CN106339536B - Comprehensive Evaluation of Water Quality based on water pollution index's method and cloud model - Google Patents

Abstract

A kind of Comprehensive Evaluation of Water Quality based on water pollution index's method and cloud model is to improve water pollution index's method using cloud model on the basis of water pollution index's method.Cloud model is combined with the WPI value in water pollution index's method, WPI value evaluation cloud model is obtained by cloud parameter determination method, single-factor WPI value cloud model is obtained by backward cloud generator, WPI value evaluation cloud atlas and single-factor WPI value cloud atlas are generated with normal state cloud generator, intuitively finds out the pollution situation of each single-factor；In conjunction with weight, comprehensive WPI value cloud model is obtained by comprehensive cloud algorithm, and showed with water dust formal intuition, determine Water Quality Evaluation grade, effectively solves the problems, such as not considering whole pollution factors in the contribution in comprehensive water quality assessment and the comparison being difficult to realize between different section in water pollution index's method.

Description

Comprehensive evaluation method of water quality based on water pollution index method and cloud model

technical field

The invention belongs to the technical field of information processing, and relates to a comprehensive water quality evaluation method based on a water pollution index method and a cloud model.

Background technique

The water environment system is a complex system integrating fuzziness and randomness. In order to ensure the rationality of water quality evaluation results, the fuzziness and randomness in the evaluation process must be comprehensively considered.

Existing literature: "China Environmental Monitoring", 2013 03 published an article "The Application of Water Pollution Index Method in River Water Quality Evaluation", the article disclosed the evaluation idea of water pollution index method:

Based on the evaluation principle of the single-factor evaluation method, according to the corresponding table of water quality category and WPI value (water pollution index value) (see Table 3), the WPI value of each water quality evaluation project of a certain section is calculated by interpolation method, The highest WPI value is taken as the WPI value of the section.

Table 3 Corresponding table of water quality category and WPI value

Water quality category Class I Class II Class III Class IV Class V Inferior Class V WPI range WPI=20 20<WPI≤40 40<WPI≤60 60<WPI≤80 80<WPI≤100 WPI>100

The calculation steps of the water pollution index method are as follows:

(1) Calculation method of index WPI value when the water pollution index does not exceed the limit of Class V water:

C _l (i)＜C(i)≤C _h (i)

In the formula, C(i) is the actual monitoring value of the i-th water quality indicator, C _l (i) and C _h (i) are the lower and upper limit values of the category standard of the i-th water quality indicator, respectively, WPI _l (i), WPI _h (i) are the index values corresponding to the lower limit value and upper limit value of the category standard of the i-th water quality indicator, respectively, and WPI(i) is the index value corresponding to the i-th water quality indicator.

In addition, when 6<pH<9,

WPI(pH)=20

(2) Calculation method of indicator WPI value when it exceeds the limit of Class V water:

In the formula, C ₅ (i) is the standard concentration limit of Class V in the water quality category of the i project.

Furthermore, when pH < 6,

WPI(pH)=100+6.67×(6-pH)

When pH>9,

WPI(pH)=100+8.00×(pH-9)

(3) Determination of comprehensive water quality WPI value

WPI=MAX(WPI(i))

Compared with the single-factor evaluation method, although the water pollution index method continues the single-factor evaluation method with the most polluted index as the judgment of the water quality category, it can quantify the water quality status. According to the quantitative results, not only can the water quality category be judged intuitively, but also the temporal and spatial changes of water quality can be reflected. According to the application of the water pollution index method and the other four water quality evaluation methods, it is known that the water pollution index method can meet the needs of quantitative water quality evaluation, identification of main pollution indicators, water quality category evaluation and comparison of water quality inferior to Class V. The comparison results See Table 4.

Table 4 Application comparison of five water quality evaluation methods

Taken together, the prior art has the following disadvantages:

(1) In the evaluation process, 100% weight is assigned to the most polluted factor, the difference in the degree of influence of different pollution indicators on water quality is not considered, and the water quality monitoring information is not fully utilized.

(2) The WPI value of the most polluted index is used as the WPI value of the section. When the main pollution index is different, the comparability of WPI values of different sections is not very strong.

(3) In the evaluation of water quality category, the objective ambiguity and uncertainty in the water environment are not considered, and the water quality evaluation result is inferior to the water quality category obtained by the comprehensive evaluation method.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a comprehensive water quality evaluation method based on the water pollution index method and the cloud model, which combines the cloud model with the WPI value in the water pollution index method, and obtains the WPI value evaluation cloud model by the cloud parameter determination method. The single-factor WPI value cloud model is obtained by the reverse cloud generator, and the WPI value evaluation cloud map and the single-factor WPI value cloud map are generated by the normal cloud generator, and the pollution status of each single factor can be seen intuitively; The WPI value cloud model is visually displayed in the form of cloud droplets to determine the comprehensive water quality evaluation level, which effectively solves the problem that the water pollution index method does not consider the contribution of all pollution factors in the comprehensive water quality evaluation and it is difficult to achieve comparison between different sections. .

The technical scheme adopted in the present invention is a comprehensive evaluation method of water quality based on the water pollution index method and cloud model, which is carried out according to the following steps:

Step 1, determine the factor set, WPI value set and weight set for evaluation;

Step 2, establish a WPI value evaluation cloud model;

Step 3, obtain a single factor WPI value cloud model;

Step 4, obtain the comprehensive WPI value cloud model;

Step 5: Determine the water quality evaluation grade.

Further, the step 1 is carried out according to the following steps:

Step a, select the evaluation index, and determine the factor set;

First select the initial indicators; then use the principal component-correlation analysis method to screen the indicators, use the principal component analysis to delete redundant indicators, and then use the correlation analysis to delete the repetitive indicators, which not only ensures that the retained indicators have a significant impact on the evaluation results, It also ensures that the information overlap between the indicators is relatively low; the final selected evaluation indicators are used as factor sets;

The selection of the initial indicators depends on the evaluation object, and can be selected through the water quality monitoring report or the existing evaluation indicators;

The basic model of principal component analysis is:

In the formula, x _i represents the ith index (i=1,2,...,p); z _j represents the jth principal component (j=1,2,...,m); l _ij represents the jth principal component The principal component load corresponding to the ith index in

The steps of principal component analysis are as follows:

1: Calculate the correlation coefficient matrix R of the standardized data of the indicators;

In the formula, r _ij is the correlation coefficient between the i-th index and the j-th index, x _ki and x _kj are the values of the i-th index and the j-th index of the k-th evaluation object, respectively, and are the average values of the i-th and j-th indicators, respectively;

2: Calculate the eigenvalue λ _i of the correlation coefficient matrix R, the eigenvector e _i (i=1,2,...,p), the variance contribution rate ω _i and the cumulative contribution rate G(m);

λ _i represents the total variance of the original indicator data explained by the i-th principal component, and the variance contribution rate ω _i of the i-th principal component to the original indicator data is:

The cumulative contribution rate G(m) is

3: Select principal components according to eigenvalues or cumulative contribution rates, and determine the number m of principal components;

Principal component selection criteria: (1) take the principal component corresponding to the eigenvalue λ _i >1; (2) take the principal component corresponding to the cumulative contribution rate G(k) ≥ 85%;

4: Calculate the principal component factor loading l _ij ;

Let e _ij be the jth component of the eigenvector e _i of the ith index, and the factor loading formula is

5: Screen the indicators according to the absolute value of the factor loadings on the principal components;

The larger the absolute value of the factor load, the more obvious the influence of the index on the evaluation results, and the more it should be retained; the smaller the absolute value of the factor load, the weaker the influence of the index on the evaluation results, the more it should be eliminated;

The specific steps of correlation analysis are as follows:

1: Calculate the correlation coefficient between each evaluation index;

Let r _ij be the correlation coefficient between the i-th index and the j-th index, x _ki and x _kj be the values of the i-th index and the j-th index of the k-th evaluation object, respectively, and are the average values of the i-th and j-th indicators, respectively;

2: Given a critical value P (0<P<1), determine the removal and retention of indicators;

When |r _ij |<P, the two evaluation indicators are kept at the same time; when |r _ij |>P, if the impact between the two indicators is similar, one of the evaluation indicators can be deleted according to the judgment of importance , if the impact and meaning of the indicators are quite different, both indicators are retained;

Step b, determine the comment set, that is, the WPI value set;

Combined with the existing water quality evaluation standards, determine the water quality evaluation comment set, and give the WPI value range of each evaluation level corresponding to the comment set, and determine the WPI value set;

Step c, determine the weight set;

The combination weighting method based on AHP-CRITIC determines the weight, uses the AHP method to determine the subjective weight, uses the CRITIC method to determine the objective weight, uses the combination weight determination formula to obtain the combined weight of the evaluation indicators, and determines the weight set;

The specific steps of the AHP method to determine the weight are as follows:

1: Determine the water quality evaluation indicators;

2: Construct judgment matrix;

The element value in the judgment matrix is a quantitative index for judging the relative importance of each element, and the 1-9 scale method is generally used; the value of each factor in the judgment matrix is judged by people on the relative importance of each factor, and then according to a certain ratio The scale is obtained by quantifying judgments;

3: Calculate the maximum eigenvalue of the judgment matrix and its eigenvector, and determine the weight vector;

At present, there are many calculation methods for eigenvalues and eigenvectors, such as square root method, sum method, eigenroot method, least square method, etc.;

4: Consistency check;

In order to ensure the rationality of the conclusion, the judgment matrix needs to be checked for consistency. The steps are as follows:

(1) Calculate the consistency test index CI;

Let λ _max judge the maximum eigenvalue of the matrix, then

(2) Find the corresponding average random consistency index RI according to Table 1, wherein n represents the order of the judgment matrix;

Table 1 Average Stochastic Consistency Index

n 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51

(3) Calculate the random consistency ratio CR, the calculation formula is

When CR<0.10, it is considered that the judgment matrix has satisfactory consistency, otherwise, it is necessary to adjust the judgment matrix to make it have satisfactory consistency;

The specific steps for determining the weight by the CRITIC method are as follows:

1: Calculate the standard deviation of the indicator sample;

Let σ _j be the standard deviation of the jth indicator, and the formula for calculating the standard deviation is:

Among them, N is the number of samples, x _i is the sample value, is the sample mean;

Step 2: Calculate

Step 3: Calculate the amount of information contained in the indicator;

Let C _j be the amount of information contained in the jth index, and its calculation formula is:

The larger the C _j , the greater the amount of information contained in the jth index, and the greater the relative importance of the index;

4: Calculate the objective weight of the indicator;

Let ω _j be the objective weight of the jth index, and its calculation formula is:

The formula for determining the combined weight is as follows:

In the formula, W _j is the combined weight of the j-th indicator, is the subjective weight of the jth index obtained by the AHP method, is the objective weight of the jth index obtained by the CRITIC method.

Further, the step 2 is carried out according to the following steps:

Step a, the determination of cloud parameters;

The cloud model uses the three numerical features of expectation Ex, entropy En and super entropy He to represent a qualitative concept as a whole;

Expectation Ex: The expectation of the distribution of cloud droplets in the universe of discourse space is the central value of the concept in the universe of discourse space, and it is the point that can best represent the qualitative concept, and its membership degree is 1, that is, it belongs to the qualitative concept 100%;

Entropy En: It is a measure of the uncertainty of a qualitative concept, which is jointly determined by the randomness and ambiguity of the qualitative concept; En is a measure of the randomness of the qualitative concept, reflecting the discrete degree of cloud droplets that can represent the qualitative concept; , En also reflects the margin of the qualitative concept and the other, and reflects the value range of cloud droplets in the universe space that can be accepted by the qualitative concept, which is a measure of the ambiguity of the qualitative concept; The larger the accepted value range of cloud droplets, the more ambiguous the qualitative concept; the use of the same digital feature to reflect randomness and ambiguity must also reflect the correlation between them;

Hyperentropy He: is a measure of the uncertainty of entropy, that is, the entropy of entropy, which is jointly determined by the randomness and ambiguity of entropy, and reflects the cohesion of the uncertainty of all points representing the language value in the universe of discourse space, Its size can represent the dispersion and thickness of the cloud;

When calculating the digital features of the comment cloud, the bilateral constraints [C _min , C _max ] are used to determine the cloud parameters, and the calculation formula is as follows:

Ex=(C _min +C _max )/2

En=(C _max -C _min )/6

He=k

Among them, k is a constant set according to the comment itself, which reflects the ambiguity of the comment; [C _min , C _max ] represents the WPI value range corresponding to each evaluation level in the comment set;

When there are only unilaterally constrained comments, the determination of cloud parameters needs to combine the upper and lower limits of the data to determine the parameters of its default boundaries, and then calculate with reference to the above formula; on the basis of the above formula, the obtained cloud parameter determination method is as follows:

1) Cloud parameters corresponding to evaluation interval 1 (0, a):

Ex1=0

En1=a/3

He1=k

2) Cloud parameters corresponding to the evaluation interval i (C _min , C _max ) (0<i<n):

Ex _i =(C _min +C _max )/2

En _i =(C _max -C _min )/6

He _i = k

3) The cloud parameters corresponding to the evaluation interval n(m, +∞):

Ex= _Cmin '+ _Cmax '

En= _Cmin '/3

He=k

C _min ' and C _max ' are the upper and lower limits of the evaluation interval n-1, respectively;

Taking the evaluation index with six evaluation intervals (0, a], (a, b], (b, c], (c, d], (d, e], [e, +∞)) as an example, its evaluation The parameters of the cloud model are determined as shown in Table 2;

Table 2 Determination method of cloud model parameters (Ex, En, He)

Cloud Ex En He C1 Ex1=0 En1=a/3 k C2 Ex2=(a+b)/2 En2=(b-a)/6 k C3 Ex3=(b+c)/2 En3=(c-b)/6 k C4 Ex4=(c+d)/2 En4=(d-c)/6 k C5 Ex5=(d+e)/2 En5=(e-d)/6 k C6 Ex6=d+e En6=d/3 k

Step b, the generation of cloud model;

Evaluate the cloud model parameters Ex, En, He according to the determined WPI value, and use the normal cloud generator to generate the corresponding WPI value evaluation cloud map;

The specific process is as follows:

Input: cloud model digital features (Ex, En, He) and the number of cloud droplets generated n;

Output: n cloud titration data x _i and the certainty y _i of the corresponding qualitative concept (i=1,2,...,n);

Algorithm steps:

1: Generate a normal random number Enn with En as the expected value and He ² as the variance;

2: Generate a normal random number _xi with Ex as the expected value and Enn ² as the variance, that is, _xi is a specific quantitative realization of the qualitative concept A in its corresponding quantitative universe, which is called cloud titration data;

3: Calculation

Define y _i as the degree of certainty that _xi belongs to qualitative concept A, and ( _xi , y _i ) as cloud droplets;

4: Repeat the above steps until n cloud droplets (x _i , y _i ) (i=n) are generated.

Further, the step 3 is carried out according to the following steps:

Step a, based on the water pollution index method, calculate the WPI value corresponding to each water quality evaluation factor;

The water pollution index method is based on the evaluation principle of the single-factor evaluation method. According to the corresponding table of water quality category and WPI value (see Table 3), the WPI value of each water quality evaluation project in a certain section is calculated by interpolation method, and the highest WPI value is obtained. value as the WPI value of the section;

Table 3 Corresponding table of water quality category and WPI value

Water quality category Class I Class II Class III Class IV Class V Inferior Class V WPI range WPI=20 20<WPI≤40 40<WPI≤60 60<WPI≤80 80<WPI≤100 WPI>100

The formula for calculating the WPI value of each water quality evaluation factor is as follows:

1) The calculation method of the WPI value of the indicator when the water pollution index does not exceed the limit of Class V water:

C _l (i)＜C(i)≤C _h (i)

In the formula, C(i) is the actual monitoring value of the i-th water quality indicator, C _l (i) and C _h (i) are the lower and upper limit values of the category standard of the i-th water quality indicator, respectively, WPI _l (i), WPI _h (i) are the index values corresponding to the lower limit value and the upper limit value of the category standard of the i-th water quality indicator, respectively, and WPI(i) is the index value corresponding to the i-th water quality indicator;

In addition, when 6<pH<9,

WPI(pH)=20

WIP (pH) is the WPI value corresponding to the indicator pH value;

2) Calculation method of indicator WPI value when it exceeds the limit of Class V water:

In the formula, C ₅ (i) is the standard concentration limit of Class V in the water quality category of the i project;

Furthermore, when pH < 6,

WPI(pH)=100+6.67×(6-pH)

When pH>9,

WPI(pH)=100+8.00×(pH-9)

Step b, generating a single-factor WPI value cloud model;

Using the inverse cloud algorithm based on X information to calculate the parameters of the cloud model, just use the quantitative values of cloud droplets X(x _i ) to restore the three parameters of the cloud, without the need for the value of the degree of certainty Y(y _i ), the process is as follows:

Input: sample point x _i (i=1,2,...,n);

Output: The cloud model numerical features (Ex, En, He) of the n cloud droplets corresponding to qualitative concepts;

Algorithm steps:

1: Calculate the sample mean of this set of data based on _xi First-order sample absolute center distance sample variance

2: Calculate expectations

3: Calculate entropy

4: Calculate super entropy

The cloud drop _xi in the algorithm refers to the WPI value corresponding to the actual monitoring data of water quality evaluation indicators collected in practical applications.

Further, the step 4 is carried out according to the following steps: using the comprehensive cloud algorithm, combining the weights, synthesizing each single factor WPI value cloud model to generate a comprehensive WPI value cloud model, and the comprehensive cloud algorithm formula is:

Further, the step 5 is performed according to the following steps: using a normal cloud generator to generate a WPI value evaluation cloud map and a comprehensive WPI value cloud map, and compare them to finally determine the comprehensive water quality evaluation level.

The beneficial effects of the invention are that the evaluation language is fuzzified, the indexes with randomness and fuzziness are quantified, and the fuzziness and randomness of the water quality evaluation factors are represented by the three digital features of the cloud model, which can not only compare Intuitively see the pollution status of each single factor, and visually display the comprehensive water quality grade of seepage in the form of cloud droplets, so that the evaluation results are more realistic, which can effectively make up for the shortcomings and deficiencies of the water pollution index method. The study of methods provides more optimized models or new ideas.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

Fig. 1 is a flow chart of the comprehensive evaluation method of water quality in the present invention.

Figure 2 is a cloud map of WPI value evaluation.

Figure 3 is a single factor WPI value cloud map.

Figure 3a is the cloud model of the WPI value of pH; Figure 3b is the cloud model of the WPI value of U; Figure 3c is the cloud model of the WPI value of Ra; Figure 3d is the cloud model of the WPI value of Σβ; Figure 3e is the cloud model of NH ₄ -N WPI value cloud model; Figure 3f is the WPI value cloud model of Mn; Figure 3g is the WPI value cloud model of F ions; Figure 3h is the WPI value cloud model of Zn.

Figure 4 is a comprehensive WPI value cloud map.

Figure 4a is the cloud map of the comprehensive WPI value of seepage in dam section A; Figure 4b is the cloud map of the comprehensive WPI value of water seepage in dam section B; Figure 4c is the cloud map of the comprehensive WPI value of water seepage in dam section C; Figure 4d is the cloud map of the comprehensive WPI value of water seepage in dam section D.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

A comprehensive water quality evaluation method based on water pollution index method and cloud model, which is specifically carried out according to the following steps:

Taking the evaluation of seepage water quality of decommissioned uranium tailings pond as an example, on the basis of the given index system and index weight of seepage water quality of decommissioned uranium tailings pond, the cloud model is used to improve the water pollution index method to conduct comprehensive evaluation of water quality. The specific steps are as follows: Figure 1.

1) Determination of factor set, comment set and weight set

According to the existing environmental monitoring reports of decommissioned uranium tailings ponds, it is determined that there are 12 primary indicators for evaluating the seepage water quality of decommissioned uranium tailings ponds: pH, U, Ra, ∑α, ∑β, ²³⁰ Th, ²¹⁰ Po, ²¹⁰ Pb, NH ₄ ^{-N, Mn, F-,} Zn, based on the principal component-correlation analysis method for index screening, determine the final seepage water quality evaluation index is pH, U, Ra, ∑β, NH ₄ -N, Mn, F ^- , Zn.

Based on the AHP-CRITIC combination weighting method, the subjective weight is determined by the AHP method as (0.0605, 0.2822, 0.1760, 0.1760, 0.049, 0.0605, 0.0386, 0.1032), and the objective weight is determined by the CRITIC method as (0.1088, 0.0783, 0.1049, 0.1808, 0.0998, 0.1731, 0.0910, 0.1632), substitute the subjective weight and the objective weight into the combined weight determination formula, and determine its weight set as (0.0534, 0.1793, 0.1498, 0.2582, 0.049, 0.1450, 0.0285, 0.1369).

Combined with GB8978 "Comprehensive Sewage Discharge Standard", on the basis of the existing water quality pollution index classification standard (see Table 5), the seepage water quality is divided into six grades, that is, the comment set is {"clean", "still clean", "Mild pollution", "moderate pollution", "severe pollution", and "severe pollution"}, respectively recorded as Ⅰ, Ⅱ, Ⅲ, Ⅳ, Ⅴ, Ⅵ; according to the WPI value in the existing water pollution index method Evaluation range (see Table 3), determine the WPI value set as {[0, 20], (20, 40], (40, 60], (60, 80], (80, 100], (100, +∞)) }; See Table 6 for the finalized evaluation criteria for seepage water quality of decommissioned uranium tailings ponds.

Table 5 The existing water pollution index classification standards

water quality index level Grading Basis <0.2 clean Multiple items not checked out 0.2～0.4 still clean The detected values are all within the standard value 0.4～0.7 light pollution 1 detected value exceeds the standard 0.7～1.0 Moderately polluted 2 detected values exceeded the standard 1.0～2.0 heavy pollution All or most items exceed the standard >2.0 heavily polluted All or most items exceed the standard > 1 times

Table 6 Classification standard of seepage water quality of decommissioned uranium tailings pond

2) Establishment of WPI value evaluation cloud model

(1) Determination of cloud parameters

When calculating the digital features of the comment cloud, the bilateral constraints [C _min , C _max ] are used to determine the cloud parameters, and the calculation formula is as follows:

Ex=(C _min +C _max )/2

En=(C _max -C _min )/6

He=k

Among them, k is a constant set according to the comment itself, which reflects the ambiguity of the comment; [C _min , C _max ] represents the WPI value range corresponding to each evaluation level in the comment set;

When there are only unilaterally constrained comments, the determination of cloud parameters needs to combine the upper and lower limits of the data to determine the parameters of its default boundaries, and then calculate with reference to the above formula; on the basis of the above formula, the obtained cloud parameter determination method is as follows:

1) Cloud parameters corresponding to evaluation interval 1 (0, a):

Ex1=0

En1=a/3

He1=k

2) Cloud parameters corresponding to the evaluation interval i (C _min , C _max ) (0<i<n):

Ex _i =(C _min +C _max )/2

En _i =(C _max -C _min )/6

He _i = k

3) The cloud parameters corresponding to the evaluation interval n(m, +∞):

Ex= _Cmin '+ _Cmax '

En= _Cmin '/3

He=k

C _min ' and C _max ' are the upper and lower limits of the evaluation interval n-1, respectively.

Taking the evaluation index with six evaluation intervals (0, a], (a, b], (b, c], (c, d], (d, e], [e, +∞)) as an example, its evaluation The parameters of the cloud model are determined as shown in Table 2.

Table 2 Determination method of cloud model parameters (Ex, En, He)

According to the water quality classification standard in Table 5, the method in Table 2 is used to determine the parameters of the WPI value evaluation cloud model, which are (0, 6.67, 1), (30, 3.33, 1), (50, 3.33, 1), (70,3.33,1), (90,3.33,1), (180,26.67,1).

(2) Generation of WPI value evaluation cloud model

The cloud model parameters Ex, En, and He are evaluated according to the determined WPI value, and the corresponding WPI value evaluation cloud map is generated by using the normal cloud generator. The process is as follows:

Input: cloud model digital features (Ex, En, He) and the number n of cloud droplets generated.

Output: n cloud titration data x _i and their corresponding qualitative concepts y _i (i=1,2,...,n).

Algorithm steps:

1: Generate a normal random number Enn with En as the expected value and He ² as the variance;

2: Generate a normal random number _xi with Ex as the expected value and Enn ² as the variance, that is, _xi is a specific quantitative realization of the qualitative concept A in its corresponding quantitative universe, which is called cloud titration data;

3: Calculation

4: Define y _i as the degree of certainty that x _i belongs to the qualitative concept A, and (x _i , y _i ) are cloud droplets;

5: Repeat the above steps until n cloud droplets are generated.

The WPI value evaluation cloud map generated by the normal cloud generator is shown in Figure 2.

3) Obtain the single factor WPI value cloud model

(1) Calculation of WPI value corresponding to each water quality evaluation factor

According to the measured data of each water quality evaluation factor, using the WPI value calculation formula, the WPI value corresponding to each water quality evaluation factor is obtained.

According to the monitoring data of the water seepage indicators of the four dam sections in 2012 (January to December), the WPI value corresponding to each index of each dam section was obtained by using the WPI value calculation formula.

Table 7 WPI value of seepage index of dam section A in 2012

(2) Single-factor WPI value cloud model generation

Using the new reverse cloud algorithm based on X information to calculate the parameters of the cloud model, just use the quantitative value of the cloud droplet X to restore the three parameters of the cloud, without the value of the degree of certainty Y, the process is as follows:

Input: sample points x _i (i=1,2,...,n).

Output: These n cloud droplets correspond to the cloud model numerical features (Ex, En, He) of qualitative concepts.

Algorithm steps:

1: Calculate the sample mean of this set of data based on x _i First-order sample absolute center distance sample variance

2: Calculate expectations

3: Calculate entropy

4: Calculate super entropy

The cloud drop _xi in the algorithm refers to the WPI value corresponding to the actual monitoring data of the water quality evaluation index collected in practical applications.

According to the WPI value of the water seepage index, and according to the new reverse cloud algorithm based on X information, the parameters of the single-factor WPI value cloud model corresponding to the eight indicators are calculated, as shown in Table 8.

Table 8 One-factor WPI value cloud model

Taking dam section A as an example, according to the obtained single-factor WPI value cloud model, the normal cloud generator is used to generate the WPI value cloud map corresponding to 8 single factors, and compare with the WPI value evaluation cloud map, as shown in Figure 3 . In the figure, gray is the single-factor WPI value cloud model, and pure black is the WPI value evaluation cloud model.

4) Determination of comprehensive WPI value cloud model

The comprehensive cloud algorithm formula is:

According to the determined single-factor WPI value cloud model, combined with the weight of the seepage index, and substituted into the comprehensive cloud calculation formula, the comprehensive WPI value cloud model of the seepage of each dam section is obtained, as shown in Table 9, and the normal cloud generator is used to generate each dam section. The comprehensive WPI value cloud map of seepage is shown in Figure 4. The red is the comprehensive WPI value cloud map of each dam section, and the blue is the WPI value evaluation cloud map.

Table 9 Comprehensive WPI value cloud model

Dam section Comprehensive WPI Value Cloud Model Dam Section A (62.6947, 5.5986, 4.2850) Dam section B (130.6272, 19.0690, 14.8482) Dam section C (69.1420,7.2551,1.0598) Dam section D (151.1792, 21.3879, 0.8707)

5) Comprehensive evaluation results and analysis

Through the comparison of the comprehensive WPI value cloud map and the WPI value evaluation cloud map, the water quality evaluation results of the seepage water in each dam section are determined in Table 10.

Table 10 Comprehensive evaluation results

Dam section Comprehensive WPI Value Cloud Model Evaluation results major pollutants Dam Section A (62.6947, 5.5986, 4.2850) Ⅳ Mn(119), pH(108) Dam section B (130.6272, 19.0690, 14.8482) VI Mn(229), pH(110) Dam section C (69.1420,7.2551,1.0598) Ⅳ NH₄-N(158), Mn(112) Dam section D (151.1792, 21.3879, 0.8707) VI Mn(248), pH(102), NH₄-N(81)

Result analysis:

(1) Single-factor pollution level: From Figure 3, we can intuitively see the single-factor pollution level in the evaluation of seepage water quality of dam section A, where the pollution levels of pH and Mn are "severe pollution", U, Ra, ∑β, The pollution level of F ions and Zn is "clean", and the pollution level of NH ₄ -N is "still clean", so it can be determined that the main pollution indicators in the seepage water of dam section A are pH and Mn. At the same time, using the normal cloud to represent the single-factor pollution level, it is possible to compare the pollution levels of different pollution indicators of the same level. It can be concluded that the pollution degree of pH is lower than that of Mn.

(2) Comprehensive evaluation grade of water quality: It can be seen from Table 9 that the water quality grade of seepage in dam section A and dam section C is "moderately polluted", and the water quality grade of seepage in dam section B and dam section D is "severely polluted" ". It can be seen intuitively from Fig. 4 that the comprehensive WPI value cloud models of water seepage in dam section A and dam section C are both located between the two clouds of "light pollution" and "moderate pollution", and the comprehensive WPI value of seepage in dam section C is located between the two clouds. The value expectation is basically close to the WPI value evaluation expectation of "moderate pollution", which is significantly higher than the comprehensive WPI value cloud model of seepage in dam section A; the comprehensive WPI value cloud model of seepage in dam section B and dam section D are both located in "severe pollution". " and "severely polluted", and closer to "severely polluted", the expected comprehensive WPI value of seepage in dam section B is lower than the expected comprehensive WPI value of seepage in dam section D, but the WPI value cloud of seepage in dam section B The dispersion degree of droplets is much higher than the dispersion degree of cloud droplets in the WPI value of seepage in dam section D. The comprehensive comparison shows that the water pollution degree of seepage water in dam section B is the highest, followed by the water seepage in dam section D and the water seepage in dam section C, and the water seepage degree in dam section A is the lowest.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims (5)

1. a comprehensive evaluation method for water quality based on water pollution index method and cloud model, is characterized in that, carry out according to the following steps: Step 1, determine the factor set, WPI value set and weight set for evaluation; Step 2, establish a WPI value evaluation cloud model; Step 3, obtain a single factor WPI value cloud model; Step 4, obtain the comprehensive WPI value cloud model; Step 5, determine the water quality evaluation grade; Said step 1 is carried out according to the following steps: Step a, select the evaluation index, and determine the factor set; First select the initial index; then use the principal component-correlation analysis method to screen the index, use the principal component analysis to delete the redundant index, and then use the correlation analysis to delete the repetitive index, and the final selected evaluation index is used as the factor set; The selection of the initial indicators is determined according to the evaluation object, and is selected through the water quality monitoring report or the existing evaluation indicators; The basic model of principal component analysis is: In the formula, x _i represents the ith index (i=1,2,...,p); z _j represents the jth principal component (j=1,2,...,m); l _ij represents the jth principal component The principal component load corresponding to the ith index in The steps of principal component analysis are as follows: 1: Calculate the correlation coefficient matrix R of the standardized data of the indicators; In the formula, r _ij is the correlation coefficient between the i-th index and the j-th index, x _ki and x _kj are the values of the i-th index and the j-th index of the k-th evaluation object, respectively, and are the average values of the i-th and j-th indicators, respectively; 2: Calculate the eigenvalue λ _i of the correlation coefficient matrix R, the eigenvector e _i (i=1,2,...,p), the variance contribution rate ω _i and the cumulative contribution rate G(m); λ _i represents the total variance of the original indicator data explained by the i-th principal component, and the variance contribution rate ω _i of the i-th principal component to the original indicator data is: The cumulative contribution rate G(m) is 3: Select principal components according to eigenvalues or cumulative contribution rates, and determine the number m of principal components; Principal component selection criteria: (1) take the principal component corresponding to the eigenvalue λ _i >1; (2) take the principal component corresponding to the cumulative contribution rate G(k) ≥ 85%; 4: Calculate the principal component factor loading l _ij ; Let e _ij be the jth component of the eigenvector e _i of the ith index, and the factor loading formula is 5: Screen the indicators according to the absolute value of the factor loadings on the principal components; The larger the absolute value of the factor load, the more obvious the influence of the index on the evaluation result, and the more it should be retained; the smaller the absolute value of the factor load, the weaker the influence of the index on the evaluation result, and the more it should be eliminated; The specific steps of correlation analysis are as follows: 1: Calculate the correlation coefficient between each evaluation index; Let r _ij be the correlation coefficient between the i-th index and the j-th index, x _ki and x _kj be the values of the i-th index and the j-th index of the k-th evaluation object, respectively, and are the average values of the i-th and j-th indicators, respectively; 2: Given a critical value P (0<P<1), determine the removal and retention of indicators; When |r _ij |<P, the two evaluation indicators are kept at the same time; when |r _ij |>P, if the impact between the two indicators is similar, one of the evaluation indicators can be deleted according to the judgment of importance , if the impact and meaning of the indicators are quite different, both indicators are retained; Step b, determine the comment set, that is, the WPI value set; Combined with the existing water quality evaluation standards, determine the water quality evaluation comment set, and give the WPI value range of each evaluation level corresponding to the comment set, and determine the WPI value set; Step c, determine the weight set; The combination weighting method based on AHP-CRITIC determines the weight, uses the AHP method to determine the subjective weight, uses the CRITIC method to determine the objective weight, uses the combination weight determination formula to obtain the combined weight of the evaluation indicators, and determines the weight set; The specific steps of determining the subjective weight by the AHP method are as follows: 1: Determine the water quality evaluation indicators; 2: Construct judgment matrix; The element value in the judgment matrix is a quantitative index for judging the relative importance of each element, and the 1-9 scale method is used; the value of each factor in the judgment matrix is judged by people on the relative importance of each factor, and then according to the ratio scale. Judgment and quantification; 3: Calculate the maximum eigenvalue of the judgment matrix and its eigenvector, and determine the weight vector; 4: Consistency check; To check the consistency of the judgment matrix, the steps are as follows: (1) Calculate the consistency test index CI; Let λ _max judge the maximum eigenvalue of the matrix, then (2) Find the corresponding average random consistency index RI according to Table 1, wherein n represents the order of the judgment matrix; Table 1 Average Stochastic Consistency Index n 1 2 3 4 5 6 7 8 9 10 11 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 1.51 (3) Calculate the random consistency ratio CR, the calculation formula is When CR<0.10, it is considered that the judgment matrix has satisfactory consistency, otherwise, it is necessary to adjust the judgment matrix to make it have satisfactory consistency; The specific steps for determining the weight by the CRITIC method are as follows: 1: Calculate the standard deviation of the indicator sample; Let σ _j be the standard deviation of the jth indicator, and the formula for calculating the standard deviation is: Among them, N is the number of samples, x _i is the sample value, is the sample mean; Step 2: Calculate Step 3: Calculate the amount of information contained in the indicator; Let C _j be the amount of information contained in the jth index, and its calculation formula is: The larger the C _j , the greater the amount of information contained in the jth index, and the greater the relative importance of the index; 4: Calculate the objective weight of the indicator; Let ω _j be the objective weight of the jth index, and its calculation formula is: The formula for determining the combined weight is as follows: In the formula, W _j is the combined weight of the j-th indicator, is the subjective weight of the jth index obtained by the AHP method, is the objective weight of the jth index obtained by the CRITIC method. 2. a kind of water quality comprehensive evaluation method based on water pollution index method and cloud model according to claim 1, is characterized in that, described step 2 is carried out according to the following steps: Step a, the determination of cloud parameters; When calculating the digital features of the comment cloud, the bilateral constraints [C _min , C _max ] are used to determine the cloud parameters, and the calculation formula is as follows: Ex=(C _min +C _max )/2 En=(C _max -C _min )/6 He=k in, Expectation Ex: The expectation of the distribution of cloud droplets in the universe of discourse space is the central value of the concept in the universe of discourse space, and it is the point that can best represent the qualitative concept, and its membership degree is 1, that is, it belongs to the qualitative concept 100%; Entropy En: It is a measure of the uncertainty of a qualitative concept, which is jointly determined by the randomness and ambiguity of the qualitative concept; En is a measure of the randomness of the qualitative concept, reflecting the discrete degree of cloud droplets that can represent the qualitative concept; , En also reflects the margin of the qualitative concept and the other, and reflects the value range of cloud droplets in the universe space that can be accepted by the qualitative concept, which is a measure of the ambiguity of the qualitative concept; The larger the accepted value range of cloud droplets, the more ambiguous the qualitative concept; the use of the same digital feature to reflect randomness and ambiguity must also reflect the correlation between them; Hyperentropy He: is a measure of the uncertainty of entropy, namely the entropy of entropy, which is jointly determined by the randomness and ambiguity of entropy, and reflects the cohesion of the uncertainty of all points representing a specific language value in the universe of discourse space, Its size can represent the dispersion and thickness of the cloud; k is a constant set according to the comment itself, which reflects the ambiguity of the comment; [C _min , C _max ] represents the WPI value range corresponding to each evaluation level in the comment set; When there are only unilaterally constrained comments, the determination of cloud parameters needs to combine the upper and lower limits of the data to determine the parameters of its default boundaries, and then calculate with reference to the above formula; on the basis of the above formula, the obtained cloud parameter determination method is as follows: 1) Cloud parameters corresponding to evaluation interval 1 (0, a): Ex1=0 En1=a/3 He1=k 2) Cloud parameters corresponding to the evaluation interval i (C _min , C _max ) (0<i<n): Ex _i =(C _min +C _max )/2 En _i =(C _max -C _min )/6 He _i = k 3) The cloud parameters corresponding to the evaluation interval n(m, +∞): Ex= _Cmin '+ _Cmax ' En= _Cmin '/3 He=k C _min ' and C _max ' are the upper and lower limits of the evaluation interval n-1, respectively; Step b, the generation of cloud model; Evaluate the cloud model parameters Ex, En, He according to the determined WPI value, and use the normal cloud generator to generate the corresponding WPI value evaluation cloud map; The specific operation process is as follows: Input: cloud model digital features (Ex, En, He) and the number of cloud droplets generated n; Output: n cloud titration data x _i and the certainty y _i of the corresponding qualitative concept (i=1,2,...,n); The specific algorithm steps are: 1: Generate a normal random number Enn with En as the expected value and He ² as the variance; 2: Generate a normal random number _xi with Ex as the expected value and Enn ² as the variance, that is, _xi is a specific quantitative realization of the qualitative concept A in its corresponding quantitative universe, which is called cloud titration data; 3: Calculation Define y _i as the degree of certainty that _xi belongs to qualitative concept A, and ( _xi , y _i ) as cloud droplets; 4: Repeat the above steps until n cloud droplets (x _i , y _i ) (i=n) are generated. 3. a kind of water quality comprehensive evaluation method based on water pollution index method and cloud model according to claim 1, is characterized in that, described step 3 is carried out according to the following steps: Step a, based on the water pollution index method, calculate the WPI value corresponding to each water quality evaluation factor; The water pollution index method is based on the evaluation principle of the single-factor evaluation method, and according to the corresponding table of water quality categories and WPI values, the WPI value of each water quality evaluation project of a certain section is calculated by interpolation method, and the highest WPI value is taken as the section's WPI value. WPI value; The formula for calculating the WPI value of each water quality evaluation factor is as follows: 1) The calculation method of the WPI value of the indicator when the water pollution index does not exceed the limit of Class V water: C _l (i)＜C(i)≤C _h (i) In the formula, C(i) is the actual monitoring value of the i-th water quality indicator, C _l (i) and C _h (i) are the lower and upper limit values of the category standard of the i-th water quality indicator, respectively, WPI _l (i), WPI _h (i) are the index values corresponding to the lower limit value and the upper limit value of the category standard of the i-th water quality indicator, respectively, and WPI(i) is the index value corresponding to the i-th water quality indicator; In addition, when 6<pH<9, WPI(pH)=20; WIP (pH) is the WPI value corresponding to the indicator pH value; 2) Calculation method of indicator WPI value when it exceeds the limit of Class V water: In the formula, C ₅ (i) is the standard concentration limit of Class V in the water quality category of the i project; Furthermore, when pH < 6, WPI(pH)=100+6.67×(6-pH) When pH>9, WPI(pH)=100+8.00×(pH-9); Step b, generating a single-factor WPI value cloud model; Using the inverse cloud algorithm based on X information to calculate the parameters of the cloud model, just use the quantitative value of the cloud droplet X( _xi ) to restore the three parameters of the cloud, without the value of the degree of certainty Y( _yi ), the operation process is as follows : Input: sample point x _i (i=1,2,...,n); Output: The cloud model numerical features (Ex, En, He) of the n cloud droplets corresponding to qualitative concepts; The algorithm steps are: 1: Calculate the sample mean of this set of data based on _xi First-order sample absolute center distance sample variance 2: Calculate expectations 3: Calculate entropy 4: Calculate super entropy Cloud drop _xi refers to the WPI value corresponding to the actual monitoring data of water quality evaluation indicators collected in practical applications. 4. a kind of water quality comprehensive evaluation method based on water pollution index method and cloud model according to claim 1, is characterized in that, described step 4 is carried out according to the following steps: use comprehensive cloud algorithm, combine weight, synthesize each single factor The WPI value cloud model generates a comprehensive WPI value cloud model. The formula of the comprehensive cloud model is: 5. a kind of water quality comprehensive evaluation method based on water pollution index method and cloud model according to claim 1, is characterized in that, described step 5 is carried out according to the following steps: utilize normal cloud generator, generate WPI value evaluation cloud map And comprehensive WPI value cloud map, and compare, and finally determine the water quality comprehensive evaluation grade.