CN105160496A - Comprehensive evaluation method of enterprise electricity energy efficiency - Google Patents

Abstract

The present invention discloses a comprehensive evaluation method of enterprise electricity energy efficiency, comprising the following steps: screening indexes through a characteristic value method and establishing an evaluation grade index model of electricity energy efficiency; in combination with expert opinions and actual situations of enterprises, using an analytic hierarchy process for calculating weight values of all indexes in the index model; using a fuzzy evaluation method for calculating and obtaining membership degrees and scores of all the indexes; and performing systematic and scientific evaluation on the enterprise electricity energy efficiency according to the scores. The method not only considers the actual situations of the enterprises, but also considers data acquisition, calculation and other actual problems, selects the indexes having both integral comprehensiveness and individual independence, calculates the weight values through combining the expert opinions with the actual situations of the enterprises, has feasibility, comparability and small calculation amount, and simultaneously effectively overcomes shortcomings of a subjective weighting method.

Description

A kind of enterprise electric energy efficiency comprehensive estimation method

Technical field

The present invention relates to electric energy efficiency management domain, particularly a kind of enterprise electric energy efficiency comprehensive estimation method.

Background technology

Current, for enterprise, especially high energy consumption enterprise, the realization of energy-saving and cost-reducing target is the needs of CSR and development.This is because industry is the rich and influential family of China's energy-consuming, and the high energy consumption enterprise in emphasis power consumption industry is again the rich and influential family of industrial energy consumption.And large for medium and small sized enterprises' quantity, the industry of covering is wide, and due to scale relatively little, so technical merit and energy consumption level uneven, also need to save energy and reduce the cost, improve the utilization ratio of the energy, to improve the competitive power of enterprise self.

At present, in China, the assessment of electric system saving energy and decreasing loss does not also form formal testing standard.The Evaluation on Energy Saving of electric system electrical equipment in the past and efficiency test, be mostly type approval test or certification test project, relevant criterion is the method for testing under laboratory condition mostly, and is the test under power frequency condition and evaluation.At present, the basic skills of energy efficiency evaluation is analytical hierarchy process, and the method mainly makes full use of the knowledge and experience of expert, assesses the energy consumption management system of high energy consumption enterprise.

Patent of invention (application number: 201210153005.9) disclose a kind of industrial user's electric energy efficiency appraisal procedure, it is applicable to the three-class power energy efficiency evaluation index of industrial user, then the weight coefficient of one-level, secondary in analytical hierarchy process determination energy efficiency indexes is used, the expert be applicable to is selected to carry out weight calculation, although the method can give full play to the experimental knowledge of expert, but result of calculation depends on the preference of expert to a great extent, it is a kind of subjective weighting method.

Patent of invention (application number: the electric energy efficiency Hopfield neural network appraisal procedure and the system that 201310146941.1) disclose a kind of enterprise customer, this system accepts all data that electric supervising device gathers, after analyzing, obtain each factor value of energy efficiency indexes system, Hopfield neural network is adopted to assess each factor group factor, the method is by setting up neural network model, can the efficiency influence factor of objective evaluation enterprise effectively, but calculating data, the process setting up Hopfield model is very complicated, each enterprise is due to its equipment, situation is different, Hopfield model needs constantly to re-establish, calculated amount is huge.

Patent of invention (application number: 201310610968.1) disclose a kind of the electric energy efficiency monitoring and evaluation system and the evaluation method thereof that are applied to enterprise, this system can carry out Real-Time Monitoring to the electric energy efficiency of enterprise, can the electric energy efficiency of objective evaluation enterprise, the method take into account the problem of data acquisition, but do not consider the actual conditions of enterprise, lack individual independence.

Summary of the invention

The object of the invention is to solve the deficiencies in the prior art, a kind of enterprise electric energy efficiency comprehensive estimation method be provided, the method can be carried out comprehensively enterprise's electric energy efficiency in conjunction with expert opinion and enterprise practical situation, the scientific evaluation of system.

For achieving the above object, the technical solution used in the present invention is: a kind of enterprise electric energy efficiency comprehensive estimation method, comprises the steps:

A) electric energy efficiency opinion rating index model is set up;

B) in conjunction with expert opinion and enterprise practical situation, analytical hierarchy process is adopted to calculate the weighted value of each index in described index model;

C) fuzzy assessment method is utilized to calculate degree of membership and the score of trying to achieve each index;

D) according to score, system, scientific evaluation are carried out to enterprise's electric energy efficiency.

Preferably, step a) in, described index model comprises economic energy efficiency indexes, power quality index, production management index, equipment control index, electric energy contamination index.

Further preferably, step a) in have employed method of characteristic to obtain the indices in described index model.

Preferably, step b) concrete implementation step as follows:

B1) judgment matrix A is between two constructed: comparing between two between ad eundem index;

$A=a_{ij(n\×n)}=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{ln}\\ a_{21}& a_{22}& ...& a_{2n}\\ ...& ...& ...& ...\\ a_{n1}& a_{n2}& ...& a_{nn}\end{array}\right]$

A _ijrepresent the important scale assignment between index i and index j, a _ji=1/a _ij, as i=j, a _ij=1;

B2) weight is calculated: the weight calculating each index based on product root method;

B3) selective goal weight threshold, screens index.

Further preferably, step b1) in, important scale is the significance level of each index in described index model, is divided into 9 scale 1-9, and 1 is of equal importance, and 9 is extremely important.

Further preferably, the proper vector w in judgment matrix corresponding to eigenvalue of maximum 1) is calculated ⁱ;

$\begin{array}{cccc}{b}_i={\left(\underset{j=1}{\overset{m}{\Σ}}{a}_{ij}\right)}^{\frac{1}{m}}& i=1,2,\mathrm{...},m& w_i=b_j/\underset{k=1}{\overset{m}{\Σ}}{b}_k& j=1,2,\mathrm{...},m\end{array}$

Bi represents the product root of the weight scale about index i, and m represents index quantity;

2) eigenvalue of maximum is calculated ${\mathrm{\λ}}_{max}=\frac{1}{m}\underset{i=1}{\overset{m}{\Σ}}\frac{\underset{j=1}{\overset{m}{\Σ}}{a}_{ij}{w}_j}{w_i};$

3) the coincident indicator C.I. of judgment matrix A is calculated,

4) random Consistency Ratio C.R. is adopted to carry out consistency check to judgment matrix A,

$C.R.=\frac{C.I.}{R.I.}$

R.I. be same order mean random mark, as C.R.<0.1, judgment matrix meets consistance, then its proper vector w _ibe the electric energy efficiency index weights of normalizing; If during C.R.>=0.1, then redefine judgment matrix A.

Further preferably, step b3) in, according to the actual conditions of described index model, selective goal weight threshold: when index number is between 7 ~ 9, index for selection weight threshold is the index of 0.07; When index number is between 4 ~ 6, the index of index for selection weight threshold 0.1; When index number≤3, do not screen.

Preferably, step b3) in, according to the actual conditions of described index model, selective goal weight threshold: when index number is between 7 ~ 9, index for selection weight threshold is the index of 0.07; When index number is between 4 ~ 6, the index of index for selection weight threshold 0.1; When index number≤3, do not screen.

Preferably, step c) comprise the steps:

C1) according to affecting each factor character of electric energy efficiency by index layering, factor layer, content layer, destination layer is respectively;

C2) the single factor test A in factor layer is calculated _k,jscore F _k,j, determine evaluation effect, and determine its weights omega _k,j, meet ω _k,j>0, and

C3) every content A in Computed-torque control layer _k(k=1,2 ..., evaluation score m) p _kfor monofactorial number, determine the evaluation effect of each content according to score;

C4) every content A is determined _kweights omega _k, ω _k> 0;

C5) the comprehensive evaluation score F of destination layer is calculated,

Further preferably, each monofactorial domain correspondence three is supposed individual fuzzy subset: good, in, poor }={ E1, E2, E3}, the membership function of its correspondence is μ ₁, μ ₂, μ ₃, adopt Triangleshape grade of membership function model, calculate monofactorial score F _k,j,

$F_{k,j}=\frac{{\mathrm{\&mu;}}_1\left(x\right)\&times;F_1+{\mathrm{\&mu;}}_2\left(x\right)\&times;F_2+{\mathrm{\&mu;}}_3\left(x\right)\&times;F_3}{\underset{i=1}{\overset{3}{\&Sigma;}}{\mathrm{\&mu;}}_i\left(x\right)}$

Wherein, F ₁, F ₂, F ₃be respectively index to belong to completely, in, mark during difference, gets 98,72,45 respectively.Due to the utilization of technique scheme, the present invention compared with prior art has following advantages: enterprise of the present invention electric energy efficiency comprehensive estimation method, do not need to pass through Holistic modeling, only need set up the model of fuzzy synthetic evaluation that enterprise's electric energy efficiency is evaluated, then set up different degree of membership Confirming model according to different pointer types, the object of comprehensive evaluation can be reached.Simply, comprehensively, efficiently, overcoming original certain methods adopts unicity to analyze the shortcoming of certain type enterprise electric energy efficiency, feature based value method is screened index, subjectiveness and objectiveness combines and analyzes, there is comparability and feasibility, not only consider actual conditions, have also contemplated that the practical problems such as data acquisition and calculated amount, do not repeat not redundancy, again there is independence comprehensively.The method of application expert and objective combination, each index is carried out to the judgement of weight, end product presents with fractional form, direct feel, and index at different levels reciprocal fraction to some extent, provide good foundation for carrying out Saving energy to enterprise.

Accompanying drawing explanation

Accompanying drawing 1 is enterprise of the present invention electric energy efficiency comprehensive estimation method block diagram.

Embodiment

Below in conjunction with specific embodiment, technical scheme of the present invention is further elaborated.

A kind of enterprise electric energy efficiency comprehensive estimation method, shown in Figure 1, specific as follows:

Step 101: set up electric energy efficiency grade evaluation index model;

Determine evaluation index, screened by method of characteristic, eliminate the High redundancy between index, consider impact direct or indirect between index, reject secondary index, form simple and clear reasonably electric energy efficiency Grade, this index model comprises economic energy efficiency indexes, power quality index, production management index, equipment control index and electric energy contamination index.

When determining evaluation index, not only needing the actual conditions comprehensively considering whole enterprise, also needing to consider the practical problems such as data acquisition, calculating simultaneously, guarantee that the index chosen neither omits key factor, can not repeated and redundant, existing entirety comprehensive, has again individual independence.

And when adopting method of characteristic to screen, final target is to set up index model with minimum topmost index.

Step 102: in conjunction with expert opinion and enterprise practical situation, adopts the weighted value of each index in analytical hierarchy process parameter model, concrete:

2-1) to the index development of judgment matrix A between two of ad eundem:

$A=\left[\begin{array}{cccc}\frac{w_1}{w_1}& \frac{w_1}{w_2}& \mathrm{...}& \frac{w_1}{w_n}\\ \frac{w_2}{w_1}& \frac{w_2}{w_2}& \mathrm{...}& \frac{w_2}{w_n}\\ \mathrm{...}& \mathrm{...}& \mathrm{...}& \mathrm{...}\\ \frac{w_n}{w_1}& \frac{w_n}{w_2}& \mathrm{...}& \frac{w_n}{w_n}\end{array}\right]=A\left(a_{ij}\right),i,j=1,2,\mathrm{...},m$

Wherein, a _ijrepresent in evaluation indice X, index x _irelative to x _jsignificance level; Then according to an expert view, adopt the 1-9 scale that Satty proposes, as shown in the table.

Judgment matrix is built as follows:

$A=a_{ij(n\&times;n)}=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{ln}\\ a_{21}& a_{22}& ...& a_{2n}\\ ...& ...& ...& ...\\ a_{n1}& a_{n2}& ...& a_{nn}\end{array}\right]$

2-2) utilize the proper vector w corresponding to product root method compute matrix eigenvalue of maximum _i;

$\begin{array}{cccc}{b}_i={\left(\underset{j=1}{\overset{m}{\&Sigma;}}{a}_{ij}\right)}^{\frac{1}{m}}& i=1,2,\mathrm{...},m& w_i=b_j/\underset{k=1}{\overset{m}{\&Sigma;}}{b}_k& j=1,2,\mathrm{...},m\end{array}$

Bi represents the product root of the weight scale about index i, and m represents index quantity;

2-3) calculate eigenvalue of maximum according to the proper vector obtained

2-4) calculate the coincident indicator C.I. of judgment matrix A.

$C.I.=\frac{{\mathrm{\&lambda;}}_{max}-m}{m-1}$

2-5) by random Consistency Ratio C.R., consistency check is carried out to judgment matrix,

$C.R.=\frac{C.I.}{R.I.}$

R.I. be same order mean random mark.As C.R.<0.1, judgment matrix meets consistance, then its proper vector is the index weights of normalizing;

2-6) according to the actual conditions of index system, selective goal weight threshold, its selection principle is: when index number is between 7 ~ 9, and index for selection weight threshold is the index of 0.07; When index number is between 4 ~ 6, the index of index for selection weight threshold 0.1; When index number≤3, do not screen.

Step 103: the actual conditions of foundation enterprise and relevant national standard carry out the data of investigation gained to determine subordinate function model and parameter to enterprise.

One, the concrete steps evaluating score are as follows:

3-1) numerous influence factors of enterprise to be evaluated are divided into many levels by its character, are generally three layers, be respectively destination layer, content layer, factor layer, here, the index screened through step 102 is in factor layer;

3-2) calculate the every single factor test A of factor layer _k,jscore F _k,j, determine evaluation effect;

3-3) for every content A _k(k=1,2 ..., m), determine the weights omega of its each factor comprised _k,j, ω be met _k,j>0, and

3-4) each content A in Computed-torque control layer _kscore F _k, p _kfor the number of two-level index, according to score determination content A _kevaluation effect;

3-5) determine the weight of each content, evaluation objective is the weights omega of every content in A _k> 0;

3-6) calculate comprehensive evaluation score F, must being divided into of evaluation objective

Two, the calculating of single factor test degree of membership:

Here, suppose that the domain of each single factor test index is corresponding three fuzzy subsets: good, in, poor }={ E1, E2, E3}, the membership function of its correspondence is μ ₁, μ ₂, μ ₃.Factor is different, and the membership function model of fuzzy subset is not identical yet.

In this example, adopt Triangleshape grade of membership function model, concrete membership function model is as follows:

(1) model 1: be applicable to the factor that value is the smaller the better, formula (1-1) ~ (1-3) be shown in by membership function model.

${\mathrm{\&mu;}}_1\left(x\right)=\left\{\begin{array}{cc}1,& x\&le;a_1\\ (x-a_2)/(a_1-a_2),& a_1<x\&le;a_2\\ 0,& x>a_2\end{array}\right.$ Formula (1-1)

${\mathrm{\&mu;}}_2\left(x\right)=\{\begin{array}{cc}0,& x\&le;a_{1}orx\&GreaterEqual;a_3\\ (x-a_1)/(a_2-a_1),& a_1<x\&le;a_2\\ (x-a_3)/(a_2-a_3),& a_2<x<a_3\end{array},a_1<a_2<a_3$ Formula (1-2)

${\mathrm{\&mu;}}_3\left(x\right)=\left\{\begin{array}{cc}0,& x\&le;a_2\\ (x-a_2)/(a_3-a_2),& a_2<x\&le;a_3\\ 1,& x>a_3\end{array}\right.$ Formula (1-3)

Model 2: be applicable to be worth the factor be the bigger the better, formula (1-4) ~ (1-6) be shown in by subordinate function model.

${\mathrm{\&mu;}}_1\left(x\right)=\left\{\begin{array}{cc}0,& x\&le;a_2\\ (x-a_2)/(a_1-a_2),& a_2<x\&le;a_1\\ 1,& x>a_1\end{array}\right.$ Formula (1-4)

${\mathrm{\&mu;}}_2\left(x\right)=\{\begin{array}{cc}0,& x\&le;a_{3}orx\&GreaterEqual;a_1\\ (x-a_3)/(a_2-a_3),& a_3<x\&le;a_2\\ (x-a_1)/(a_2-a_1),& a_2<x<a_1\end{array},a_3<a_2<a_1$ Formula (1-5)

${\mathrm{\&mu;}}_3\left(x\right)=\left\{\begin{array}{cc}1,& x\&le;a_3\\ (x-a_2)/(a_3-a_2),& a_3<x\&le;a_3\\ 0,& x>a_2\end{array}\right.$ Formula (1-6)

Model 3: be applicable to the factor of value in certain fixed interval.Formula (1-7) ~ (1-9) be shown in by subordinate function model. ${\mathrm{\&mu;}}_1\left(x\right)=\left\{\begin{array}{cc}1,& a_{11}<x\&le;a_{12}\\ (x-a_{11})/(a_{11}-a_{21}),& a_{21}<x\&le;a_{11}\\ (x-a_{22})/(a_{12}-a_{22}),& a_{12}<x\&le;a_{22}\\ 0,& x\&le;a_{21}orx>a_{22}\end{array}\right.$ Formula (1-7)

${\mathrm{\&mu;}}_2\left(x\right)=\left\{\begin{array}{cc}0,& a_{31}<x<a_{11}\\ (x-a_{31})/(a_{11}-a_{31}),& a_{21}\&le;x<a_{11}\\ (x-a_{11})/(a_{21}-a_{11}),& a_{21}<x\&le;a_{11}\\ (x-a_{12})/(a_{22}-a_{12}),& a_{12}<x\&le;a_{22}\\ (x-a_{32})/(a_{22}-a_{32}),& a_{22}<x<a_{32}\end{array}\right.,a_{31}<a_{21}<a_{11}<a_{12}<a_{22}<a_{32}$ Formula (1-8)

${\mathrm{\&mu;}}_3\left(x\right)=\left\{\begin{array}{cc}1,& x\&le;a_{31}orx\&GreaterEqual;a_{32}\\ (x-a_{21})/(a_{31}-a_{21}),& a_{31}<x\&le;a_{21}\\ (x-a_{22})/(a_{32}-a_{22}),& a_{22}<x\&le;a_{32}\\ 0,& a_{21}\&le;x\&le;a_{22}\end{array}\right.$ Formula (1-9)

A ₁, a ₂, a ₃represent definition index good, in, difference fuzzy membership parameter.

μ ₁, μ ₂, μ ₃represent that this single factor test belongs to, in, the degree of membership of difference, and μ ₁, μ ₂, μ ₃meet following relation: μ ₁+ μ ₂+ μ ₃=1, determine that the step of single index Fuzzy evaluation model is as follows:

Determine to select which kind of model according to index value, subordinate function parameter should be determined in conjunction with actual conditions.

According to each single factor test A _k,jsubordinate function model and parameter determines that each index belongs to, in, the degree of membership μ of difference ₁, μ ₂, μ ₃, then calculate each factor evaluation score F _k,j, according to score determination opinion rating.Each single factor test index score formula:

$F_{k,j}=\frac{{\mathrm{\&mu;}}_1\left(x\right)\&times;F_1+{\mathrm{\&mu;}}_2\left(x\right)\&times;F_2+{\mathrm{\&mu;}}_3\left(x\right)\&times;F_3}{\underset{i=1}{\overset{3}{\&Sigma;}}{\mathrm{\&mu;}}_i\left(x\right)}$

Wherein, F ₁, F ₂, F ₃be respectively index to belong to completely, in, mark during difference, gets 98,72,45 respectively.

Then, according to index score, enterprise's electric energy efficiency is carried out to the evaluation of system, science.

Above-described embodiment is only for illustrating technical conceive of the present invention and feature; its object is to person skilled in the art can be understood content of the present invention and be implemented; can not limit the scope of the invention with this; all equivalences done according to Spirit Essence of the present invention change or modify, and all should be encompassed in protection scope of the present invention.

Claims (9)

1. enterprise's electric energy efficiency comprehensive estimation method, is characterized in that, comprises the steps:

A) electric energy efficiency opinion rating index model is set up;

B) in conjunction with expert opinion and enterprise practical situation, analytical hierarchy process is adopted to calculate the weighted value of each index in described index model;

C) fuzzy assessment method is utilized to calculate degree of membership and the score of trying to achieve each index;

D) according to score, system, scientific evaluation are carried out to enterprise's electric energy efficiency.

2. enterprise according to claim 1 electric energy efficiency comprehensive estimation method, it is characterized in that, step a) in, described index model comprises economic energy efficiency indexes, power quality index, production management index, equipment control index, electric energy contamination index.

3. enterprise according to claim 2 electric energy efficiency comprehensive estimation method, is characterized in that, step a) in have employed method of characteristic to obtain the indices in described index model.

4. enterprise according to claim 1 electric energy efficiency comprehensive estimation method, is characterized in that, step b) concrete implementation step as follows:

B1) judgment matrix A is between two constructed: comparing between two between ad eundem index;

$A=a_{ij(n\&times;n)}=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ a_{21}& a_{22}& ...& a_{2n}\\ ...& ...& ...& ...\\ a_{n1}& a_{n2}& ...& a_{nn}\end{array}\right]$

A _ijrepresent the important scale assignment between index i and index j, a _ji=1/a _ij, as i=j, a _ij=1;

B2) weight is calculated: the weight calculating each index based on product root method;

B3) selective goal weight threshold, screens index.

5. enterprise according to claim 4 electric energy efficiency comprehensive estimation method, is characterized in that, step b1) in, important scale is the significance level of each index in described index model, be divided into 9 scale 1-9, and 1 is of equal importance, 9 is extremely important.

6. enterprise according to claim 4 electric energy efficiency comprehensive estimation method, is characterized in that, step b2) in, calculate weight based on product root method, specific as follows:

1) the proper vector w in judgment matrix corresponding to eigenvalue of maximum is calculated ⁱ;

$\begin{array}{cc}{b}_i={\left(\underset{j=1}{\overset{m}{\&Pi;}}{a}_{ij}\right)}^{\frac{1}{m}},j=1,2,...,m& w_i=b_j/\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{b}_k,j=1,2,...,m\end{array}$

Bi represents the product root of the weight scale about index i, and m represents index quantity;

2) eigenvalue of maximum is calculated ${\mathrm{\&lambda;}}_{max}=\frac{1}{m}\underset{i=1}{\overset{m}{\&Sigma;}}\frac{\underset{j=1}{\overset{m}{\&Sigma;}}{a}_{ij}{w}_j}{w_i};$

3) the coincident indicator C.I. of judgment matrix A is calculated,

4) random Consistency Ratio C.R. is adopted to carry out consistency check to judgment matrix A,

$C.R.=\frac{C.I.}{R.I.}$

R.I. be same order mean random mark, as C.R.<0.1, judgment matrix meets consistance, then its proper vector w _ibe the electric energy efficiency index weights of normalizing; If during C.R.>=0.1, then redefine judgment matrix A.

7. enterprise according to claim 4 electric energy efficiency comprehensive estimation method, it is characterized in that, step b3) in, according to the actual conditions of described index model, selective goal weight threshold: when index number is between 7 ~ 9, index for selection weight threshold is the index of 0.07; When index number is between 4 ~ 6, the index of index for selection weight threshold 0.1; When index number≤3, do not screen.

8. enterprise according to claim 1 electric energy efficiency comprehensive estimation method, is characterized in that, step c) comprise the steps:

C1) according to affecting each factor character of electric energy efficiency by index layering, factor layer, content layer, destination layer is respectively;

C2) the single factor test A in factor layer is calculated _k,jscore F _k,j, determine evaluation effect, and determine its weights omega _k,j, meet ω _k,j>0, and

C3) every content A in Computed-torque control layer _k(k=1,2 ..., evaluation score m) p _kfor monofactorial number, determine the evaluation effect of each content according to score;

C4) every content A is determined _kweights omega _k, ω _k> 0;

C5) the comprehensive evaluation score F of destination layer is calculated,

9. enterprise according to claim 8 electric energy efficiency comprehensive estimation method, is characterized in that: suppose each monofactorial domain correspondence three individual fuzzy subset: good, in, poor }={ E1, E2, E3}, the membership function of its correspondence is μ ₁, μ ₂, μ ₃, adopt Triangleshape grade of membership function model, calculate monofactorial score F _k,j,

$F_{k,j}=\frac{{\mathrm{\&mu;}}_1\left(x\right)\&times;F_1+{\mathrm{\&mu;}}_2\left(x\right)\&times;F_2+{\mathrm{\&mu;}}_3\left(x\right)\&times;F_3}{\underset{i=1}{\overset{3}{\&Sigma;}}{\mathrm{\&mu;}}_i\left(x\right)}$

Wherein, F ₁, F ₂, F ₃be respectively index to belong to completely, in, mark during difference, gets 98,72,45 respectively.