Pollution tracing method based on three-dimensional fluorescence spectrum

Abstract

The invention discloses a pollution source tracing method based on a three-dimensional fluorescence spectrum, and relates to the technical field of water quality monitoring, a water sample of an upstream specific pollution discharge unit is collected as a pollution source, a mixed water sample formed by mixing different pollution source water samples according to different volumes or proportions is prepared, and a pollution source database adapted to parallel factor analysis is established; according to the pollution source tracing method, a good linear model between a pollution source and a three-dimensional fluorescence spectrum characteristic value is established, a shortest path algorithm is proposed to quantitatively solve the volume or proportion of the pollution source in a monitored water sample, and the pollution source is traced to a specific pollution discharge unit. According to the invention, the sewage three-dimensional fluorescence spectrum database established by mixing upstream pollution source drainage improves the application range of parallel factor analysis, so that the three-dimensional fluorescence spectrums of sewage of different industrial systems can be subjected to effective parallel factor analysis together; compared with a parallel factor analysis-Tucker congruence coefficient method, the method has the advantage that the pollution source can be traced to a more specific level.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality monitoring, and in particular to a pollution source tracing method based on three-dimensional fluorescence spectroscopy. Background Art

[0002] As an important part of environmental issues, water environment is closely related to other environmental issues such as soil environment, so the importance of water pollution control is self-evident. Centralized treatment of sewage will become the primary task of water pollution control, and whether sewage discharge units centrally treat sewage according to standards and discharge according to standard quotas directly affects the completion of this task. Therefore, water quality monitoring and pollution tracing methods are one of the current research focuses, which will provide important means and important references for water pollution control.

[0003] To cope with different sewage discharge conditions, the main methods currently include hydrodynamic modeling, behavioral planning algorithms, and concentration gradient algorithms. However, due to the complex composition of actual sewage, there are many reasons for the change of a certain water quality monitoring indicator. The above methods require extensive collection of various traditional water quality monitoring indicators. The establishment and adjustment of models or algorithms also rely on experience. A single model cannot be well extended to other watersheds that need to be monitored. Compared with traditional water quality monitoring indicators, three-dimensional fluorescence spectroscopy has unique advantages in the analysis of actual sewage with complex components due to its high sensitivity, multi-dimensional information, and comprehensive chemical composition reflection. In recent years, it has attracted more and more attention from scholars. After being irradiated by short-wavelength excitation light, the molecules of a substance absorb photon energy, and the electrons in the molecules transition from the ground state to the excited state. The molecules in the excited state are unstable and return to the ground state by radiation transition, and emit long-wavelength fluorescence that can reflect the characteristics of the substance. The light absorbed by the molecules of the substance to be tested when they become excited is the excitation light, and the light generated when the molecules in the excited state return to the ground state is the emission light. The excitation spectra and emission spectra of different material molecules are different. Therefore, the spectrum obtained by recording the excitation and emission wavelength and intensity information by the instrument is called a three-dimensional fluorescence spectrum (EEM). It can reflect the type and content of organic matter in the sample and has the characteristic of one-to-one correspondence with the sample.

[0004] At present, the main analytical methods for water quality monitoring or pollution tracing based on three-dimensional fluorescence spectroscopy are:

[0005] 1. Single wavelength or multi-wavelength spectroscopy: A method of selecting an excitation or emission spectrum at one or more wavelengths to establish a linear relationship between the fluorescence intensity in the spectrum and the information of the substance to be detected.

[0006] 2. Fluorescence area integration method: Based on experience, the three-dimensional fluorescence spectrum is divided into different regions, the fluorescence intensity in different regions is integrated, and the integral is used to represent the cumulative fluorescence response of water-soluble organic matter (DOM) with similar characteristics.

[0007] 3. Feature extraction and matching method of two-dimensional single-shot surface data: After normalizing the three-dimensional fluorescence spectrum, the area is divided according to the intuitive contour line convexity in the form of a contour map, the maximum difference in fluorescence intensity and the center position in the area are extracted as feature values, and a Reeb diagram is drawn for matching.

[0008] 4. Parallel Factor Analysis - Tucker Congruence Coefficient Method: After parallel factor analysis of the three-dimensional fluorescence spectrum, several component graphs split according to the original three-dimensional fluorescence spectrum are obtained. The Tucker Congruence Coefficient Method is used to calculate the similarity of different component graphs with the three-dimensional fluorescence spectra of one or several types of pollutants that may have similar characteristics.

[0009] Parallel factor analysis-Tucker congruence coefficient method is a pollution tracing method for tracing the industry of the pollution source. According to the relationship between the flow direction of the monitored river and the main stream and direct current, this method sets up dozens of water sample collection points in the monitoring area, and performs three-dimensional fluorescence scanning and other detection indicators on the water samples. The three-dimensional fluorescence spectra of the water samples at all collection points are sorted into a data set and then subjected to parallel factor analysis to obtain several component graphs split according to the three-dimensional fluorescence spectra. After judging the type of dissolved organic matter that the high fluorescence intensity area of ​​the three-dimensional fluorescence spectra of different components may represent based on experience, the Tucker congruence coefficient method is used to calculate the similarity between the three-dimensional fluorescence spectrum of the component graph and the three-dimensional fluorescence spectrum of the industrial wastewater that discharges this type of dissolved organic matter. If the similarity reaches 0.85 or above, it is considered that the water body is polluted by the wastewater of this industry.

[0010] Because the parallel factor analysis-Tucker congruence coefficient method stops tracing pollution sources at the industry level, technicians in this field are committed to developing a method to monitor pollution excess emissions in downstream areas with multiple sources and complex pollution and trace them back to specific emission units. Summary of the invention

[0011] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is to provide a method for monitoring pollution excess discharge in downstream areas with multiple sources and complex pollution and tracing it back to specific emission units.

[0012] To achieve the above object, the present invention provides a pollution source tracing method based on three-dimensional fluorescence spectroscopy, which comprises the following steps:

[0013] Step 1: Collect the independent drainage of n pollution sources in the upstream of the monitoring basin, and collect 1 portion of mixed drainage when the pollution sources merge and discharge in the downstream, which is recorded as "standard". Use the independent drainage collected upstream to configure i portions of mixed sewage in different volume proportions, and together with the "standard", form a total of i+1 portions of mixed sewage samples, record the drainage configuration information of the pollution sources contained in each portion of mixed sewage, and scan the three-dimensional fluorescence spectrum of the mixed sewage to form a one-to-one corresponding database with the configuration information;

[0014] Step 2: Collect actual water samples at the downstream end where all sewage discharge units in the monitoring basin converge, scan their three-dimensional fluorescence spectra, and perform parallel factor analysis together with the three-dimensional fluorescence spectra in the database to split them into m components, and extract the peak value of the fluorescence peak of each component as the characteristic value;

[0015] Step 3: Arrange the configuration information of the mixed sewage into a matrix I of size (i+1)×n, add a column of unit column vectors to the left of the matrix I, and record it as matrix IN; Arrange the eigenvalues ​​of the mixed sewage into a matrix C of size (i+1)×m, and solve the linear model matrix M so that IN·M=C;

[0016] Step 4: Arrange the eigenvalues ​​of the actual water sample into an m-dimensional column vector c, record the quantified information of the drainage composition of the n sewage discharge units contained in the actual water sample as an n-dimensional column vector x, and use the matrix equation M T [1; x] = c finds the relationship between the elements in x;

[0017] Step 5: When n is less than or equal to m, x has a unique solution; when n is greater than m, the specific composition of the elements in the column vector x is [x1, x2, …, x m ,x m+1 ,…,x n-1 ,x n ] T , the specific elements of the discharge information column vector of the conventional discharge water sample "standard" are [s1, s2, ..., s n ] T , construct the Euclidean distance function f(x1,x2,…,x n )=(x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 , using the matrix equation M T ·[1;x]=c Eliminate the function f and get f(x1,x2,…,x m ,x m+1 ,…x n )=g(x1,x2,…,xm ), find the unique solution x that minimizes the function g;

[0018] Step 6: By comparing the configuration information of the mixed water sample with the standard discharge situation, it is possible to determine whether there is over-discharge upstream and to determine the scope of possible sewage discharge units that may cause over-discharge.

[0019] In a preferred embodiment of the present invention, the number i of the configured mixed sewage is not less than 20.

[0020] In a preferred embodiment of the present invention, the configuration information is any one of the volume, volume proportion or concentration ratio of the wastewater of each sewage discharge unit in the mixed sewage.

[0021] In a preferred embodiment of the present invention, the wavelength range of the excitation wavelength and the emission wavelength of the three-dimensional fluorescence spectrum is between 200nm and 900nm, and the scanning wavelength interval is between 1nm and 5nm.

[0022] In a preferred embodiment of the present invention, the components obtained by the parallel factor analysis splitting should be cross-validated.

[0023] In a preferred embodiment of the present invention, the method for solving the linear model matrix is ​​any one of the least squares method, principal component analysis, partial least squares method or support vector machine, and its content format should be organized into the first line as the constant term of each component expression, and the next n lines are the coefficients of the drainage configuration information of n sewage discharge units in each component expression.

[0024] In a preferred embodiment of the present invention, the expression of the function f is (x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 or |x1-s1|+|x2-s2|+…+|x n -s n |One of them.

[0025] In a preferred embodiment of the present invention, the sewage over-discharge situation in step 6 is specifically to determine whether there is over-discharge and to determine the range of possible sewage discharge units that may cause over-discharge.

[0026] In a preferred embodiment of the present invention, the value of each element in the column vector x corresponds to the composition relationship of the wastewater discharged by each upstream sewage discharge unit in the actual water sample collected downstream.

[0027] In a preferred embodiment of the present invention, when judging the actual over-discharge situation and the source of the over-discharged sewage, an increase of 10% to 20% in the configuration information value of the monitored water sample quantification compared to the configuration information value of the standard discharge quantification indicates a low confidence level of over-discharge.

[0028] The present invention uses the method of parallel factor analysis to reveal the fluorescence peak information that is covered by the overlapping fluorescence areas of different substances in the three-dimensional fluorescence spectrum, and has a more accurate fluorescence peak position than the prior art for the feature extraction of the three-dimensional fluorescence spectrum. The present invention improves the application scope of parallel factor analysis by establishing a three-dimensional fluorescence spectrum database of sewage by mixing the drainage of upstream pollution sources, so that the three-dimensional fluorescence spectra of sewage from different industrial systems can be effectively analyzed together in parallel factors, and can trace the source of pollution to a more specific level compared to the parallel factor analysis-Tucker congruence coefficient method. The model establishment method of the pollution source information and the three-dimensional fluorescence spectrum characteristics proposed by the present invention is universal and will not be restricted by different pipe networks or rivers. The pollution source tracing method using three-dimensional fluorescence spectrum proposed by the present invention has the advantages of fewer water quality indicators to be monitored, convenient database establishment, more reliable extraction of three-dimensional fluorescence spectrum feature information, the method of model establishment does not rely on experience and the accuracy of pollution source tracing, and has broad application prospects in the field of water quality monitoring.

[0029] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of a pollution source tracing method of three-dimensional fluorescence spectrum according to a preferred embodiment of the present invention;

[0031] Figure 2 is a kernel consistency function diagram of embodiment 1 of the present invention;

[0032] Figure 3 is a three-component cross-validation diagram of Example 1 of the present invention;

[0033] Figure 4 This is a first component decomposition diagram of Example 1 of the present invention;

[0034] Figure 5 This is a diagram of the second component decomposition of Example 1 of the present invention;

[0035] Figure 6 This is a diagram of the third component decomposition of Example 1 of the present invention;

[0036] Figure 7 is the linear fitting result of the drainage volume ratio of the five factories in Example 1 of the present invention and the peak value of the fluorescence of the first component;

[0037] Figure 8 It is the linear fitting result of the drainage volume ratio of the five factories in Example 1 of the present invention and the peak value of the fluorescence of the second component;

[0038] Fig. 9 It is the linear fitting result of the drainage volume ratio of the five factories in Example 1 of the present invention and the peak value of the fluorescence of the third component;

[0039] Fig.10 This is the increase in the drainage volume of the five factories in Example 1 of the present invention compared with the standard drainage volume. DETAILED DESCRIPTION

[0040] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.

[0041] In the drawings, components with the same structure are indicated by the same numerical reference numerals, and components with similar structures or functions are indicated by similar numerical reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the components is appropriately exaggerated in some places in the drawings.

[0042] The present invention provides a pollution source tracing method based on three-dimensional fluorescence spectroscopy. The method establishes a database that can apply parallel factor analysis by mixing sewage with different components due to different industries and showing obvious differences in three-dimensional fluorescence spectra, so that the pollution source and the water sample of the monitored basin are subjected to parallel factor analysis as a whole, and the peak values ​​of the fluorescence peaks of different components are extracted as characteristic values, and a linear model of the characteristic values ​​and the emission conditions is established, so as to solve the actual tracing problem. The shortest path search algorithm is applied, and when the over-emission situation is monitored, the specific pollution over-emission unit can be traced back.

[0043] The purpose of the present invention can be achieved by the following technical solutions, the complete process is shown in Figure 1 :

[0044] The method for monitoring pollution over-discharge and tracing the source for a downstream section with multiple sources and complex pollution according to the present invention comprises the following steps:

[0045] (1) Collect the independent drainage of n pollution sources in the upstream of the monitoring basin, and collect 1 portion of mixed drainage when the pollution sources merge and discharge in the downstream, which is recorded as the "standard". Use the independent drainage collected upstream to configure i portions of mixed sewage in different volume proportions, and together with the "standard", form a total of i+1 portions of mixed sewage samples. Record the drainage or configuration information of the pollution sources contained in each portion of mixed sewage, and scan the three-dimensional fluorescence spectrum of the mixed sewage to form a one-to-one corresponding database with the configuration information;

[0046] (2) Collect actual water samples at the downstream end of the confluence of all sewage discharge units in the monitoring basin, scan their three-dimensional fluorescence spectra, and perform parallel factor analysis together with the three-dimensional fluorescence spectra in the database to split them into m components, and extract the fluorescence peak value of each component as the characteristic value;

[0047] (3) Arrange the configuration information of the mixed sewage into a matrix I of size (i+1)×n, add a column of unit column vectors to the left of the matrix I, and record it as matrix IN; Arrange the eigenvalues ​​of the mixed sewage into a matrix C of size (i+1)×m, and solve the linear model matrix M so that IN·M=C;

[0048] (4) The eigenvalues ​​of the actual water sample are organized into an m-dimensional column vector c, and the quantized information of the drainage composition of the n sewage discharge units contained in the actual water sample is recorded as an n-dimensional column vector x. Through the matrix equation M T [1; x] = c finds the relationship between the elements in x;

[0049] (5) When n is less than or equal to m, x has a unique solution; when n is greater than m, the specific composition of the elements in the column vector x is [x1, x2, …, x m ,x m+1 ,…,x n-1 ,x n ] T , the specific elements of the discharge information column vector of the conventional discharge water sample "standard" are [s1, s2, ..., s n ] T , construct the Euclidean distance function f(x1,x2,…,x n )=(x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 , using the matrix equation M T ·[1;x]=c Eliminate the function f and get f(x1,x2,…,x m ,x m+1 ,…x n )=g(x1,x2,…,x m ), find the unique solution x that minimizes the function g;

[0050] (6) The numerical value of each element in the column vector x corresponds to the composition relationship of the wastewater of each upstream sewage discharge unit in the actual water sample collected downstream. By comparing with the configuration information of the mixed water sample under the standard discharge condition, it is possible to judge whether there is over-discharge in the upstream and determine the range of possible sewage discharge units that may cause over-discharge.

[0051] In the technical solution provided by the present invention, the number of mixed sewage portions i should be no less than 20;

[0052] On the basis of the technical solution provided by the present invention, the configuration information is any one of the volume, volume proportion or concentration ratio of the wastewater of each sewage discharge unit in the mixed sewage;

[0053] On the basis of the technical solution provided by the present invention, the wavelength range of the excitation wavelength and the emission wavelength of the three-dimensional fluorescence spectrum is between 200nm and 900nm, and the scanning wavelength interval is 1nm to 5nm;

[0054] Based on the technical solution provided by the present invention, the components obtained by parallel factor analysis splitting should be cross-validated;

[0055] On the basis of the technical solution provided by the present invention, the method for solving the linear model matrix is ​​any one of the least squares method, principal component analysis, partial least squares method or support vector machine, and its content format should be organized into the first line as the constant term of each component expression, and the following n lines are the coefficients of the drainage configuration information of n sewage discharge units in each component expression;

[0056] Based on the technical solution provided by the present invention, the expression of function f is (x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 or |x1-s1|+|x2-s2|+…+|x n -s n |One of;

[0057] On the basis of the technical solution provided by the present invention, when judging the actual over-discharge situation and the source of the over-discharged sewage, it is considered that the confidence level of over-discharge is very low if the value of the quantified configuration information of the monitored water sample increases by 10% to 20% compared with the value of the quantified configuration information of the standard discharge.

[0058] Embodiment 1:

[0059] An application of a method for monitoring pollution over-discharge and tracing the source of over-discharge at a downstream end with multiple sources and complex pollution based on three-dimensional fluorescence, specifically comprising the following steps:

[0060] (1) Wastewater discharged from five factories A, B, C, D, and E in an industrial park was collected. According to the emission quotas of the five factories and the average flow rate of non-industrial polluted water from other sources, mixed sewage samples with standard discharge conditions of 3:3:3:3:3:3:10 were prepared, and 20 additional mixed sewage samples were prepared to simulate various non-standard discharge conditions as much as possible. All configuration information was recorded in the form of volume ratio, and then the excitation wavelength range was set to 200nm-500nm, the emission wavelength was set to 250nm-550nm, and the wavelength interval was 5nm for scanning. The 21 three-dimensional fluorescence spectra obtained were used to establish a database corresponding to the configuration information.

[0061] (2) A water sample of the day when the discharge volume of Factory A exceeded the quota by 50% was collected from the downstream pipeline after the rainwater and sewage were combined in the industrial park as the actual water sample. The excitation wavelength range was set to 200nm-500nm, the emission wavelength was set to 250nm-550nm, and the wavelength interval was 5nm for scanning to obtain a three-dimensional fluorescence spectrum. The spectrum was subjected to parallel factor analysis together with the other 21 three-dimensional fluorescence spectra in the database of step (1). The kernel consistency function was used to confirm that the number of components was 3, see Figure 2 After further cross-validation, the three-dimensional fluorescence spectrum of each sample was split into three components of uniform form. Figure 3 , the three components are shown in Figure 4-6 . Extract the peak value of fluorescence in the three-dimensional fluorescence spectrum of each component as a characteristic value;

[0062] (3) The configuration information of each mixed sewage in (1) is organized into a matrix IN, and the characteristic values ​​of each mixed sewage in (2) are organized into a matrix C. The data in the matrix IN and the matrix C are used as dependent variables and independent variables to perform partial least squares fitting with 5 principal components to obtain the fitting results of the three component characteristic values ​​and the linear model matrix M. The fitting results are shown in Figure 7-9 ;

[0063] (4) The eigenvalues ​​of the actual water sample are sorted into a 3D column vector c, and the matrix equation M T ·[1;x]=c to obtain the drainage volume ratio column vector x of the five sewage discharge units in the actual water sample;

[0064] (5) Because the number of pollution sources 5 is greater than the number of components 3, the specific values ​​of each element in x cannot be directly obtained. Therefore, the Euclidean distance function f(x1, x2, ..., x5) = (x1-3) is constructed between the volume ratio of wastewater discharged by each factory in the actual water sample and the volume ratio of wastewater discharged by each factory in the mixed wastewater under standard discharge conditions. 2 +(x2-3) 2 +(x3-3) 2 +(x4-3)2 +(x5-3) 2 , using the quantitative relationship between the drainage volume ratios of each sewage discharge unit reflected in the column vector x, the function f is eliminated and simplified to a binary function g(x4,x5). In MATLAB, x4=3.2872 and x5=3.3313 are obtained to minimize the function g. Therefore, x1=4.7790, x2=3.3312, and x3=3.0317. Therefore, the actual samples come from the sewage discharged by the five factories A, B, C, D, and E, and the volume ratio is 4.7790:3.3312:3.0317:3.2872:3.3313. Compared with the volume ratio of the standard discharge, it is judged that there is an over-discharge situation, and the A factory obviously over-discharges. The volume of the B, C, D, and E factories increases by less than 15% compared with the standard situation. It is within the allowable fluctuation range and is not considered to be over-discharged, which is consistent with the actual situation of the actual water samples collected. Fig.10 .

[0065] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A pollution source tracing method based on three-dimensional fluorescence spectroscopy, characterized in that: The method comprises the following steps: Step 1: Collect the independent drainage of n pollution sources in the upstream of the monitoring basin, and 1 portion of mixed drainage when the pollution sources merge and discharge in the downstream, which is recorded as "standard"; use the independent drainage collected upstream to configure i portions of mixed sewage according to different volume ratios, and together with the "standard", form a total of i+1 portions of mixed sewage samples, record the drainage configuration information of the pollution sources contained in each portion of mixed sewage, and scan the three-dimensional fluorescence spectrum of the mixed sewage to form a one-to-one corresponding database with the configuration information; Step 2: Collect actual water samples at the downstream end where all sewage discharge units in the monitoring basin converge, scan their three-dimensional fluorescence spectra, and perform parallel factor analysis together with the three-dimensional fluorescence spectra in the database to split them into m components, and extract the peak value of the fluorescence peak of each component as the characteristic value; Step 3: Arrange the configuration information of the mixed sewage into a matrix I of size (i+1)×n, add a column of unit column vectors to the left of the matrix I, and record it as matrix IN; Arrange the eigenvalues ​​of the mixed sewage into a matrix C of size (i+1)×m, and solve the linear model matrix M so that IN·M=C; Step 4: Arrange the eigenvalues ​​of the actual water sample into an m-dimensional column vector c, record the quantified information of the drainage composition of the n sewage discharge units contained in the actual water sample as an n-dimensional column vector x, and use the matrix equation M T [1; x] = c finds the relationship between the elements in x; Step 5: When n is less than or equal to m, x has a unique solution; when n is greater than m, the specific composition of the elements in the column vector x is [x1, x2, …, x m ,x m+1 ,…,x n-1 ,x n ] T , the specific elements of the discharge information column vector of the conventional discharge water sample "standard" are [s1, s2, ..., s n ] T , construct the Euclidean distance function f(x1,x2,…,x n )=(x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 , using the matrix equation M T ·[1;x]=c Eliminate the function f and get f(x1,x2,…,x m ,x m+1 ,…x n )=g(x1,x2,…,x m ), find the unique solution x that minimizes the function g; Step 6: Determine the wastewater over-discharge situation by comparing the configuration information of the mixed water sample with the standard discharge situation.

2. The method according to claim 1, characterized in that The number i of the mixed sewage prepared is not less than 20.

3. The method according to claim 1, characterized in that The configuration information is any one of the volume, volume proportion or concentration ratio of the wastewater of each sewage discharge unit in the mixed sewage.

4. The method according to claim 1, characterized in that The wavelength range of the excitation wavelength and the emission wavelength of the three-dimensional fluorescence spectrum is between 200nm and 900nm, and the scanning wavelength interval is between 1nm and 5nm.

5. The method according to claim 1, characterized in that The components obtained by the parallel factor analysis splitting should be cross-validated.

6. The method according to claim 1, characterized in that The method for solving the linear model matrix is ​​any one of the least squares method, principal component analysis, partial least squares method or support vector machine. The content format should be organized into the first line as the constant term of each component expression, and the next n lines are the coefficients of the drainage configuration information of n sewage discharge units in each component expression.

7. The method according to claim 1, characterized in that The expression of the function f is (x1-s1) 2 +(x2-s2) 2 +…+(x n -s n ) 2 or |x1-s1|+|x2-s2|+…+|x n -s n |One of them.

8. The method according to claim 1, characterized in that The sewage over-discharge situation in step 6 is specifically to determine whether there is over-discharge and to determine the scope of possible sewage discharge units that may cause over-discharge.

9. The method according to claim 1, characterized in that The value of each element in the column vector x corresponds to the composition relationship of the wastewater discharged by each upstream sewage discharge unit in the actual water sample collected downstream.

10. The method according to claim 1, characterized in that When judging the actual over-discharge situation and the source of the over-discharged sewage, an increase of 10% to 20% in the quantified configuration information value of the monitored water sample compared with the quantified configuration information value of the standard discharge indicates a low confidence level of over-discharge.

Images