing the three-dimensional fluorescence spectrum data of the obtained water sample at the river channel according to the pollution source data preprocessing step, carrying out Raman correction, and removing a Raman Rayleigh scattering part.
And extracting the fluorescence intensity value of the specific area, and calculating fluorescence parameters such as the percentage of fluorescent components, the fluorescence index, the humification index, the biological index and the like. The calculation results are shown in figures 4-6
By adopting parallel factor analysis, the maximum group fraction which can be analyzed by splitting is 3, and the number 16 point in the fitting process has higher leverage.
And (3) processing the data of the monitoring point position 16, inputting the processed data into a support vector machine model for recognition, setting the label to be 0, and finally recognizing the data as the chemical fiber dyeing and finishing industry, namely recognizing the data to be 1.
Two chemical fiber dyeing and finishing enterprises A and B are arranged around the point, and the similarity with the water sample at the point is 0.9 and 0.996 respectively, so that the enterprise B is more likely to be a pollution input source.
Claims (7)
1. The method for tracing the organic pollutants in the river channel based on the three-dimensional fluorescent LPP-SVM is characterized by comprising the following steps of:
(1) Collecting sewage of wading pollution enterprises in a region to be traced, and carrying out three-dimensional fluorescence spectrum scanning on sewage samples to obtain three-dimensional fluorescence spectrum data corresponding to each sample;
(2) Sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the acquired three-dimensional fluorescence spectrum data, and classifying the processed data according to the pollution industry category to obtain a classified training set;
(3) Inputting the training set into a support vector machine model for distinguishing the types of the pollution industry for training to obtain a pollution source three-dimensional fluorescence identification model for distinguishing the types of the pollution industry;
(4) Collecting river water samples at different sampling points of a pollutant discharge river channel in a tracing area, and performing three-dimensional fluorescence spectrum scanning on the river water samples to obtain three-dimensional fluorescence spectrum data of each river water sample;
(5) Screening out river water samples with abnormal three-dimensional fluorescence spectrum data according to the three-dimensional fluorescence spectrum data obtained in the step (4);
(6) Sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the three-dimensional fluorescence spectrum data of the abnormal river sample screened in the step (5) to obtain sample data to be traced;
(7) Inputting the sample data to be traced in the step (6) into the pollution source three-dimensional fluorescent identification model in the step (3) to obtain pollution industry types matched with the sample to be traced;
(8) And (3) locking a pollution enterprise from the area to be traced according to the pollution category in the step (7).
2. The method for tracing organic pollutants in a river channel based on three-dimensional fluorescent LPP-SVM as set forth in claim 1, wherein in step (2), the raman correction process includes the steps of:
step (2-1), carrying out three-dimensional fluorescence spectrum scanning on the ultrapure water to obtain three-dimensional fluorescence spectrum data of the ultrapure water;
step (2-2) of calculating a Raman peak integrated value A of the ultrapure water by using the formula (1) rp The calculation formula is as follows:
(1),for a specific lambda ex Lower corresponding lambda em Raman integral value of (a); lambda (lambda) ex Represents the excitation wavelength; lambda (lambda) em Representing the emission wavelength; d represents the integral symbol, < >>Is a certain lambda ex Lower lambda em The measured fluorescence intensity of the Raman spectrum; />And->Starting and ending points of an integration interval;
step (2-3) of dividing all the three-dimensional fluorescence spectrum data obtained in step (1) by A of ultrapure water rp The method comprises the steps of obtaining spectral data of a sewage sample in units of (R.U.), wherein the calculation formula is as follows:
in the formula (2), the amino acid sequence of the compound,is of arbitrary lambda ex 、λ em The corresponding corrected data, i.e. fluorescence intensity in raman (r.u.); />To correct for the former arbitrary lambda ex 、λ em The corresponding fluorescence intensity, in (a.u.); arp is the integral value of the Raman peak of ultrapure water.
3. The method for tracing organic pollutants in river channels based on three-dimensional fluorescent LPP-SVM according to claim 1, wherein in step (2), the steps between Raman correction processing and data normalization processing further comprise: removal of raman rayleigh scattering and tiling of data;
step (2-4), the method for removing the raman rayleigh scattering comprises the following steps:
removing the regions of Em < Ex+/-20 nm and Em >2 Ex+/-10 nm, inserting 0 value into the removed region for replacement, and retaining the most obvious region of fluorescence characteristics;
step (2-5), the tiling method of the data comprises the following steps:
expanding the corrected data along the direction of the excitation wavelength i, and connecting the data points between adjacent rows end to end; the method comprises the steps of converting a matrix of a×b into a vector form of 1×m, merging N samples into a matrix form of n×m, wherein N is the number of samples, M=a×b, M is a characteristic quantity, a is the number of rows for acquiring three-dimensional fluorescent matrix data, and b is the number of columns for acquiring the three-dimensional fluorescent matrix data;
step (2-6), the normalization processing mode comprises:
and normalizing the matrix by using a mapmin max function to convert the sample into a row vector, wherein the normalization formula is as follows:
in the formula (3), x' represents the value of the single data of each sample characteristic, min is the minimum value of the sample characteristic data, and max is the maximum value of the sample characteristic data.
4. The method for tracing organic pollutants in river channels based on three-dimensional fluorescent LPP-SVM as set forth in claim 1, wherein in step (2), the method for reducing dimension includes:
and inputting the normalized NxM data into an LPP dimension reduction model to obtain dimension reduced data Nxm and M < M >, wherein N is the number of samples, M is the characteristic quantity, and M is the characteristic quantity after dimension reduction.
5. The method for tracing organic matter pollution of a river channel based on three-dimensional fluorescent LPP-SVM as set forth in claim 1, wherein in step (3), the method for training the model includes:
step (3-1), training the training set classified in the step (2) by using a Libsvm tool box, wherein the number of samples of each training set is more than 20;
step (3-2), optimizing penalty parameters c and kernel function parameters g by adopting an SVMcgForclass function in the training process;
and (3-3) randomly selecting a plurality of unmodeled samples from the sewage samples of each pollution industry category collected in the step (1) as a prediction set, and checking the recognition performance of the model.
6. The method for tracing organic matter pollution of a river channel based on the three-dimensional fluorescent LPP-SVM according to claim 1, wherein in the step (5), a parallel factor analysis method or a fluorescent data analysis method is adopted as the method for screening river water samples with abnormal three-dimensional fluorescent spectrum data;
the parallel factor analysis method comprises the following steps:
step (5-1), inputting the three-dimensional fluorescence spectrum data of the river water sample in the step (4) into a DOMFluor tool box in Matlab software for parallel factor analysis, and obtaining a fluorescence intensity load matrix A;
constructing a matrix B by using a fluorescence intensity load matrix A according to a formula (4), and calculating a leverage ratio L according to a formula (5);
L i ＝b ii i＝1,2,…,I (5)
in the formulas (4) and (5), L i Leverage for the ith sample, b ii The matrix B is a main diagonal element, and I is the number of samples; matrix A is the fluorescence intensity load matrix of each component, A H Is a conjugate matrix of matrix A, (A) H A) + Is A H A pseudo-inverse matrix of A;
step (5-3), when the i-th sample leverage ratio is more than 0.5, identifying the river water sample of the sampling point as an abnormal river water sample;
the fluorescence data analysis method comprises the following steps:
step (5-I), finding out characteristic substances from the three-dimensional fluorescence spectrum data obtained in the step (4); the characteristic substance is tyrosine-like peak B, em=305 nm; ex=275 nm; tryptophan-like peak T, em=340 nm; ex=275 nm; humic substances a, em=400-460 nm, ex=260 nm; humic-like substances M, em=370-410 nm, ex=290-310 nm; humic substances C, em=420-460 nm, ex=320-360 nm; wherein Em is the emission wavelength of the substance to be measured, and Ex is the excitation wavelength of the substance to be measured;
calculating the numerical value of the characteristic index corresponding to the characteristic substance in the step (5-I), wherein the characteristic index is a fluorescence area integral, a fluorescence index, a humification index, a biological index, freshness and a fluorescence parameter;
step (5-III), carrying out statistical difference analysis on the calculated characteristic index value by using a triple standard difference method to obtain an abnormal value; and identifying the sampling point river water sample with the abnormal value as an abnormal river water sample.
7. The method for tracing organic matter pollution of a river channel based on a three-dimensional fluorescent LPP-SVM according to claim 1, wherein in the step (8), if the locked pollution enterprise is one, the enterprise is determined to be a pollution source; if the locked polluted enterprises are more than two, performing the step (8-1);
step (8-1), collecting the sewage samples of a plurality of locked polluted enterprises, carrying out three-dimensional fluorescence spectrum scanning on the sewage samples to obtain three-dimensional fluorescence spectrum data, and sequentially carrying out Raman correction processing, data normalization processing and dimension reduction processing on the obtained three-dimensional fluorescence spectrum data;
step (8-2), performing similarity matching on the locked multiple polluted enterprises by adopting a formula (6) to obtain an included angle cosine value; the calculation method of the cosine value of the included angle is as follows
In the formula (6), a is a vector of 1×m after the three-dimensional fluorescent sample of the abnormal river sample screened in the step (5) is spread, b i 1 Xm vector after expansion for the ith locked enterprise three-dimensional fluorescence data, |a| is the modulus of the a vector, |b i The I is a modulus of bi vectors, cos theta is cosine values of the two vectors, and m is the characteristic dimension after expansion;
the cosine value range of the included angle is [0,1], and the closer the obtained value is to 1, the more similar the two groups of data are represented; and comparing the cosine values of different included angles obtained by the locked multiple polluted enterprises to obtain the polluted enterprise corresponding to the cosine value of the included angle closest to 1, and judging the polluted enterprise as a pollution source.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311245608.6A CN117309831A (en) | 2023-09-25 | 2023-09-25 | Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311245608.6A CN117309831A (en) | 2023-09-25 | 2023-09-25 | Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117309831A true CN117309831A (en) | 2023-12-29 |
Family
ID=89284303
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311245608.6A Pending CN117309831A (en) | 2023-09-25 | 2023-09-25 | Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117309831A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522653A (en) * | 2024-01-05 | 2024-02-06 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
| CN117783088A (en) * | 2024-02-23 | 2024-03-29 | 广州贝拓科学技术有限公司 | Control model training method, device and equipment of laser micro-Raman spectrometer |
| CN118013258A (en) * | 2024-04-08 | 2024-05-10 | 山东星芭克生物科技有限公司 | Information acquisition method of intelligent water-soluble fertilizer production line |
-
2023
- 2023-09-25 CN CN202311245608.6A patent/CN117309831A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522653A (en) * | 2024-01-05 | 2024-02-06 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
| CN117522653B (en) * | 2024-01-05 | 2024-03-26 | 同济大学 | Rain and sewage hybrid joint traceability analysis method based on three-dimensional fluorescence and municipal pipe network BIM model |
| CN117783088A (en) * | 2024-02-23 | 2024-03-29 | 广州贝拓科学技术有限公司 | Control model training method, device and equipment of laser micro-Raman spectrometer |
| CN117783088B (en) * | 2024-02-23 | 2024-05-14 | 广州贝拓科学技术有限公司 | Control model training method, device and equipment of laser micro-Raman spectrometer |
| CN118013258A (en) * | 2024-04-08 | 2024-05-10 | 山东星芭克生物科技有限公司 | Information acquisition method of intelligent water-soluble fertilizer production line |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117309831A (en) | Pollution tracing method for river channel organic matters based on three-dimensional fluorescent LPP-SVM | |
| US11965871B2 (en) | Method and system for intelligent source tracing of organic pollution of water body | |
| Azcarate et al. | Modeling second-order data for classification issues: Data characteristics, algorithms, processing procedures and applications | |
| US11249023B2 (en) | Rapid detection method for condition of landfill leachate polluting groundwater and application thereof | |
| CN113011478A (en) | Pollution source identification method and system based on data fusion | |
| CN110554013B (en) | Method for realizing rapid identification and comparison by using fluorescence spectrum characteristic information | |
| CN110702648B (en) | Fluorescent spectrum pollutant classification method based on non-subsampled contourlet transformation | |
| CN113267462A (en) | Method for characterizing organic pollution characteristics of water body by applying ultraviolet spectrum | |
| CN105334186A (en) | Infrared spectral analysis method | |
| CN117309838A (en) | Industrial park water pollution tracing method based on three-dimensional fluorescence characteristic data | |
| CN112098361A (en) | Corn seed identification method based on near infrared spectrum | |
| CN114184599B (en) | Single-cell Raman spectrum acquisition number estimation method, data processing method and device | |
| CN116595461A (en) | Rain inlet sunny-day pollution discharge tracing method based on random forest identification | |
| CN114563381A (en) | Water body pollution tracing method | |
| CN112630202A (en) | Method for identifying source of overflowing sewage in rainy days of urban drainage system | |
| CN107247033B (en) | Identify the method for Huanghua Pear maturity based on rapid decay formula life cycle algorithm and PLSDA | |
| CN116263444B (en) | High-resolution mass spectrum non-targeted analysis water pollution source identification and tracing method | |
| CN117331913A (en) | Construction method of industrial park water pollution source fluorescence fingerprint database | |
| CN116858822A (en) | Quantitative analysis method for sulfadiazine in water based on machine learning and Raman spectrum | |
| CN116297936A (en) | Method, system, device and medium for tracing pollutant | |
| CN115575363B (en) | Method and system for acquiring ecological influence mechanism | |
| Burlacu et al. | Convolutional Neural Network detecting synthetic cannabinoids | |
| CN113295673B (en) | Laser-induced breakdown spectroscopy weak supervision feature extraction method | |
| CN113311081B (en) | Pollution source identification method and device based on three-dimensional liquid chromatography fingerprint | |
| CN113406037B (en) | Infrared spectrum online rapid identification analysis method based on sequence forward selection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |