CN106682416A - Sewage enterprise water pollution source assessment method based on multi-index evaluation algorithm - Google Patents

Abstract

The invention provides a sewage enterprise water pollution source assessment method based on a multi-index evaluation algorithm. The method comprises the following steps of building sewage enterprise water pollution source factors; building a sewage enterprise water pollution source evaluation set; adopting a fuzzy transformation method for performing fuzzy evaluation on sewage enterprise water pollution source data; adopting a fuzzy evaluation matrix for performing primary data selection; adopting a fuzzy clustering analysis method for classifying primarily-selected object set, and screening out typical objects. A sewage source enterprise is comprehensively assessed through a fuzzy evaluation algorithm, the comprehensive, independent and operable principle for index selection is sufficiently considered, the evaluation result conforms to the subjective and objective consistency principle, and the evaluation result has the certain guiding significance in final water environment treatment and improvement construction.

Description

Sewage enterprise water pollution source appraisal procedure based on multiple index evaluation algorithm

Technical field

The present invention relates to sewage enterprise evaluation technical field, and in particular to a kind of sewage based on multiple index evaluation algorithm is looked forward to Industry water pollution source appraisal procedure.

Background technology

Water is a kind of irreplaceable resource, is the important material base of social development and economic development, is people's lives Important leverage, be one of important support of social sustainable development.With developing rapidly for China's economy, city-building and work Industry production constantly expands, and water consumption increases severely, and quantity of wastewater effluent is also sharply increased.Many water bodys suffer trade effluent, sanitary sewage Pollution, water pollution becomes the instant environmental problem of China.

Water quality evaluation is to carry out environmental key-element analysis to a certain water environment region, and quantitative commentary is made to it.It is logical Cross environmental quality assessment, understand fully the rule of regional environmental quality Change and Development, be Regional Environmental System pollution control planning and Formulate Regional Environmental System engineering proposal and foundation is provided.

The U.S. is the country with legal form certainly of the one one or two environmental evaluation in the world, makes a general survey of the development of environmental evaluation, Nineteen sixty-five U.S. L.A.Zadeh professors are famous《Fuzzy set》One text is delivered, and indicates the birth and quickly of fuzzy mathematics Grow up.Due to there are a large amount of uncertain factors in water environment, water quality level, criteria for classification are all some misty ideas, Therefore fuzzy mathematics is used widely in Water Quality Evaluation.

Water quality evaluation first has to make the source of sewage clear, and on the one hand the degree of reliability of the source results of sewage takes Certainly in the accuracy of Monitoring Data, the evaluation methodology of science, including the selection of technology are on the other hand depended on.Current generation has From single goal to multiple target, from single environmental key-element to multi-environment key element, from simple natural environment system to natural environment and society The integrated system of meeting environment, from static analysis to the trend of dynamic analysis development.

Dynamic comprehensive assessment to sewage enterprise water dirt material source, is one of main task of water environment treatment and construction, It is that there is critically important realistic meaning for further instructing and improving water environment, therefore the foundation of evaluation model needs two The points for attention of aspect：

(1) object for choosing reflected appraisal that can be comprehensive of index；

(2) the subjective and objective combination of whole evaluation procedure so that evaluation result has credibility.

In order to the sewage source for objectively reflecting industry is constituted, needs filter out real with representative from by respondent The user of property, i.e. typical user.For the selection of typical user, first, should be according to trade classification, according to practical experience at each The user that several are chosen in industry can reflect the sector production feature is investigated, and this is used according to the typical case of expertise Family primary election.Secondly, on the basis of user is investigated in primary election, prior art is by fuzzy C-means clustering FCM methods, actual And atypical user removes, and then carries out the synthesis of industrial nature.But the method is needed to smoothing factor and initial center square Battle array is selected, and the index evaluated is excessively single, it is impossible to accurately react actual situation.

The content of the invention

The defect that overall merit is carried out for sewage source system, the present invention is there is no to provide one in order to solve prior art Plant based on the sewage enterprise water pollution source appraisal procedure of multiple index evaluation algorithm, specifically include following steps：

1) sewage enterprise water pollution source factors system is set up, all specific targets are constituted in water pollution source factors system Set of factors X={ x₁,x₂,…,x_m, wherein m represents the number of factor；

2) sewage enterprise water pollution source appraisement system is set up, all specific targets are constituted in the appraisement system of water pollution source Evaluate collection Y={ y₁,y₂,…,y_n, wherein n represents the number of evaluation；

3) fuzzy evaluation is carried out to sewage enterprise water pollution derived data using the method for blurring mapping, obtains single factor test and comment Valency matrix：

4) primary election is carried out to data using fuzzy evaluating matrix：

Define factor weight distribution A=(a₁,a₂,…,a_m), wherein a_i>=0 andObtaining fuzzy evaluating matrix is

Wherein, each element in fuzzy evaluating matrix B is chosen according to maximum subjection principle：b_i=max { min (a_i, r_1i),…,min(a_m,r_mi)}；Typical subject is selected according to evaluate collection Y, primary election object set U={ u are obtained₁,u₂,…,u_p}；

5) primary election object set is classified using method of fuzzy cluster analysis, screens typical subject.

Step 1) in, sewage enterprise water pollution source factors include economic, employment, prediction and environment；Step 2) in, sewage The source evaluation of enterprise's water pollution includes important, more important, less important and inessential.

Step 5) in, method of fuzzy cluster analysis specifically includes following steps：

51) using the sewage source percentage ratio of primary election object set, the domain of sewage source enterprise is set up：

Wherein k is the number of pollutant kind, and p is the number of primary election object；

52) normalization process is carried out to data, obtains normalized matrix：

Normalization process adopts equation below：

Wherein u '_kmax、u′_kminIt is u '_1k, u '_2k..., u '_pkIn maximum and minima.

53) fuzzy similarity matrix is set up：

54) suitable cut set λ is chosen, primary election object is classified, obtain preferred typical subject.Choose different taking During cut set λ, different typical user's groups can be produced, usual index value difference is bigger, can more reflect the gap for being evaluated enterprise.

Sewage source modeling is one of problem of urgent need to resolve in current water utilities systematic analysiss.Carrying out systematic analysiss and pre- When surveying calculating, the result inconsistent with practical situation will be obtained come source model using different.Therefore by for evaluation index In qualitative index set up Comment gathers to realize the quantization to evaluation index, and establish membership function to realize to qualitative index Evaluation just can well be estimated and predict calculating.Meanwhile, by the cluster analyses to same user, and summarize one's own profession Industry characterisitic parameter, can finally determine sewage source parameter, and this is also the important prerequisite of modeling.

From above technical scheme, the present invention carries out comprehensive assessment by fuzzy evaluation algorithm to sewage source enterprise, The comprehensive of selecting index, independence and operability principle are taken into full account, it is former that evaluation result meets subjective and objective concordance Then, evaluation result is to final water environment treatment and improves construction with certain guidance meaning.

The present invention is not only simplified index compared to traditional method, and causes the weight point of each factor In the case of fitting over using a large amount of historical datas, more rationally, its evaluation result discrimination substantially, as a result more conforms to reality, Make the reflected appraisal result that evaluation index is more effective.

Description of the drawings

Fig. 1 is method of the present invention flow chart.

Specific embodiment

Below in conjunction with the accompanying drawings a kind of preferred implementation of 1 couple of present invention is described in detail.

Needed to be modeled using Component Based before assessment, wherein sewage source modeling approach：

A) in network-wide basis, each class industry user to being divided chooses some more representational users and adjusts Look into, determine that it uses the Capacity Ratio of sewage source constitution state and all kinds of sewage；

B) according to the industry overall characteristic that each industry user is determined per the average characteristics of class sewage source；

C) the industry composition and its capacity ratio of sewage source are determined, the composite wastewater needed for going out carrys out source model.

Below so that certain environmental administration of city carries out the assessment of water pollution source to the sewage enterprise of city as an example, illustrate.

Step one：Obtain sewage enterprise history samples data

By carrying out investigation statisticses to the sewage enterprise of city, the emission performance data of sewage enterprise are obtained.Jing investigation statisticses Ranking, preferably going out can represent the typical user of blowdown.9 users of primary election first, as shown in table 1.

The sewage business survey table of table 1

As can be seen from Table 1：9 enterprises of object to be assessed, the parameter of each enterprise={ solid pollutant, aerobic dirt Dye thing, nutrient pollutant, soda acid pollutant, toxic pollutant, oil pollutant, biological pollutant }, numerical value represents sewage Source proportion.User's sewage source statistical result is shown in Table 2

The sewage source volume percent statistical table of table 2

Step 2：Carry out fuzzy evaluation cluster

Water utilities, economic development expert are combined while environmental administration investigates to user's overall merit.Set of factors X=is set { economic, employment, prediction, environment }, evaluate collection Y={ important, more important, less important, inessential }.Expert to user it is each because Element is evaluated, the weight distribution of each factor be A=(0.3,0.1,0.5,0.1).Such as the shadow of 1 pair of prediction of user enterprise Ring, 40% expert thinks that the user is important, and 35% expert thinks important, and 25% expert is thought than less important. The overall merit of enterprise is as shown in table 3.

The Enterprise Integrated evaluation table of table 3

Single factor evaluation matrix is set up, under fuzzy operation overall merit is obtained：

Matrix is：

As a result show that user's significance level is 0.4, important degree is 0.35, be 0.25 than less important degree, Inessential degree is 0.1, by the visible enterprise-essential of maximum subjection principle, can be alternative as typical user.Obtain it in the same manner The synthetic evaluation matrix of his 8 enterprises is：

B₂=(0.25,0.3,0.35,0.1), B₃=(0.2,0.2,0.4,0.2),

B₄=(0.5,0.3,0.1,0.1), B₅=(0.25,0.25,0.35,0.15),

B₆=(0.45,0.1,0.15,0.3), B₇=(0.35,0.3,0.25,0.1),

B₈=(0.35,0.25,0.3,0.1), B₉=(0.3,0.35,0.2,0.1)

As can be seen here enterprise 1, enterprise 4, enterprise 6, enterprise 7, enterprise 8, enterprise 9 can be alternative as typical user.

Recycle fuzzy clustering to classify { enterprise 1, enterprise 4, enterprise 6, enterprise 7, enterprise 8, enterprise 9 }, screen allusion quotation Type user, concrete grammar is：

The sewage source percentage ratio of 6 primary election enterprises is set up with the survey data of gained, so as to set up sewage source enterprise Domain U '

Normalization process adopts equation below：

Normalization process is carried out to data, normalized matrix is obtained：

Set up fuzzy similarity matrix：

If taking cut set λ=0.256, enterprise i, j of rij >=0.256 are one for a class, i.e. enterprise's { Isosorbide-5-Nitrae, 6,7,8,9 } Big class is now actual not classified；If taking cut set λ=0.421, r_ij>=0.421 enterprise i, j is a class, is now looked forward to Industry { Isosorbide-5-Nitrae, 6,7,8 } is a class, and { 9 } are a class, then can be used as typical user's group comprising the class more than element.

Step 3：Evaluated using prior art FCM method

1) user's Subject Matrix D

According to aforementioned thought with FCM methods to typical user's cluster analyses, you can determine Subject Matrix D and cluster centre P. It is determined that classification number c=2, m=2, by formulaFormulaIt is calculated typical user person in servitude Category matrix D：

In formula：T belongs to the generic degree of cluster point for equipment.

2) center matrix P

Matrix D is a fuzzy classified matrix, represents that a typical user belongs to the degree of membership of a certain class per a line, each Row are represented in an apoplexy due to endogenous wind without such degree of membership.According to this classification matrix, according to the maximum subjection principle in fuzzy set The belonging kinds of each typical user are can determine that, enterprise's { Isosorbide-5-Nitrae, 5,6,7,8 } is a class, and { 2,3,9 } are a class.According to including The most principle of element, enterprise's { Isosorbide-5-Nitrae, 5,6,7,8 } center matrix of cluster is represented as selected typical user's matroid P, A cluster centre is represented per a line, each row are represented in this apoplexy due to endogenous wind sewage source proportion, gained center matrix P：

Step 4：Cluster result is analyzed

User to be clustered is extracted Typical Representative as industry by FCM methods by membership function, used as Component Based The characteristics of basis of sewage source modeling has description user's general character.From the center matrix of FCM methods, the main dirt of the first kind Water source is solid pollutant and biological pollutant, and it is solid pollutant and oil pollutant that Equations of The Second Kind is main.From cluster knot Guo Kan enterprises 5 can be to regard typical user as, but, the algorithm depends on initial cluster center necessary in advance while selection It is determined that the number of cluster, for parameter m, is also called smoothing factor, control to share degree between fuzzy class, to realize fuzzy poly- Class must just select suitable m, be the special case of m=2 in this algorithm, and optimal m values still lack at present theoretical direction.Thus may be used See, the quality of the method strong depend-ence initialization data, thus have very strong randomness, reduce the accuracy of algorithm.

Cluster analyses of the present invention based on fuzzy overall evaluation are by considering sewage on the basis of expert opinion is solicited Impact and enterprise of the source analysis to prediction selects typical user to the comprehensive effect of society.First, by fuzzy synthesis Evaluating first selecting can be used as the candidate enterprise of typical user.Secondly, by the cluster analyses based on equivalence relation, choose suitable Cut set, cluster result is divided into two classes, comprising the most class of element be selected typical user because such reflects industry The general character of Most users, i.e., main sewage source is solid pollutant and biological pollutant, while the weight in overall evaluation system The property wanted degree is higher.

Not only taken on a new look stiff algorithm based on the cluster of fuzzy evaluation, and combined the Rational Decision of expert, moreover it is possible to Preferably clustered, therefore, when the cluster analyses of sewage source modeling are studied, the overall merit based on fuzzy analysis is clustered Method has promotional value.

The above embodiment is only that the preferred embodiment of the present invention is described, not to the model of the present invention Enclose and be defined, on the premise of without departing from design spirit of the present invention, technical side of the those of ordinary skill in the art to the present invention Various modifications and improvement that case is made, all should fall in the protection domain of claims of the present invention determination.

Claims (4)

1. a kind of sewage enterprise water pollution source appraisal procedure based on multiple index evaluation algorithm, it is characterised in that including as follows Step：

1) sewage enterprise water pollution source factors system is set up, all specific targets constituent elements in water pollution source factors system Collection X={ x₁,x₂,…,x_m, wherein m represents the number of factor；

2) sewage enterprise water pollution source appraisement system is set up, all specific targets are constituted and evaluated in the appraisement system of water pollution source Collection Y={ y₁,y₂,…,y_n, wherein n represents the number of evaluation；

3) fuzzy evaluation is carried out to sewage enterprise water pollution derived data using the method for blurring mapping, obtains single factor evaluation square Battle array：

$R=\left\{\begin{array}{c}{R}_1\\ R_2\\ .\\ .\\ .\\ R_m\end{array}\right\}=\left\{\begin{array}{cccc}{r}_{11}& r_{12}& \mathrm{...}& r_{1n}\\ r_{21}& r_{22}& \mathrm{...}& r_{2n}\\ & & .& \\ & & .& \\ & & .& \\ r_{m1}& r_{m2}& \mathrm{...}& r_{mn}\end{array}\right\};$

4) primary election is carried out to data using fuzzy evaluating matrix：

Define factor weight distribution A=(a₁,a₂,…,a_m), wherein a_i>=0 andObtaining fuzzy evaluating matrix is

Wherein, each element in fuzzy evaluating matrix B is chosen according to maximum subjection principle：b_i=max { min (a_i,r_1i),…,min (a_m,r_mi)}；Typical subject is selected according to evaluate collection Y, primary election object set U={ u are obtained₁,u₂,…,u_p}；

5) primary election object set is classified using method of fuzzy cluster analysis, screens typical subject.

2. sewage enterprise according to claim 1 water pollution source appraisal procedure, it is characterised in that step 1) in, sewage Enterprise's water pollution source factors include economic, employment, prediction and environment；Step 2) in, bag is evaluated in sewage enterprise water pollution source Include important, more important, less important and inessential.

3. sewage enterprise according to claim 1 water pollution source appraisal procedure, it is characterised in that step 5) in, obscure Clustering methodology specifically includes following steps：

51) using the sewage source percentage ratio of primary election object set, the domain of sewage source enterprise is set up：

$U^{\′}=\left\{\begin{array}{cccc}{u}_{11}& u_{12}& \mathrm{...}& u_{1k}\\ u_{21}& u_{22}& \mathrm{...}& u_{2k}\\ & & .& \\ & & .& \\ & & .& \\ u_{p1}& u_{p2}& \mathrm{...}& u_{pk}\end{array}\right\};$

Wherein k is the number of pollutant kind, and p is the number of primary election object；

52) normalization process is carried out to data, obtains normalized matrix：

$U^{\′\′}=\left\{\begin{array}{cccc}{u}_{11}^{\′}& u_{12}^{\′}& \mathrm{...}& u_{1k}^{\′}\\ u_{21}^{\′}& u_{22}^{\′}& \mathrm{...}& u_{2k}^{\′}\\ & & .& \\ & & .& \\ & & .& \\ u_{p1}^{\′}& u_{p2}^{\′}& \mathrm{...}& u_{pk}^{\′}\end{array}\right\};$

53) fuzzy similarity matrix is set up：

$z_{ij}=\frac{\underset{k=1}{\overset{m}{\Σ}}min(u_{ik},u_{ik})}{\underset{k=1}{\overset{m}{\Σ}}max(u_{ik},u_{ik})}$

$Z={\left(z_{ij}\right)}_{p\×p}=\left\{\begin{array}{cccc}{z}_{11}& z_{12}& \mathrm{...}& z_{1p}\\ z_{21}& z_{22}& \mathrm{...}& z_{2n}\\ & & .& \\ & & .& \\ & & .& \\ z_{p1}& z_{p2}& \mathrm{...}& z_{pp}\end{array}\right\};$

54) suitable cut set λ is chosen, primary election object is classified, obtain preferred typical subject.

4. sewage enterprise according to claim 3 water pollution source appraisal procedure, it is characterised in that step 52) in, it is regular Change is processed and adopts equation below：

$u_{ik}=\frac{u_{ik}^{\′}-u_{kmin}^{\′}}{u_{kmax}^{\′}-u_{k\min }^{\′}};$

Wherein u '_kmax、u′_kminIt is u '_1k, u '_2k..., u '_pkIn maximum and minima.