CN115236045A

CN115236045A - Method and device for determining chemical oxygen demand in water by using fluorescence spectrometry and storage medium

Info

Publication number: CN115236045A
Application number: CN202210798150.6A
Authority: CN
Inventors: 魏峨尊; 高贝贝; 贺颖
Original assignee: Hua Xia An Jian Wu Lian Technology Qingdao Co ltd
Current assignee: Hua Xia An Jian Wu Lian Technology Qingdao Co ltd
Priority date: 2022-07-01
Filing date: 2022-07-01
Publication date: 2022-10-25
Also published as: CN115236045A8

Abstract

A method, apparatus and storage medium for determining chemical oxygen demand in water by using fluorescence spectrometry, the method obtains the fluorescence spectrum data of the water sample to be measured; analyzing and acquiring spectral data of each component; calculating the concentration of the component using spectral data analysis; and calculating the COD value corresponding to the component concentration according to the component concentration, and acquiring the COD value of the water sample to be detected according to the COD value of the corresponding concentration of each component. The method includes constructing a model for calculating the concentration of the constituent based on the constituent spectral data. The method comprises a method for obtaining the COD value based on the component concentration. The method realizes the rapid determination of the chemical oxygen demand in the water environment through fluorescence spectroscopy, and obtains the information of main pollutants constituting the water body.

Description

Method and device for determining chemical oxygen demand in water by using fluorescence spectrometry and storage medium

Technical Field

The invention belongs to the technical field of optical means testing or water environment analysis in physical testing, and particularly relates to a method for determining chemical oxygen demand in water by using a fluorescence spectrometry.

Background

The Chemical Oxygen Demand (COD) is an important organic matter pollution parameter which is an index for measuring the content of organic matters in water by measuring the amount of reducing matters needing to be oxidized in a water sample by a chemical method. The traditional COD determination method needs to perform determination through chemical reagents such as permanganate and dichromate, for example, the environmental protection standard of determination of chemical oxygen demand for water quality (dichromate method for determining chemical oxygen demand) issued by the original environmental protection department of 5, 1 and 5 months in 2017 (HJ 828-2017) stipulates that potassium dichromate is used as a test reagent, and a mercuric sulfate solution is used for removing chloride interference. According to incomplete statistics, the amount of mercury discharged to the environment by waste liquid generated by COD measurement in China is measured by tons each year. Although the sampling amount is reduced by the new national standard, the COD measurement is based on a chemical method, some chemical reagents which can cause secondary pollution are inevitably used, and in addition, the COD measurement by the chemical method has long time consumption, large equipment maintenance difficulty and high cost, and is inconvenient for long-time unattended monitoring. With the higher and higher requirements of relevant national departments on the refinement of water quality regulation, people pay attention to the COD value in the water body and hope to know the main pollutants in the water body and the source and change rules thereof, but the existing national standard COD measurement method cannot identify the main pollutants in the water body.

In order to solve the above problems, people try to obtain the COD value in water by inversion calculation by using the absorption spectrum of the organic compounds in the water in the visible ultraviolet region, and a certain result is obtained. In practice, the following problems have been found to exist in calculating COD in water by using visible ultraviolet spectroscopy: on one hand, the single potassium hydrogen phthalate substance is used as a standard substance for determination, and the ultraviolet visible absorption spectrum of the single potassium hydrogen phthalate substance cannot completely represent or reflect organic matters in the actual water body, so that the accuracy of the result is influenced. On the other hand, the turbidity of the water body seriously interferes with the visible ultraviolet absorption spectrum data. In order to eliminate the interference, the turbidity in the water body is often measured, the interference of the turbidity to the absorption spectrum is deducted, and then the relation between the absorption spectrum and the COD of the water body is established. In addition, for a data set acquired by using a spectroscopy, since the number of information variables (sample dimension) for acquiring a certain wavelength spectrum is often greater than the number of samples, and sometimes the number of sample data is even less than the sample dimension, the data matrix cannot be inverted, and thus a good enough standard linear model cannot be obtained.

In view of the defects in the prior art, the technical problems to be solved by the invention are that part of reagents of the national standard monitoring method are toxic and harmful, secondary pollution is easy to cause, and the operation and maintenance cost of online equipment is high; the COD detection by the ultraviolet-visible spectrum method can be monitored on line in real time, but has low precision and limited accuracy, and has poor applicability to different water qualities, particularly water qualities with high turbidity.

Disclosure of Invention

Aiming at the defects of the prior art, the invention aims to provide a method for determining the chemical oxygen demand in water by using a fluorescence spectrometry, the method uses a fluorescence spectrometry determination mode, performs spectrum recognition and characterization on overlapped objects of fluorescence spectra in a multi-component system by a parallel factor decomposition method, determines specific components of a water sample to be determined as standard substances for COD determination by a water quality analysis method or by combining the characteristics of the water sample to be determined, and constructs a regression model of the fluorescence spectra of the component solution and the concentration of the solution aiming at the specific components. And the regression model is improved by using a partial least square regression method, a ridge regression method, an elastic network regression method, an LASSO regression method and the like, so that errors caused by analysis methods such as a parallel factor decomposition method and the like are further reduced. The calculated concentration values of the component solutions were used based on the improved regression model. The contribution of each component solution to the COD value of the mixed solution is obtained based on the concentration value of the component solution. The COD contributions of the component solutions were summed to obtain the COD of the mixed solution.

In order to achieve the above object, the method for determining chemical oxygen demand in an aqueous environment by fluorescence spectroscopy according to the present invention comprises:

(1) A method for determining chemical oxygen demand in water by fluorescence spectroscopy, comprising:

acquiring fluorescence spectrum data of a water sample to be detected;

separating the fluorescence spectrum data of the water sample to be detected by using a parallel factor method or a non-negative matrix analysis method to obtain the fluorescence spectrum data of the specific component of the water sample to be detected, and analyzing;

based on the concentration of the specific component of each water sample to be detected and a fluorescence spectrum regression model, acquiring the concentration of the specific component of each water sample to be detected in the water sample to be detected by using fluorescence spectrum data of the specific component of each water sample to be detected;

based on the relation between the concentration of the specific component of each water sample to be detected and the chemical oxygen demand, acquiring the chemical oxygen demand of the specific component of the water sample to be detected by using the concentration of the specific component of each water sample to be detected;

acquiring the chemical oxygen demand of each water sample to be detected based on the chemical oxygen demand of specific components of the water sample to be detected;

the specific component of the water sample to be detected is at least one component with the highest chemical oxygen demand contribution rate in the water sample to be detected or a component selected based on the characteristics of the water sample to be detected;

the concentration of the specific component of each water sample to be detected and the fluorescence spectrum regression model are constructed by a principal component analysis method, and the specific construction method comprises the following steps:

preparing m mixed standard solutions with different concentrations by using k specific component substances of all selected water samples to be detected; constructing a determination matrix according to the fluorescence intensities of m mixed standard solutions with different concentrations at n different wavelengths; constructing a concentration matrix according to the concentrations of the k selected main components in the mixed standard solution in the m mixed standard solutions; performing linear regression on the measurement matrix and the concentration matrix; extracting component pairs one by one from the standardized measuring matrix and the standardized concentration matrix by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method, and performing principal component analysis to calculate the number of principal components; based on principal component analysis and linear regression of the determination matrix and the concentration matrix, obtaining a coefficient matrix by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method; and constructing a regression model based on constant terms in the coefficient matrix and the original variable coefficients.

(2) The method for determining COD in water by fluorescence spectrometry as described in item (1), wherein the components having a COD contribution ratio of not less than 95% in the sample to be tested are used as the specific components of the sample to be tested.

(3) The method for determining chemical oxygen demand in water by fluorescence spectrometry according to any one of items (1) to (2), wherein the sample of water to be measured is surface water, and at least one of humic substances, fulvic acids substances, proteinoid substances or amino acid-like substances is selected as the specific component of the sample of water to be measured.

(4) The method for determining chemical oxygen demand in water by fluorescence spectrometry according to any one of (1) to (3), wherein at least one of Fluorescent Dissolved Organic substances (FDOM) is selected as the specific component of the water sample to be measured.

(5) The method for measuring chemical oxygen demand in water by fluorescence spectrometry according to any one of items (1) to (4), wherein the relationship between the concentration of the specific component of the water sample to be measured and the chemical oxygen demand is obtained based on a linear fit equation of the concentration of the specific component of the water sample to be measured and the chemical oxygen demand, which is measured in advance, or on data obtained in advance on the relationship between the concentration of the specific component and the chemical oxygen demand.

(6) The method for determining chemical oxygen demand in water by fluorescence spectrometry according to any one of items (1) to (5), wherein after the spectral data of the water sample to be measured is acquired, the spectral data of the water sample to be measured is corrected by removing Rayleigh scattering or/and Raman scattering.

(7) The method for determining chemical oxygen demand in water by using fluorescence spectrometry according to any one of items (1) to (6), wherein a water sample to be tested is subjected to quantitative analysis, and if the composition components of the current water sample to be tested are changed compared with a specific water sample to be tested when the specific components of the water sample to be tested are determined before, the specific components of the water sample to be tested are adjusted or the change of the composition components of the water sample to be tested is prompted.

(8) The method for determining chemical oxygen demand in water by using fluorescence spectrometry according to any one of (1) to (7), wherein the maximum fluorescence peak value of the spectral data of the water sample to be measured and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value are obtained; acquiring the maximum fluorescence peak value of each component spectral data and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value; and quantitatively analyzing the acquired spectral data of the water sample to be detected, the maximum fluorescence peak value of the spectral data of each component, and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value.

(9) The method for determining chemical oxygen demand in water by using fluorescence spectrometry according to any one of items (1) to (8), wherein the fluorescence spectrum data of a specific component of the water sample to be measured is obtained by separating the fluorescence spectrum data of the water sample to be measured using a parallel factor method or a non-negative matrix analysis method.

(10) The method for determining chemical oxygen demand in water by using fluorescence spectrometry according to any one of items (1) to (9), wherein the components of the water sample to be tested, the percentage of each specific component in the total components, and the fluorescence spectrum data of each specific component are obtained by analyzing the spectrum data of the water sample to be tested using a nuclear consistency diagnostic analysis method or a Tucker coefficient analysis.

(11) An apparatus for measuring chemical oxygen demand in water by fluorescence spectroscopy, which can perform the method according to any one of (1) to (10), comprising:

the spectrum data separation module is used for acquiring fluorescence spectrum data of the water sample, and separating and acquiring the fluorescence spectrum data of the components of the water sample by using a parallel factor method or a non-negative matrix analysis method and analyzing the fluorescence spectrum data;

the composition analysis module is used for acquiring fluorescence spectrum data of the water sample, acquiring fluorescence spectrum data and analysis data of composition components of the water sample from the spectrum data separation module, and selecting at least one composition component with the highest chemical oxygen demand contribution rate in the water sample to be detected as a specific component of the water sample to be detected or selecting the specific component of the water sample to be detected based on the characteristics of the water sample to be detected;

the regression model building module is used for configuring a specific component sample solution of the water sample to be detected by using specific component substances of the water sample to be detected selected by the component analysis module, building a regression model of concentration and fluorescence spectrum by using component spectrum data and component concentrations of sample solutions with different concentrations and based on a principal component analysis method, and improving the regression model of concentration and fluorescence spectrum by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method;

the concentration analysis module acquires the fluorescence spectrum data of the specific component of the water sample to be detected from the spectrum data separation module, and analyzes the concentration of the specific component in the water sample to be detected based on the concentration of the specific component of the water sample to be detected acquired by the autoregressive model construction module and the fluorescence spectrum regression model;

and the chemical oxygen demand analysis module is used for receiving the concentrations of the specific components of the water samples to be detected analyzed by the concentration analysis module, analyzing and obtaining the chemical oxygen demand of the specific components of the water samples to be detected according to the relationship between the concentrations of the specific components of the water samples to be detected and the chemical oxygen demand, and summing the chemical oxygen demands of the specific components of the water samples to be detected to obtain the chemical oxygen demand of the water samples to be detected.

(12) The apparatus for determining chemical oxygen demand in a water environment by using fluorescence spectrometry as described in item (11), wherein the composition analysis module selects a composition having a chemical oxygen demand contribution rate of not less than 95% in the water sample to be measured as the specific component of the water sample to be measured.

(13) The apparatus for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy as set forth in any one of (11) to (12), wherein the composition analyzing module selects at least one of Fluorescent Dissolved Organic Matter (FDOM) as the specific component of the sample water to be measured.

(14) The apparatus for determining chemical oxygen demand in an aqueous environment by using fluorescence spectrometry according to any one of items (11) to (13), wherein the aqueous sample to be measured is surface water, and the composition analysis module selects at least one of humic substances, fulvic acids substances, proteinoid substances, or amino acid-like substances as the specific component of the aqueous sample to be measured.

(15) The apparatus for determining chemical oxygen demand in a water environment by using fluorescence spectroscopy as described in any one of (11) to (14), wherein the composition analysis module comprises a composition adjustment module for obtaining a maximum fluorescence peak value of spectral data of a water sample to be measured and an excitation wavelength and an emission wavelength corresponding to the maximum fluorescence peak value; acquiring the maximum fluorescence peak value of each component of spectral data and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value; and quantitatively analyzing the acquired spectral data of the water sample to be detected, the maximum fluorescence peak value of the spectral data of each component, and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value, wherein if the composition of the current water sample to be detected changes compared with the specific water sample to be detected when the specific composition of the water sample to be detected is determined before, the composition division analysis module adjusts the specific composition of the water sample to be detected or prompts the condition that the composition of the water sample to be detected changes.

(16) The apparatus for determining chemical oxygen demand in a water environment by using fluorescence spectrometry according to any one of items (11) to (15), wherein the apparatus comprises a component concentration and fluorescence spectrum regression model information database, the component concentration and fluorescence spectrum regression model of the specific water sample to be determined is stored, and the concentration analysis module can retrieve from the component concentration and fluorescence spectrum regression model information database when the specific water sample to be determined is determined.

(17) The apparatus for determining chemical oxygen demand in a water environment by using fluorescence spectrometry according to any one of items (11) to (16), wherein the apparatus comprises a device for determining a relationship between concentration and chemical oxygen demand, and a device for determining a relationship between concentration of a specific component of a water sample to be determined and chemical oxygen demand, and the relationship is used for the chemical oxygen demand analysis module to call.

(18) The apparatus for determining chemical oxygen demand in a water environment using fluorescence spectroscopy according to any one of claims (11) to (17), wherein the apparatus comprises a component concentration and chemical oxygen demand information database, and stores a relationship between component concentrations and chemical oxygen demand for calling by a chemical oxygen demand analysis module.

(19) The apparatus for measuring chemical oxygen demand in an aqueous environment by fluorescence spectrometry according to any one of items (11) to (18), wherein a fluorescence spectrometry apparatus is provided for acquiring fluorescence spectrum data of a water sample to be measured.

(20) The apparatus for measuring chemical oxygen demand in an aqueous environment using fluorescence spectroscopy as described in any one of (11) to (19), wherein a correction module is included to correct fluorescence spectrum data by removing rayleigh scattering or/and raman scattering.

(21) In another aspect, the present invention provides an apparatus for determining chemical oxygen demand in water by fluorescence spectroscopy, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements all or part of the steps of any one of (1) to (10) of the method for determining chemical oxygen demand in water environment by fluorescence spectroscopy.

(22) Another aspect of the present invention provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements all or part of the steps of the method for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy according to any one of (1) to (10).

The invention adopts fluorescence spectrum data to carry out chemical oxygen demand measurement, and the measurement sensitivity of the fluorescence spectrum measurement to soluble organic matters (DOM) is 10-1000 times that of the conventional ultraviolet-visible spectrum measurement, so that abundant fluorescence information of organic pollutants can be obtained, and the fluorescence spectrum measurement is increasingly emphasized in water quality monitoring. Since the fluorescence spectrum is used for measuring the excitation spectrum and the emission spectrum of the sample in the direction perpendicular to the incident light, rather than obtaining the light transmittance from the direction of the incident light, the influence of the turbidity of the solution on the absorbance is eliminated to a certain extent.

Secondly, the invention adopts specific components in the water sample to be detected selected by analysis as standard substances, or selects and uses the humoid and the proteinoid as the standard substances according to the properties of the water sample to be detected such as surface water, reservoir water and the like, and can more accurately realize the change rule of organic matters in the water body by establishing the relationship between the specific components of the water sample to be detected, such as fluorescence spectrum data of the humoid and the proteinoid and the like, and COD (chemical oxygen demand) of the water body, thereby providing detailed basis for fine management of water quality change and avoiding secondary pollution and waste of manpower and material resources caused by the addition of chemical reagents.

Thirdly, the invention adopts mathematical methods such as parallel factorization and the like to complete spectrum identification and characterization of the overlapping objects of the fluorescence spectra in the multi-component system. The soluble organic matter in the water environment of the tracer sample is analyzed by using a parallel factor, and various fluorescent components contained in a sample excitation-emission matrix (EEMs) are calculated, and the number of the fluorescent components is verified by a nuclear Consistency Diagnostic (Core Consistency Diagnostic) analysis method or a Tucker coefficient (Tucker Congreence coeffients). The fluorescence intensity of each fluorescent fraction in each sample is expressed in F _ max (RU), and the ratio of F _ max of each fraction to the sum of F _ max of all fractions is approximately considered to be consistent with the ratio of fractions in DOM.

And secondly, constructing a regression model by EEMs of all components of the mixed solution through parallel factor analysis, and improving the regression model by using a partial least square regression method, a ridge regression method, an elastic network regression method, a LASSO regression method and the like to further reduce errors caused by analysis methods such as a parallel factor decomposition method and the like. Based on the established fluorescence spectrum data of each component solution and the regression equation of the solution concentration, the COD value of the solution can be further obtained according to the relation between the solution concentration and the COD value through the solution concentration value obtained by calculation. Since the COD of the mixed solution is considered to be a contribution from each component in the mixed solution, if each component and the component concentration of the mixed solution are known, the COD value of the mixed solution can be determined from the concentration of each component.

In addition, when the water sample to be detected is detected and analyzed, quantitative analysis is carried out on the basis of the spectral data of the water sample to be detected and the spectral data of the components, and whether the components of the water sample to be detected have changes compared with the previous determination is detected. If the change exists, the specific components of the water sample to be detected are adjusted or early-warning is carried out.

Therefore, at least the following beneficial technical effects can be achieved by adopting the method for measuring the chemical oxygen demand in the water body by using the fluorescence spectrum:

1. the invention realizes the method for monitoring COD in the water body by using the fluorescence spectrum, has high precision and a large amount of information, and is suitable for the requirements of various water quality fine management.

2. The invention has the advantages of rapid monitoring, no chemical reagent consumption, convenient use of unattended stations for a long time and the like.

Drawings

FIG. 1 is a flow chart of a first embodiment of a method, an apparatus and a storage medium for determining chemical oxygen demand in water by fluorescence spectroscopy according to the present invention.

FIG. 2 is a schematic diagram of a method, an apparatus and a storage medium for determining chemical oxygen demand in water by fluorescence spectrometry according to the present invention.

FIG. 3 is a fluorescence spectrum of a mixed solution of tryptophan, tyrosine and fulvic acid configured during the construction of a regression model in a fifth embodiment of the method, apparatus and storage medium for determining chemical oxygen demand in water by fluorescence spectroscopy according to the present invention.

FIG. 4 is a fluorescence spectrum of a mixed solution of tryptophan, tyrosine and fulvic acid, which is prepared during the construction of a regression model, after parallel factorization, in a fifth embodiment of a method, an apparatus and a storage medium for determining chemical oxygen demand in water by fluorescence spectroscopy.

Detailed Description

The technical means adopted by the invention to achieve the preset purpose are further described below by combining the accompanying drawings and the preferred embodiments of the invention.

As shown in fig. 1, the present invention provides a first embodiment of a method for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy.

The first embodiment is as follows:

summary of the examples a method:

analyzing the determined specific components of the water sample to be detected, and performing COD determination by using the specific components of the water sample to be detected as standard substances in COD detection of the water sample to be detected;

establishing a regression model of the solution fluorescence spectrum and the concentration of a specific component of the water sample to be detected, and establishing or obtaining a fitting equation of the concentration of the component solution and COD;

acquiring fluorescence spectrum data of a water sample to be detected;

separating and acquiring fluorescence spectrum data of specific components of the water sample to be detected based on the fluorescence spectrum data of the water sample to be detected, and analyzing the fluorescence spectrum data;

acquiring the chemical oxygen demand of the specific component of each water sample to be detected based on the concentration of the specific component of each water sample to be detected;

and acquiring the chemical oxygen demand of the water sample to be detected based on the chemical oxygen demand of the specific component of each water sample to be detected.

A first preparation step: analyzing and determining specific components of a water sample to be detected for determining the chemical oxygen demand:

and (I) analyzing the components of the water sample to be detected, and determining the specific components of the water sample to be detected.

In this example, the parallel factor analysis method is used as an example, and the component analysis can be performed by any conventional component analysis method.

The method for analyzing the components in this example is as follows:

1.1, filtering a water sample to be detected with a certain volume by a 0.45-micron filter membrane, and scanning on a fluorescence spectrometer to obtain three-dimensional fluorescence spectrum data of the water sample to be detected;

1.2 the fluorescence spectrometer sets the scanning parameters of the water sample to be that the scanning range of the excitation wavelength is 220-500nm, the excitation bandwidth is 5nm, the scanning range of the emission wavelength is 250-650nm, the emission bandwidth is 5nm, the scanning speed is 10000nm/min, and the voltage is 900V;

1.3, according to a Rayleigh scattering formula, deducting first-level and second-level Rayleigh scattering, or deducting Rayleigh scattering and Raman scattering interference by matlab2018a and a tool box DOMFluor to obtain correction data of the water sample to be detected;

1.4, carrying out parallel factor (PARAFAC) analysis and calculation on the corrected data matrix, and verifying by a Core Consistency Diagnostic analysis method or a Tucker coefficient (Tucker Consistency Coefficients) to obtain the component number of the water sample solution to be detected, the data matrix of each component and a fluorescence spectrum;

1.5 determining the composition of the water sample to be detected according to the fluorescence spectrum characteristics of each component.

And (3) preparing a simulation solution for simulating the water sample to be detected based on the determined composition simulation, and determining a component with the total contribution rate of Chemical Oxygen Demand (COD) not less than 95% to the water sample to be detected as the specific component of the water sample to be detected in a manner of determining the recovery rate of the solution sample.

(II) determining specific components of the water sample to be detected based on the property of the water sample to be detected

The protein-like substances are often related to life activities, represent the influence of human activities to a certain extent, and have a certain correlation with COD. The fluorescence spectra of surface waters, particularly water sources, rivers and lakes, are mainly attributed to humoid and proteinoid substances. When the water sample to be detected is surface water, at least one of humic substances, fulvic acid substances, proteinoid substances or amino acid substances is selected as a specific component of the water sample to be detected.

In addition, when measuring the chemical oxygen demand to be measured by using fluorescence spectroscopy, a Fluorescent Dissolved Organic Matter (FDOM) in a water sample to be measured can be selected as a specific component.

The selection of the specific components of the water sample to be detected can also be carried out by the method (I) and the method (II), and when the components with the chemical oxygen demand contribution rate not less than 95% of the water sample to be detected are selected as the specific components of the water sample to be detected, the specific components are selected by preferentially combining the properties of the water sample to be detected. For example, FDOM is preferably selected as the specific component, or when the sample to be tested is surface water, humic substance, fulvic acid substance, proteinoid substance or amino acid substance is preferably selected as the specific component of the sample to be tested.

A second preparation step: construction of regression model of specific component concentration and fluorescence spectrum of water sample to be detected

Taking a certain amount of k specific component substances of the water sample to be detected, preparing a specific component sample solution of the water sample to be detected, diluting and mixing the specific component sample solution of the water sample to be detected according to a certain proportion, and obtaining m series of water sample samples with different concentrations.

And measuring the excitation spectrum and the emission spectrum of the series of water samples by using a fluorescence spectrometer, forming an excitation-emission spectrum data matrix by using the excitation spectrum and the emission spectrum, and deducting the Raman scattering interference by using the excitation-emission spectrum matrix of the ultrapure water.

And (3) deducting first-order and second-order Rayleigh scattering according to a Rayleigh scattering formula or deducting Rayleigh scattering and Raman scattering interference by using matlab2018a and a tool box DOMFluor to obtain a correction data matrix of the water sample.

The fluorescence spectrum of a water sample consisting of specific components of a water sample to be detected is not the simple superposition of the fluorescence spectrum of each component, so that the concentration values of each component in the water sample can be obtained through a model which is obtained through mathematical separation and is used for mathematically separating each component in the water sample and establishing the model between each component and the original component.

Decomposing the correction data matrix of the water sample by a parallel factor (PARAFAC) method, and obtaining the composition components of the water sample and the composition component data matrix of the water sample by a nuclear Consistency diagnosis (Core Consistency Diagnostic) analysis method or a Tucker coefficient verification (Tucker Consistency Coefficients).

The parallel factorization method (PARAFAC) is one of the tools for mathematically separating the components of the fluorescence spectrum of a multi-component mixed solution. In addition, mathematical separation of each component of the fluorescence spectrum of the multi-component mixed solution can be realized by methods such as non-negative matrix decomposition and the like.

And calculating the maximum fluorescence peak value (F _ max) of the water sample correction data matrix and each component data matrix and the corresponding excitation wavelength (Ex _ max) and emission wavelength (Em _ max).

X is a determination matrix formed by the fluorescence intensities of the m water samples at n wavelengths, and Y is a concentration matrix (the concentration can be measured according to spectral data) of k components (namely, the specific components of each water sample to be determined) of the water samples in the m water samples. Decomposing the X matrix according to principal component analysis

T _(m×h) ＝X _(m×n) P _(n×h) +E _(m×h)

Wherein P is the eigenvector (the first h eigenvectors selected after sorting according to the eigenvalue) of the covariance matrix of X, T is the projection of X on P, namely the principal component score matrix, and h is the number of principal components. And E is a residual matrix.

X can be linearly expressed as T, so the measurement matrix X and the concentration matrix Y can be linearly regressed:

Y _(m×k) ＝T _(m×h) B _(h×k)

for a data set acquired by using a spectroscopy method, since the number of information variables (sample dimension) for acquiring a certain wavelength spectrum is often greater than the number of samples, and sometimes the number of sample data is even less than the sample dimension, a data matrix cannot be inverted, and a good enough linear model cannot be obtained.

The ratio of the concentration score of a certain component to the total score obtained by the parallel factorization method is considered to be the concentration ratio of the component in the mixed solution, but it has been found experimentally that the concentration ratio determined by this method has a certain error from the mixing ratio of the components in the actual solution due to the interference of the fluorescent substance. Therefore, the difference between the calculation result and the true value is larger according to the concentration of each component in the inversion solution in the concentration score ratio.

In order to solve the problems that the error between the concentration factor score and the true value calculated by directly using a parallel factor method is large and the dimension of the data set sample obtained by the spectrum is insufficient, an improved regression equation model can be established by further using a partial least square regression method, a ridge regression method, an elastic network regression method, an LASSO regression method and the like, and the problems can be better corrected and solved.

Preferably, the regression model is established using partial least squares.

The specific calculation method of the partial least square method is that the water sample correction data matrix is subjected to data standardization, then the mean value, standard deviation and correlation coefficient matrix of each series of correction data matrix are respectively calculated, and a standardized determination matrix X and a standardized concentration matrix Y are constructed;

extracting components t from the standardized measurement matrix X and concentration matrix Y by partial least squares regression ₁ And u ₁ I.e. the 1 st pair of components, to establish a regression equationA process;

extracting a component t by using a residual error matrix of the regression equation ₂ And u ₂ Namely the 2 nd component pair, establishing a regression equation fitting the 2 components; and so on.

Every time a component pair is extracted once, the number standard of the component pairs is provided by using cross validity setting, for example, for a regression equation fitting h components, the sum of the squares of prediction errors of all dependent variables Y is defined as PRESSh,

is provided with

Defining the sum of the squared errors of all dependent variables Y as SS _h ，

Is provided with

For all dependent variables Y, the component t _h The cross-validation is defined as:

that is, if

Increasing the component logarithm to improve the prediction model, and repeating the above steps until the next component pair is extracted from the normalized X and Y by partial least squares regression

Finishing the iteration;

and (4) carrying out principal component analysis on the independent variables, unifying the sum of the characteristic values into 100, and calculating the number h of the principal components after the characteristic values reach the contribution rate.

Based on the linear regression of the measurement matrix X and the concentration matrix Y, a coefficient matrix B can be obtained:

B＝(T′T)-1T′Y

according to the coefficient matrix B obtained by calculation, the constant term alpha is divided into ₀ Coefficient of original variation alpha ₁ 、α ₂ 、α ₃ 、...α _p Substituting the following formula to construct a regression equation of the concentration values and fluorescence values of the components:

y＝α ₀ +α ₁ x ₁ +α ₂ x ₂ +…+α _p x _p

y represents the concentration value of a specific component of a solution to be measured, and x ₁ 、x ₂ ...x _p The fluorescence value calculated from the spectral data of the specific component of the sample solution is expressed, and preferably, the fluorescence value calculated from the spectral data of the specific component of the sample solution is normalized.

According to the established regression equation, the concentration of the specific component of the certain solution to be detected can be calculated by the fluorescence spectrum of the specific component of the certain solution to be detected in the mixed standard solution.

A third preparation step: obtaining a linear fitting equation of the concentration of the specific component of the solution to be measured and the COD contribution value

Preparing a series of standard solutions with certain concentration gradients for specific components of different solutions to be detected respectively.

Measuring the COD value of each standard solution with different concentration in each series by adopting a standard method established by the state; the national standard method for measuring COD comprises the method for measuring COD by a dichromate method _Cr And COD determined by potassium permanganate method _Mn 。

According to the concentration of the standard solution of the specific component of each solution to be detected and the corresponding COD value, constructing a linear fitting equation of the COD and the concentration of the specific component of each solution to be detected:

y′＝ax′+b

wherein y 'represents the contribution value of a specific component of a solution to be detected to the COD of the solution, and x' represents the concentration value of the specific component of the solution to be detected in the solution.

A detection step: COD determination of water sample to be tested

Filtering the collected water sample to be detected and ultrapure water prepared in a laboratory through a 0.45-micron filter membrane, setting instrument measurement parameters, respectively measuring the excitation spectrum and the emission spectrum of the water sample to be detected and the ultrapure water by using a fluorescence spectrometer, and forming an excitation _ emission data matrix EEMs (electronic emission measurement systems) by using the obtained excitation spectrum and emission spectrum data.

And subtracting the excitation-emission data matrix of ultrapure water from the excitation-emission data matrix of the water sample to be detected, and subtracting the interference of Raman scattering.

And (4) according to a Rayleigh scattering formula, deducting first-level and second-level Rayleigh scattering or deducting Rayleigh scattering and Raman scattering interference by using matlab2018a and a tool box DOMFluor to obtain a correction data matrix of the water sample to be detected.

And (3) analyzing and calculating the correction data matrix by using a parallel factor (PARAFAC), and verifying the correction data matrix by using a Core Consistency Diagnostic analysis (Core Consistency Diagnostic) analysis method or a Tucker coefficient (Tucker Consistency Coefficients) to obtain the data matrix of each component of the water sample solution to be detected.

And (3) calculating the maximum fluorescence peak value F _ max, the corresponding excitation wavelength Ex _ max and the emission wavelength Em _ max of the data matrix of the correction data matrix and each component, and carrying out qualitative and quantitative analysis on the water sample to be detected and each component according to the maximum fluorescence peak value F _ max, the corresponding excitation wavelength Ex _ max and the emission wavelength Em _ max so as to find out whether the components of the water sample to be detected are changed compared with the specific water sample to be detected when the specific components of the water sample to be detected are determined.

If the water sample is changed, the water sample is used as an abnormal water sample, the preparation step needs to be executed again on the abnormal water sample to be detected, the selection of the specific components of the solution to be detected is adjusted, the concentration and fluorescence spectrum regression model is reconstructed, and the linear fitting equation of the concentration and the COD contribution value is reconstructed. Optionally, the change of the composition of the current water sample to be detected can be prompted or early warned to inform the change of the water quality.

If the composition of the water sample to be detected is not changed or the change is within the threshold value, substituting the fluorescence value obtained by calculating according to the spectral data of the data matrix of the specific component of each solution to be detected into a regression equation of a concentration and fluorescence spectrum regression model, and calculating to obtain the concentration value of the specific component of each solution to be detected. Calculating the COD contribution value of the specific component of each solution to be detected to the solution based on the linear fitting equation of the concentration of the specific component of each solution to be detected and the COD, and summing the COD contribution values of the specific components of each solution to be detected to obtain the COD value of the water sample to be detected.

In accordance with another aspect of the present invention, as shown in FIG. 2, there is provided an apparatus for performing the method for measuring chemical oxygen demand in an aqueous environment using fluorescence spectroscopy as described above.

The second embodiment:

referring to fig. 2, an apparatus 1 for determining chemical oxygen demand in water by fluorescence spectroscopy according to the present invention includes a composition analysis module 11, a spectral data separation module 12, a regression model construction module 13, a concentration analysis module 14, and a chemical oxygen demand analysis module 15.

The apparatus 1 for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy may be further provided with a fluorescence spectroscopy apparatus 2, and the fluorescence spectroscopy apparatus 2 may be a conventional fluorescence spectrometer. The fluorescence spectrum measuring device 2 can measure the spectrum data of the water sample to be measured.

The apparatus 1 for determining chemical oxygen demand in water environment by fluorescence spectroscopy may further comprise a correction module 16 for correcting fluorescence spectrum data of a detected water sample by removing rayleigh scattering or/and raman scattering. Preferably, after the water sample to be measured is filtered by a filter membrane with the aperture of 0.45 mu m, the fluorescence spectrum measuring device 2 measures the excitation spectrum and the emission spectrum of the water sample to be measured and the ultrapure water. The calibration module 16 forms an excitation-emission data matrix from the obtained excitation spectrum and emission spectrum data, subtracts the excitation-emission data matrix of ultrapure water from the excitation-emission data matrix of the water sample, and subtracts the interference of raman scattering to obtain calibration spectrum data. The correction module 16 can subtract the first and second rayleigh scattering according to the rayleigh scattering formula to obtain the corrected spectrum data. The correction module 16 may also obtain corrected spectrum data with rayleigh scattering and raman scattering interference subtracted by matlab2018a and DOMFluor.

And the spectrum data separation module 12 is used for acquiring fluorescence spectrum data of the water sample to be detected and separating and acquiring the fluorescence spectrum data of the specific component of the water sample to be detected by using a parallel factor method or a non-negative matrix analysis method. The spectrum data separation module 12 can obtain fluorescence spectrum data of specific components of the water sample to be detected for chemical oxygen demand analysis of the water sample to be detected. The spectrum data separation module 12 can also analyze and determine specific components of the water sample to be detected for the composition analysis module 11, and the regression model construction module 13 constructs a regression model to separate and obtain fluorescence spectrum data of the water sample to be detected or the water sample.

And the composition analysis module 11 is used for analyzing or selecting and determining specific components of the solution to be measured, which are required to be used as COD (chemical oxygen demand) determination standard substances.

The composition analysis module 11 analyzes and determines the specific components of the water sample to be tested by the following methods:

a water sample to be detected with a certain volume is taken, filtered by a 0.45 micron filter membrane and scanned by a fluorescence spectrum measuring device 2, and three-dimensional fluorescence spectrum data of the water sample to be detected is obtained. Preferably, the correction module 16 subtracts the first and second rayleigh scatterings according to the rayleigh scattering formula to obtain the corrected spectrum data. The composition analysis module 11 performs parallel factor (PARAFAC) analysis and calculation on the spectrum data matrix, performs verification (tuner Consistency diagnostics) through a Core Consistency Diagnostic analysis method or a tuner coefficient to obtain the composition number of the solution of the water sample to be tested, the data matrix and the fluorescence spectrum of each component, and determines the composition of the water sample to be tested according to the fluorescence spectrum characteristics of each component. And (3) preparing a simulation solution for simulating the water sample to be detected based on the determined composition simulation, and determining a component with the total chemical oxygen demand contribution rate of not less than 95% to the water sample to be detected as the specific component of the water sample to be detected in a manner of determining the solution sample recovery rate.

When the specific component required to be measured as the standard substance is determined based on the selection, the main component analysis module selects the specific component to be measured as the standard substance based on the characteristics of the water sample to be measured. Preferably, at least one of Fluorescent Dissolved Organic Matter (FDOM) is selected as the specific component of the water sample to be tested. Preferably, when the water sample to be detected is surface water, at least one of humic substances, fulvic acid substances, proteinoid substances or amino acid substances is selected as the specific component of the water sample to be detected.

Preferably, the composition analysis module 11 can adopt the two technical schemes to determine the specific components of the water sample to be detected, and preferentially combines the properties of the water sample to be detected and selects the specific components when the components with the chemical oxygen demand contribution rate not less than 95% of the water sample to be detected are selected as the specific components of the water sample to be detected. For example, FDOM is preferably selected as the specific component, or when the sample to be tested is surface water, humic substance, fulvic acid substance, proteinoid substance or amino acid substance is preferably selected as the specific component of the sample to be tested.

The composition analysis module 11 may further include a composition adjustment module 111, which obtains the spectral data of the water sample to be tested for quantitative analysis, so that the composition analysis module 11 can determine whether to adjust the specific components of the water sample to be tested.

And the regression model building module 13 is used for building a regression model of the concentration and the fluorescence spectrum of the specific component of the water sample to be detected.

The regression model building module 13 builds a concentration and fluorescence spectrum regression model for the specific components of the water sample to be tested of the current water sample to be tested by the following method:

the specific component sample solution of the water sample to be detected is prepared by all selected specific component substances of the water sample to be detected, serial water sample samples with different concentrations are obtained after dilution and mixing according to a proportion, the serial water sample samples are scanned by the fluorescence spectrum measuring device 2, and three-dimensional fluorescence spectrum data of the water sample are obtained, preferably, the fluorescence spectrum data can be corrected by the correction module 16. The spectrum data separation module 12 separates the fluorescence spectrum data to obtain a water sample and a spectrum data matrix of each component. The regression model construction module 13 constructs a determination matrix according to fluorescence intensity data of water samples with different concentrations at different wavelengths; constructing a concentration matrix according to the concentrations of the components in the water sample with different concentrations; performing linear regression on the determination matrix and the concentration matrix by using principal component analysis; calculating a coefficient matrix by using a principal component analysis method based on linear regression of the determination matrix and the concentration matrix; preferably, the measurement matrix and the concentration matrix are respectively standardized, component pairs are extracted one by one from the standardized measurement matrix and concentration matrix by using a partial least square regression method, a ridge regression method, an elastic network regression method, an LASSO regression method and other methods, a regression equation fitting all extracted components is constructed, scores are determined by using cross validity verification, and a coefficient matrix is obtained by calculation based on linear regression of the measurement matrix and the concentration matrix; and extracting coefficients corresponding to components required by the regression model one by one from the coefficient matrix to construct a concentration and fluorescence spectrum regression model. The specific construction flow is as in "preparation step two" of the first embodiment: specific components of the water sample to be tested and the construction of the regression model of the concentration and fluorescence spectrum of each specific component are not described herein.

Preferably, the apparatus 1 for determining chemical oxygen demand in water environment by using fluorescence spectroscopy may further include a component concentration and fluorescence spectrum regression model information database 17 for storing the specific component concentration and fluorescence spectrum regression model of each water sample to be determined corresponding to the specific water sample to be determined, and the concentration analysis module 14 may be capable of retrieving from the component concentration and fluorescence spectrum regression model information database 17 when determining the specific water sample to be determined. And the concentration analysis module 14 is used for acquiring the fluorescence spectrum data of the specific component of the water sample to be detected from the spectrum data separation module 12, and analyzing the concentration of the specific component in the water sample to be detected in the regression model establishment module 13 or the concentration and fluorescence spectrum regression model information database 17 corresponding to the concentration of the specific component of the water sample to be detected, and the fluorescence spectrum regression model.

And the chemical oxygen demand analysis module 15 receives the concentrations of the specific components analyzed by the concentration analysis module 14, analyzes and obtains the chemical oxygen demand of each specific component according to the relationship between the concentration of each specific component and the chemical oxygen demand, and sums the chemical oxygen demands of each specific component to obtain the chemical oxygen demand of the water sample to be detected. The relation between the concentration of each specific component and the chemical oxygen demand can be obtained by measuring or manually measuring the relation between the concentration and the chemical oxygen demand by the concentration and chemical oxygen demand measuring device 19 and then provided to the chemical oxygen demand analyzing module 15, or can be stored in the component concentration and chemical oxygen demand information database 18 in advance and called by the chemical oxygen demand analyzing module 15.

The concentration and chemical oxygen demand relation measuring device 19 can measure the COD value of each standard solution with different concentrations in each series by using a national standard method for the prepared series of standard solutions of the specific components of the water sample to be measured with a certain concentration gradient, and construct a linear fitting equation of the COD and the concentration of the specific components of each solution to be measured according to the concentration of the standard solution of the specific components of each solution to be measured and the corresponding COD value. The concentration and chemical oxygen demand relation determination device can store the constructed linear fitting equation of the COD and the concentration of the specific component of the solution to be measured in the component concentration and chemical oxygen demand information database 18.

When the chemical oxygen demand of a specific water sample to be measured is measured, the collected water sample to be measured and ultrapure water prepared in a laboratory are filtered by a 0.45-micron filter membrane, and then the fluorescence spectrum measuring device 2 is used for respectively measuring the spectral data of the water sample to be measured and the ultrapure water, preferably, the spectral data can comprise an excitation spectrum and an emission spectrum, and the obtained excitation spectrum and the obtained emission spectrum data form an excitation-emission data matrix EEMs. Preferably, the calibration module 16 forms an excitation-emission data matrix from the obtained excitation spectrum and emission spectrum data, subtracts the excitation-emission data matrix of ultrapure water from the excitation-emission data matrix of the water sample, and subtracts the interference of raman scattering to obtain the calibration spectrum data. The correction module 16 subtracts the first and second rayleigh scatterings according to the rayleigh scattering formula to obtain the corrected spectrum data. The correction module 16 may also obtain corrected spectrum data with rayleigh scattering and raman scattering interference subtracted by matlab2018a and DOMFluor.

The spectral data separation module 12 performs parallel factor (parafacc) analysis and calculation on the spectral data matrix, and obtains a data matrix of each component of the water sample solution to be tested by a nuclear Consistency diagnosis (Core Consistency diagnosis) analysis method or a tracker coefficient verification (tracker Consistency Coefficients).

The component adjusting module 111 of the component analyzing module 11 acquires the maximum fluorescence peak value of the spectral data of the water sample to be detected and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value; acquiring the maximum fluorescence peak value of each component spectral data and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value; and quantitatively analyzing the acquired spectral data of the water sample to be detected, the maximum fluorescence peak value of the spectral data of each component, and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value, and determining whether the components of the currently determined water sample to be detected are changed compared with the specific water sample to be detected when the specific components of the water sample to be detected are determined. If the water sample is changed, the water sample is used as an abnormal water sample, the component analysis module 11 is required to adjust the selection of the specific components of the solution to be detected of the abnormal water sample to be detected again, the regression model construction module 13 is used for reconstructing the concentration and fluorescence spectrum regression model, and the linear fitting equation of the constructed concentration and the COD contribution value is determined again. Optionally, the change of the composition of the current water sample to be detected can be prompted or early warned to inform the change of the water quality. If the composition of the water sample to be tested does not change or the change is within the threshold, the concentration analysis module 14 obtains the fluorescence spectrum data of the specific component of the water sample to be tested from the spectrum data separation module 12, and the regression model construction module 13 or the concentration of the specific component of the water sample to be tested obtained from the component concentration and fluorescence spectrum regression model information database 17 and the concentration of the specific component of the water sample to be tested in the water sample to be tested are analyzed by the fluorescence spectrum regression model. And the chemical oxygen demand analysis module 15 receives the concentrations of the specific components of the water samples to be detected, which are analyzed by the concentration analysis module 14, analyzes the concentrations to obtain the chemical oxygen demand of the specific components of the water samples to be detected according to the relationship between the concentrations of the specific components of the water samples to be detected and the chemical oxygen demand, and sums the chemical oxygen demands of the specific components of the water samples to be detected to obtain the chemical oxygen demand of the water samples to be detected.

Example three:

the invention also provides a device for measuring the chemical oxygen demand in water by using fluorescence spectroscopy, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to realize part or all of the steps of any method for measuring the chemical oxygen demand in water by using fluorescence spectroscopy.

Example four:

the present invention also provides a computer readable storage medium storing a computer program which, when executed by a processor, performs part or all of the steps of any one of the methods of measuring chemical oxygen demand in water using fluorescence spectroscopy described herein.

An embodiment of the present invention is described below by taking a water sample containing amino acids and fulvic acids as an example, so that the technical contents thereof will be more clear and easily understood.

Example five:

the method is characterized in that a water sample containing amino acid-like substances and fulvic acid-like substances is determined by component analysis, and specific components with the total COD contribution rate of more than 95% in the water sample are tryptophan-like substances, fulvic acid-like substances and tyrosine-like substances.

Tryptophan and fulvic acid standard stock solutions with the concentration of 1000mg/L and tyrosine standard stock solutions with the concentration of 300mg/L are respectively prepared. Respectively absorbing a plurality of volumes from standard stock solutions of tryptophan, fulvic acid and tyrosine, and mixing and diluting the three solutions into a mixed solution according to the analysis results of the components of the tryptophan, the fulvic acid and the tyrosine in a water sample containing amino acid substances and fulvic acid substances, so that the concentration of the tryptophan in the mixed solution is 1mg/L-100mg/L, the concentration of the fulvic acid is 5mg/L-100mg/L, and the concentration of the tyrosine is 2mg/L-80mg/L; respectively calculating the concentration of each standard substance in the mixed solution, preparing a certain volume by using standard stock solution, and measuring the COD of the solution by using a standard method;

after the mixed solution is filtered by a 0.45 micron filter membrane, measuring an excitation spectrum and an emission spectrum of the mixed solution and ultrapure water by using a fluorescence spectrometer, and forming excitation-emission data matrixes EEMs by using the obtained excitation spectrum and emission spectrum data, wherein FIG. 3 is a fluorescence spectrum diagram of the mixed solution;

subtracting the excitation-emission data matrix of the ultrapure water from the excitation-emission data matrix of the mixed solution, and subtracting the interference of Raman scattering;

according to a Rayleigh scattering formula, primary and secondary Rayleigh scattering is deducted, or a correction data matrix with Rayleigh scattering and Raman scattering interference deducted is obtained by utilizing matlab2018a and DOMFluor;

performing parallel factor (PARAFAC) analysis and calculation on the correction data matrix, and verifying by a Core Consistency Diagnostic analysis method or a Tucker coefficient (Tucker Consistency Coefficients) to obtain the component number of the mixed solution, the percentage of each component in the total component and the data matrix of each component, wherein FIG. 4 is a 3 component spectrogram of a fluorescence spectrum of a certain mixed solution after parallel factor decomposition;

establishing regression equations for the data matrix of the 3 components and the concentrations of the tyrosine solution, the fulvic acid solution and the tryptophan solution when the mixed solution is prepared, and establishing fitting equations for the concentrations of the components in the mixed solution and the COD of the solution respectively;

after parallel factor analysis and calculation

The regression equation for component one tyrosine fluorescence spectrum and concentration is:

y _tyrosine ＝17.061+(0.072*x1)+(0.038*x2)+(-0.064*x3)+(-0.037*x4)

The regression equation of the fluorescence spectrum and the concentration of the fulvic acid component II is as follows:

y _{fulic acid} ＝31.735+(0.533*x1)+(-0.367*x2)+(-0.145*x3)+(0.068*x4)

The regression equation of the fluorescence spectrum and the concentration of the components of the tryptophan is as follows:

y _tryptophan ＝24.059+(0.164*x1)+(-0.171*x2)+(0.109*x3)+(-0.088*x4)

The fitting equation of the component-tyrosine solution concentration and COD is as follows:

y1′ _COD ＝1.83y _tyrosine +0.19

The fitting equation of the concentration of the component bifurcate solution and COD is as follows:

y2′ _COD ＝0.81y _{fulic acid} +1.32

The fitting equation of the concentration and COD of the trichromatic ammonia acid solution is as follows:

y3′ _COD ＝1.71y _tryptophan +2.33

Collecting 10 water samples containing amino acid-like substances and fulvic acid-like substances, dividing each water sample into two parts, measuring the chemical oxygen demand of one water sample according to a national standard method, filtering the other water sample by a filter membrane with the aperture of 0.45 mu m, measuring the excitation spectrum and the emission spectrum of the water sample and ultrapure water by using a fluorescence spectrometer, and forming an excitation-emission data matrix EEMs by using the obtained excitation spectrum and emission spectrum data;

subtracting the excitation-emission data matrix of ultrapure water from the excitation-emission data matrix of the water sample, and subtracting the interference of Raman scattering;

performing parallel factor (PARAFAC) analysis and calculation on the correction data matrix, and verifying by a Core Consistency Diagnostic analysis method or a Tucker coefficient (Tucker Consistency coeffients) to obtain the percentage of the total components of the water sample solution, namely the tryptophan, the fulvic acid and the tyrosine, and the data matrix of each component; (percentages refer to the fractions of the components, dimensionless, concentrations refer to the specific values, typically milligrams per liter, dimensionless)

Calculating the maximum fluorescence peak value F _ max of the correction data matrix and the corresponding excitation wavelength Ex _ max and emission wavelength Em _ max, and performing qualitative and quantitative analysis on the water sample to be detected and each component according to the maximum fluorescence peak value F _ max and the corresponding excitation wavelength Ex _ max and emission wavelength Em _ max so as to find an abnormal water sample and adjust and select the components of the standard solution;

and substituting the data matrix of the 3 components into the established regression equation, respectively calculating the concentration values of the 3 components, respectively substituting the concentration of each component into the established solution concentration and COD linear equation of each component, calculating the COD value of each component, and taking the sum of the COD values of each component as the COD value of the water sample solution of the reservoir.

TABLE 1 determination and calculation of COD of water sample containing amino acids and fulvic acids

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the present application are generated in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., coaxial cable, fiber optic, digital subscriber line) or wirelessly (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk), among others.

Although the present invention has been described with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the present invention.

Claims

1. A method for determining chemical oxygen demand in water by fluorescence spectrometry, comprising:

acquiring fluorescence spectrum data of a water sample to be detected;

separating the fluorescence spectrum data of the water sample to be detected by using a parallel factor method or a nonnegative matrix analysis method to obtain the fluorescence spectrum data of the specific component of the water sample to be detected;

the specific component of the water sample to be detected is at least one component with the highest chemical oxygen demand contribution rate in the water sample to be detected or at least one component selected based on the characteristics of the water sample to be detected;

preparing m mixed standard solutions with different concentrations by using k specific component substances of all selected water samples to be detected; constructing a determination matrix according to the fluorescence intensities of m mixed standard solutions with different concentrations at n different wavelengths; constructing a concentration matrix according to the concentrations of the k selected main components in the mixed standard solution in the m mixed standard solutions; performing linear regression on the measurement matrix and the concentration matrix; extracting component pairs one by one from the standardized measuring matrix and the standardized concentration matrix by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method, and performing principal component analysis to calculate the number of principal components; based on principal component analysis and linear regression of the determination matrix and the concentration matrix, a regression model is constructed by obtaining a coefficient matrix by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method.

2. The method for determining COD by fluorescence spectrometry of claim 1, wherein the specific component of the water sample to be tested is a composition with COD contribution rate not less than 95%.

3. The method for determining COD by fluorescence spectrometry of claim 1, wherein the sample of water to be tested is surface water, and at least one of humic substances, fulvic acids substances, proteinoid substances or amino acid substances is selected as the specific component of the sample of water to be tested.

4. The method of claim 1, wherein at least one of Fluorescent Dissolved Organic substances (FDOM) is selected as a specific component of the sample.

5. The method for determining COD in water by fluorescence spectrometry according to claims 1 to 4, wherein the relation between the concentration of the specific component of the sample to be tested and the COD is obtained according to a pre-determined linear fitting equation of the concentration of the specific component of the sample to be tested and the COD or pre-obtained relation data of the concentration of the specific component and the COD.

6. The method for determining COD by fluorescence spectrometry according to any one of claims 1 to 4, wherein the spectral data of the water sample to be tested is corrected by removing Rayleigh scattering or/and Raman scattering after the spectral data of the water sample to be tested is obtained.

7. The method according to any one of claims 1 to 4, wherein the quantitative analysis is performed on the water sample to be tested, and if the composition of the water sample to be tested is changed compared with the specific water sample to be tested when the composition of the specific water sample to be tested is determined before, the specific composition of the water sample to be tested is adjusted or the change of the composition of the water sample to be tested is prompted.

8. The method for determining the chemical oxygen demand in water by utilizing the fluorescence spectrometry according to any one of claims 7, wherein the maximum fluorescence peak value of the spectral data of a water sample to be measured and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value are obtained; acquiring a maximum fluorescence peak value of each composition spectral data of a water sample to be detected and an excitation wavelength and an emission wavelength corresponding to the maximum fluorescence peak value; and quantitatively analyzing the acquired spectral data of the water sample to be detected, the maximum fluorescence peak value of the spectral data of each component, and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value.

9. An apparatus for determining chemical oxygen demand in water by fluorescence spectroscopy, comprising:

the regression model building module is used for configuring a specific component sample solution of the water sample to be detected by using specific component substances of the water sample to be detected based on specific components of the water sample to be detected selected by the composition analysis module, building a regression model of concentration and fluorescence spectrum based on a principal component analysis method by using component spectrum data and component concentrations of the sample solutions with different concentrations, and improving the regression model of concentration and fluorescence spectrum by using a partial least square method, a ridge regression analysis method, a LASSO analysis method or an elastic network analysis method;

the concentration analysis module acquires fluorescence spectrum data of the specific component of the water sample to be detected from the spectrum data separation module, and analyzes the concentration of the specific component in the water sample to be detected based on the concentration of the specific component of the water sample to be detected acquired by the autoregressive model construction module and the fluorescence spectrum regression model;

10. The apparatus for determining chemical oxygen demand in water environment by fluorescence spectrometry according to claim 9, wherein the composition analysis module selects the composition having the chemical oxygen demand contribution rate of not less than 95% in the water sample to be measured as the specific component of the water sample to be measured.

11. The apparatus for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy of claim 9, wherein the composition analysis module selects at least one of Fluorescent Dissolved Organic Matter (FDOM) as the specific component of the sample water to be measured.

12. The apparatus for determining chemical oxygen demand in an aqueous environment by fluorescence spectrometry according to claim 9, wherein the sample of water to be measured is surface water, and the composition analysis module selects at least one of humic substances, fulvic acids substances, proteinoid substances, or amino acid-like substances as the specific component of the sample of water to be measured.

13. The apparatus for determining chemical oxygen demand in water environment by fluorescence spectrometry according to any one of claims 9 to 12, wherein the composition analysis module comprises a composition adjustment module for obtaining the maximum fluorescence peak value of the spectral data of the water sample to be determined and the excitation wavelength and emission wavelength corresponding to the maximum fluorescence peak value; acquiring the maximum fluorescence peak value of each component of spectral data and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value; and quantitatively analyzing the acquired spectral data of the water sample to be detected, the maximum fluorescence peak value of the spectral data of each component, and the excitation wavelength and the emission wavelength corresponding to the maximum fluorescence peak value, wherein if the composition of the current water sample to be detected changes compared with the specific water sample to be detected when the specific composition of the water sample to be detected is determined before, the composition division analysis module adjusts the specific composition of the water sample to be detected or prompts the condition that the composition of the water sample to be detected changes.

14. An apparatus for determining chemical oxygen demand in an aqueous environment by fluorescence spectrometry according to any one of claims 9 to 12, comprising a concentration-chemical oxygen demand relationship determining means for determining the relationship between the concentration of a specific component of the sample water to be measured and the chemical oxygen demand for the chemical oxygen demand analysis module to call.

15. An apparatus for determining chemical oxygen demand in an aqueous environment using fluorescence spectrometry as claimed in any one of claims 9 to 12, wherein a fluorescence spectrometry apparatus is provided for acquiring fluorescence spectrum data of the water sample.

16. An apparatus for determining chemical oxygen demand in an aqueous environment using fluorescence spectroscopy as claimed in any one of claims 9 to 12 comprising a correction module for correcting the fluorescence spectroscopy data by removing rayleigh scattering or/and raman scattering.

17. An apparatus for determining chemical oxygen demand in water using fluorescence spectroscopy, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor when executing the computer program implements the steps of the method of any one of claims 1 to 8.

18. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 8.