CN104182816A - Method for evaluating power quality comprehensively based on the Vague sets and the improved technique for order preference by similarity to ideal solution and application thereof - Google Patents

Abstract

The invention discloses a method for evaluating power quality comprehensively based on the Vague sets and the improved technique for order preference by similarity to ideal solution and the application thereof. The method consists of the steps of: monitoring the common connection points of an offshore wind power plant, and acquiring the simulation data of power quality indicators; determining the weight vectors (i)W(/i) of the power quality indicators of each evaluation object; indicating the scores of the indicators of each evaluation object with Vague values; determining an absolute positive ideal solution and an absolute negative ideal solution; calculating the distance between each evaluation object and the absolute positive ideal solution and the distance between each evaluation object and the absolute negative ideal solution and the relative closeness of each evaluation object, and comprehensively evaluating and sequencing the power quality according to the calculated results. By means of this invention, the weights are determined more objectively and less subjectively, and in the meantime, the weight of indicators and the weights of the true membership function, the false membership function and the abstention part of the indicators are considered, so that the evaluation result can be closer to the actual situation. The method plays a very significant role in enhancing the transparency of the power market and improving the power supply quality of the power system.

Description

The energy quality comprehensive assessment method and the application thereof that approach ideal solution based on Vague collection and improvement

Technical field

The present invention relates to a kind of based on Vague collection and improve approach ideal solution energy quality comprehensive assessment method and application thereof, belong to electrical engineering and electric energy quality synthesis evaluation field.

Background technology

Wind energy has obtained the whole world as environmental protection, reproducible resource and has extensively paid attention to.In recent years, offshore wind farm development rapidly, has that wind speed is large, generated energy large and disturb the advantages such as few.But the undulatory property of wind speed, intermittence and randomness, wake flow, turbulent flow, blower fan tower shadow effect etc. can cause power swing, and then affect the quality of power supply, and add that mostly Large Scale Offshore Wind Farm is to access compared with weak link place at power distribution network, can electric power networks be impacted and be disturbed.Under Power Market, electric energy is fixed the price according to the quality, high quality and favourable price and evaluation point examination rewards and punishments requirement people pay close attention to power quality problem, is necessary therefore the power quality index of marine wind electric field is assessed.

The difficult point of electric energy quality synthesis evaluation is: how to set up rational power quality index system; How objectively reasonably determine the each index weights of the quality of power supply; How to find more perfect, comprehensive, quantitative, objective assessment method to be applied in the assessment of Modern Electric Power Quality, to obtain more accurately reasonably assessment result.Existing document has been made useful exploration to the comprehensive assessment of the quality of power supply, many energy quality comprehensive assessment methods are proposed: analytical hierarchy process (Analytic Hierarchy Process, AHP), fuzzy theory, probability statistics and vector algebra method, neural network, information entropy principle, sudden change Decision Method, core vector space model, D-S evidence theory, genetic algorithm etc.In determining power quality index weight, can be subject in various degree the impact of artificial subjective factor, subjective evaluation method is to study how farthest to reduce subjectivity, improves the method for the reliability of assessment result.But the uncertainty that a lot of othernesses of all not considering that expert is familiar with of the method for existing definite weight cause, makes the subjectivity of result larger.With respect to subjective appraisal procedure, objective evaluation method has the advantages that assessment result objectivity is stronger, a lot of intelligent methods require training sample data enough large, just can obtain accurate model, and then make the evaluation work of other data of existing model, but often data are insufficient in reality, so often cause assessment result and actual conditions to differ larger.Also some objective assessment method is not considered the feature of different evaluation object power quality indexs self, causes assessment result and actual conditions not to meet.Due to the many comprehensive estimation methods of above reason can apply in practice actually rare.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of energy quality comprehensive assessment method and application thereof that approaches ideal solution based on Vague collection and improvement is provided.

For technical solution problem, the technical solution used in the present invention is:

Comprise the steps: based on Vague collection and the energy quality comprehensive assessment method that improvement approaches ideal solution

(1) on PSCAD/EMTDC software platform, offshore grid-connected wind farm is simulated, obtain the emulated data of 6 power quality indexs at offshore grid-connected wind farm points of common connection place;

(2) judgment matrix that uses analytical hierarchy process in conjunction with the Pair Analysis Construction of A Model of Set Pair Analysis Theory understanding homogeneity matrix and cognitive diversity matrix, consistent matrix method is processed homogeneity matrix and is obtained corresponding weight value, objectively entropy power method is processed otherness matrix and is obtained corresponding weight, finally by Set Pair Analysis Theory, both combinations is obtained to power quality index weight vectors W;

(3) invite expert to evaluate the quality of evaluation object and each power quality index, adopt the scoring to each index of Vague value representation each evaluation object, definite method of the Vague value of each index is: in conjunction with the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of use Vague value representation;

(4) determine definitely positive ideal solution and definitely negative ideal solution, in the time only considering the weight of index, using the ideal value in GB and limit value as absolute ideal solution and definitely negative ideal solution, join together in the middle of evaluation object, when this ideal value refers to that working as 6 adopted power quality index values is tending towards 0, using numerical value 0 as ideal value;

(5) method that adopts improved weighting to approach ideal solution on the basis of Vague collection theory, calculates each evaluation object and the absolute just distance of ideal solution distance with definitely negative ideal solution the weight vectors W of the power quality index that institute's weighted value is each evaluation object;

(6) according to the distance of the definitely positive ideal solution having calculated distance with definitely negative ideal solution calculate the relative approach degree σ of each evaluation object _i,

${\mathrm{\σ}}_i=\frac{{\mathrm{\Γ}}_i^{*}}{{\mathrm{\Γ}}_i^{-}+{\mathrm{\Γ}}_i^{*}},i=\mathrm{1,2},...,n$

The relative approach degree σ of evaluation object _iless, represent that evaluation object approaches positive ideal solution and away from negative ideal solution, by approaching degree principle, the quality of power supply carried out to comprehensive assessment and sequence.

Described step (1) is:

Marine wind electric field emulation adopts the situation of two kinds of capacity: 50MW and 75MW, wherein, 50MW includes the double-fed wind power generator group of 10 unit 5MW, 75MW includes the double-fed wind power generator group of 15 unit 5MW, every typhoon group of motors is elevated to 35kV outlet voltage from 0.69kV through machine end step-up transformer, then pass through current collection network in wind energy turbine set, be connected to the low-pressure side of offshore boosting station, by offshore boosting station, the voltage that collects of 35kV is elevated to 220kV, then by the subsea cable access Infinite bus system of 20km;

The power quality data of having chosen the marine wind electric field that in two kinds of situations, electric pressure is 220kV is as evaluation object: 1. in the time that wind speed situation is different, and the power quality data at offshore grid-connected wind farm point place; 2. in the time that the generator installed capacity of marine wind electric field is different, and the power quality data at site place, totally three groups of emulated datas: a, wind speed steadily, when installed capacity 50MW; B, fluctuations in wind speed are large, when installed capacity 50MW; C, fluctuations in wind speed are large, when installed capacity 75MW;

6 power quality index data are all to adopt the large value of 95% probability.The large value of 95% probability is that all measured values are arranged from big to small, gives up front 5% and gets remaining maximal value, obtains the corresponding large value of 95% probability according to the power quality index data in three kinds of situations of following marine wind electric field,

The method of the definite power quality index weight described in step (2) is:

1) Set Pair Analysis is with Pair Analysis model, the certainty and uncertainty of things to be processed as a system, to the characteristic spread analysis of a set antithetical phrase, sets up these two the I. D.C connexion formula μ that are integrated under relevant issues background

μ＝a+bi+cj (1)

Wherein, a, b, c is respectively institute's analects and is combined in identical degree, diversity factor, the opposition degree under relevant issues background, and i is diversity factor coefficient, is defined in [1,1] in interval, look different situation value, j is opposition degree coefficient, is defined as-1, and a and c determine relatively, b is relatively uncertain, because of changeability, the complicacy of objective objects, be familiar with ambiguity and subjectivity when portraying objective things, can cause this relatively uncertain;

Be provided with r position expert, index set X={X _k, k=1,2 ..., n, every expert compares index between two, Judgement Matricies M _zkl, z=1,2 ..., r, k=1,2 ..., n, l=1,2 ..., n, is shown below, and represents the view of z position expert to the relative important relationship between any two indexs, x _z23represent that the 2nd index compare with the 3rd index, the 3rd important degree of index of the 2nd ratio;

$M_{\mathrm{zkl}}=\left[\begin{array}{cccc}{x}_{z11}& x_{z12}& ...& x_{z1n}\\ x_{z21}& x_{z22}& ...& x_{z2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ x_{\mathrm{zn}1}& x_{\mathrm{zn}2}& ...& x_{\mathrm{znn}}\end{array}\right]---\left(2\right)$

Wherein $x_{\mathrm{zkl}}=\frac{1}{x_{\mathrm{zlk}}};$

Utilize the matrix form of Pair Analysis to set up the Pair Analysis model μ of description indexes relative importance relation _qkl, that is:

${\mathrm{\μ}}_{\mathrm{qkl}}=A_{\mathrm{kl}}+B_{\mathrm{kl}}i=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ a_{21}& a_{22}& ...& a_{2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ a_{n1}& a_{n2}& ...& a_{\mathrm{nn}}\end{array}\right]+\left[\begin{array}{cccc}{b}_{11}& b_{12}& ...& b_{1n}\\ b_{21}& b_{22}& ...& b_{2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ b_{n1}& b_{n2}& ...& b_{\mathrm{nn}}\end{array}\right]i---\left(3\right)$

Wherein,

$a_{\mathrm{kl}}=\left\{\begin{array}{cc}\underset{z}{\min }\left\{x_{\mathrm{zkl}}\right\}& x_{\mathrm{zkl}}\≥1\\ \underset{z}{\max }\left\{x_{\mathrm{zkl}}\right\}& x_{\mathrm{zkl}}\≤1\end{array}\right.---\left(4\right)$

z＝1,2,…,r；k＝1,2,…,n；l＝1,2,…,n

$b_{\mathrm{kl}}=\left\{\begin{array}{cc}\left|\underset{z}{\max }\right\{x_{\mathrm{zkl}}\}-\underset{z}{\min }\{x_{\mathrm{zkl}}\left\}\right|& x_{\mathrm{zkl}}\≥1\\ (-1)\left|\underset{z}{\max }\right\{x_{\mathrm{zkl}}\}-\underset{z}{\min }\{x_{\mathrm{zkl}}\left\}\right|& x_{\mathrm{zkl}}\≤1\end{array}\right.---\left(5\right)$

z＝1,2,…,r；k＝1,2,…,n；l＝1,2,…,n

A _klfor describing the understanding homogeneity matrix of expert to the relative importance between each index; B _klfor describing the cognitive diversity matrix of expert to the relative importance between each index.For research object herein, the situation of opposition does not have, therefore j=0;

A _klwhat represent is the common understandings of different experts to evaluation index relative importance relation, is complete relation that can be definite; To x _zkl≤ 1 and x _zkl>=1 situation is value respectively, represents different experts' consensus of opinion, works as x _zlk≤ 1 o'clock, a _lk=max{x _zlk, because $x_{\mathrm{zkl}}=\frac{1}{x_{\mathrm{zlk}}},$ So $a_{\mathrm{lk}}=\max \left\{x_{\mathrm{zlk}}\right\}=\frac{1}{\min \left\{x_{\mathrm{zkl}}\right\}};$ Work as x _zkl>=1 o'clock, X _kwith X _lrelation at least will to meet scale value be min{x _zklcorresponding situation, the trend of its variation is scale value y > min{x _zklcorresponding situation but be no more than scale value y=max{x _zklcorresponding state.Obtain thus, adopt tectonic link degree μ in this way _qklin homogeneity matrix A _kl, this matrix makes to represent index X _lwith X _krelation two coefficient correspondences get up;

2) consistent matrix method is processed homogeneity matrix A _kl, obtain basis matrix D _kl, wherein D _kl=(d _kl), d _klbasis matrix D _klelement, and d satisfies condition _kk=1 He homogeneity matrix A _klcorresponding index weights ω _kcalculated by following formula:

${\mathrm{\ω}}_k=\frac{c_k}{\underset{s=1}{\overset{n}{\mathrm{\Σ}}}{c}_s},k=\mathrm{1,2},...,n---\left(6\right)$

Wherein, $c_s=\sqrt[n]{\underset{l=1}{\overset{n}{\mathrm{\Π}}}{d}_{\mathrm{kl}}}---\left(7\right)$

3) entropy power method is processed otherness matrix B _kl, obtain B _klcorresponding otherness weights, these weights have reflected the different opinions of brainstrust, concrete steps are as follows:

A) according to the entropy of the each evaluation index of concept definition of traditional entropy be:

$H_k=-\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{f}_{\mathrm{kl}}\ln f_{\mathrm{kl}}\right)/\ln n,k=\mathrm{1,2},...,n;l=\mathrm{1,2},...,n---\left(8\right)$

In formula: b _klit is otherness matrix B _klelement;

Obviously work as f _kl=0 o'clock, lnf _klmeaningless, therefore to f _klcalculating revised, be defined as

$f_{\mathrm{kl}}=(1+b_{\mathrm{kl}})/\underset{l=1}{\overset{n}{\mathrm{\Σ}}}(1+b_{\mathrm{kl}})---\left(9\right)$

B) entropy that calculates indices is weighed

$\mathrm{\η}={\left({\mathrm{\η}}_k\right)}_{1\×n},{\mathrm{\η}}_k=(1-H_k)/(n-\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{H}_k)---\left(10\right)$

And meet $\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\η}}_k=1;$

When getting while determining i value, obtain determining evaluation index weight, work as n=6, try to achieve index weights W by formula (6) and formula (10);

4) the concrete computation process of weight:

By 9 experts, 6 of the wind energy turbine set quality of power supply indexs are compared between two, use nine scale standards in analytical hierarchy process, obtain evaluating matrix group M _kl=(M _1kl, M _2kl..., M _9kl), evaluate power quality index in matrix and arrange in the following order: supply voltage deviation I ₁/ %, voltage fluctuation I ₂/ %, Short Term Flicker I ₃, percent harmonic distortion I ₄/ %, three-phase imbalance I ₅/ % and frequency departure I ₆/ Hz, provides two expert opinion matrixes below as example:

$M_{1\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/2& 1& 1& 1/3\\ 2& 1& 1& 2& 2& 1/2\\ 2& 1& 1& 2& 2& 1/2\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 3& 2& 2& 3& 3& 1\end{array}\right]$ $M_{2\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/3& 1/2& 1& 1/3\\ 2& 1& 1& 1& 2& 1/2\\ 3& 1& 1& 1& 2& 1/2\\ 2& 1& 1& 1& 2& 1/2\\ 1& 1/2& 1/2& 1/2& 1& 1/3\\ 3& 2& 2& 2& 3& 1\end{array}\right]---\left(11\right)$

Obtain homogeneity matrix by formula (4):

$A_{\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/2& 1& 1& 1/3\\ 2& 1& 1& 1& 2& 1/2\\ 2& 1& 1& 1& 2& 1/2\\ 1& 1& 1& 1& 1& 1/2\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 3& 2& 2& 2& 3& 1\end{array}\right]---\left(12\right)$

Obtain otherness matrix by formula (5):

$B_{\mathrm{kl}}=\left[\begin{array}{cccccc}0& 0& -1/6& -1/2& 0& 0\\ 0& 0& 0& 1& 0& 0\\ 1& 0& 0& 1& 0& 0\\ 1& -1/2& -1/2& 0& 1& -1/6\\ 0& 0& 0& -1/2& 0& 0\\ 0& 0& 0& 1& 0& 0\end{array}\right]---\left(13\right)$

Adopt consistent matrix method to process homogeneity matrix A _kl:

$D_{\mathrm{kl}}=\left[\begin{array}{cccccc}1& 0.5888& 0.5888& 0.7418& 1& 0.3240\\ 1.6984& 1& 1& 1.2599& 1.6984& 0.5503\\ 1.6984& 1& 1& 1.2599& 1.6984& 0.5503\\ 1.3480& 0.7937& 0.7937& 1& 1.3480& 0.4368\\ 1& 0.5888& 0.5888& 0.7418& 1& 0.3240\\ 3.0862& 1.8171& 1.8171& 2.2894& 3.0862& 1\end{array}\right]---\left(14\right)$

According to formula (7), we obtain:

C _s(6×1)＝[0.6608,1.1224,1.1224,0.8908,0.6608,2.0395] ^T

Obtain homogeneity matrix A by formula (6) _klcorresponding weight:

W _k(6×1)＝[ω ₁,ω ₂,ω ₃,ω ₄,ω ₅,ω ₆] ^T＝(0.1017,0.1728,0.1728,0.1371,0.1017,0.3139) ^T (15)

Obtain otherness matrix B by formula (8)～formula (10) _klcorresponding weight:

η _(1×6)＝[η ₁,η ₂,η ₃,η ₄,η ₅,η ₆] ^T＝[0.0701,0.1265,0.1697,0.4379,0.0692,0.1265] ^T (16)

Get i=0.08, by formula (15) and formula (16) must power quality index weight be:

W＝[0.0994,0.1694,0.1726,0.1594,0.0993,0.3000] ^T (17)。

Described step (3) is:

Adopt the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of Vague value representation that the power quality index data value table of each evaluation object is shown as to Vague value; By 21 grades of linguistic variables of Vague value representation be:

Grade	Vague value	Abstention situation
			AG	[1,1]	0
AG -	[0.95,0.975]	0.025
			VG	[0.9,0.95]	0.05
VG -	[0.85,0.925]	0.075
			G	[0.8,0.9]	0.1
G -	[0.75,0.875]	0.125
			FG	[0.7,0.85]	0.15
FG -	[0.65,0.825]	0.175
			MG	[0.6,0.8]	0.2
MG -	[0.55,0.75]	0.2
			M	[0.5,0.5]	0
M -	[0.45,0.65]	0.2
			MP	[0.4,0.6]	0.2
MP -	[0.35,0.525]	0.175
			FP	[0.3,0.45]	0.15
FP -	[0.25,0.375]	0.125
			P	[0.2,0.3]	0.1
P -	[0.15,0.225]	0.075
			VP	[0.1,0.15]	0.05
VP -	[0.05,0.075]	0.025
			AP	[0,0]	0

The quality of power supply grade classification of the each single index under 220kV electric pressure is:

Described step (4) is: form respectively optimal value vector Z by indices optimal value and the most bad value ⁺the most bad value vector Z ^-, i.e. definitely positive ideal solution and definitely negative ideal solution

$Z^{+}=(z_1^{+},z_2^{+},...z_p^{+});$ $Z^{-}=(z_1^{-},z_2^{-},...,z_p^{-}),---\left(18\right)$

Wherein, $z_j^{+}=\max \{z_{1j},z_{2j},...,z_{\mathrm{nj}}\},$ $z_j^{-}=\min \{z_{1j},z_{2j},...,z_{\mathrm{nj}}\},j=\mathrm{1,2},...,p$

In the time only considering the weight of index, consider using the ideal value in GB and limit value as definitely positive ideal solution and definitely negative ideal solution.

Described step (5) is:

If U is a nonempty set, its element represents with x, and a Vague collection A on U refers to a pair of subordinate function t on U _aand f _a,

t _A：U→[0,1]，f _A：U→[0,1]

Meet t _a(x)+f _a(x)≤1, and 0≤t _a(x)≤1,0≤f _a(x)≤1, wherein: t _afor the true subordinate function of Vague collection A, express support for the degree of membership lower bound of the evidence of x ∈ A; f _afor the false subordinate function of Vague collection A, the degree of membership lower bound of the evidence of the x ∈ A that makes difficulties; If x ∈ is U, claim closed interval [t _a(x), 1-f _a(x)] be Vague collection A in the Vague of x value,

Two kinds of Vague value Similarity Measures,

1) Similarity Measures between improved Vague value is as follows:

$M_z(x,y)=1-\frac{|t_x-t_y-(f_x-f_y)|}{8}-\frac{|t_x-t_y+f_x-f_y|}{4}-\frac{|t_x-t_y|+|f_x-f_y|}{8}---\left(19\right)$

M _zthe value of (x, y) is larger, represents that Vague value x is more similar with y,

2) improved weighting Vague value Similarity Measures is as follows:

If x, y is two Vague values, and the weight of x and y is all w, wherein, w=(a, b, c), a represents the weight of true subordinate function part, b represents the weight of false subordinate function part, c represents the weight of unknown portions, considers to affect three factors of Vague value similarity measure simultaneously, and improved weighting Vague value Similarity Measures is:

$\begin{array}{c}{M}_w^z(x,y)=1-\frac{|a*(t_x-t_y)-b*(f_x-f_y)|}{2(a+b+c)}\\ +\frac{a*|t_x-t_y|+b*|f_x-f_y|+c*|t_x-t_y+f_x-f_y|}{2(a+b+c)}\end{array}---\left(20\right)$

In formula (20), a, b, c>=0, and a+b+c > 0, value larger, represent that Vague value x is more similar with y.Work as a=1, b=1, when c=2, formula (20) is exactly (19);

Calculating each evaluation object with the distance method of definitely positive ideal solution and definitely negative ideal solution is:

1. only consider the weight of index, the similarity measure formula of Vague value is formula (19)

${\mathrm{\Γ}}_i^{*}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VPIS}),i=\mathrm{1,2},...,(n+2)---\left(21\right)$

${\mathrm{\Γ}}_i^{-}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VNIS}),i=\mathrm{1,2},...,(n+2)---\left(22\right)$

2. not only consider the weight of index, consider the weight of the true subordinate function of each index, false subordinate function, abstention part simultaneously

${\mathrm{\Γ}}_i^{*}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_w^z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VPIS}),i=\mathrm{1,2},...,(n+2)---\left(23\right)$

${\mathrm{\Γ}}_i^{-}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_w^z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VNIS}),i=\mathrm{1,2},...,(n+2)---\left(24\right).$

Based on Vague collection with improve the energy quality comprehensive assessment method that approaches ideal solution and be applied to the power quality controlling of marine wind electric field: parallel connection type active electric filter APF, improves the harmonic wave of marine wind electric field; Static Var Compensator SVC, STATCOM STATCOM, compensate the reactive power of marine wind electric field.

Compared with prior art, beneficial effect of the present invention is:

1, the present invention is determining when weight, and the analytical hierarchy process that consistent matrix method is more traditional, without the judgment matrix that carries out consistency desired result and just can directly be met coherence request, has been simplified computation process; Entropy power method energy Objective Weight; Method of Set Pair Analysis organically combines certainty and uncertainty, and the uncertainty that the otherness that description and processing expert are familiar with in the time determining index weights causes has reduced the subjectivity randomness judging.

2, improved TOPSIS method of the present invention is different with a lot of intelligent methods, data are not had to strict requirement, be not subject to sample content, data distribution pattern, the how many restriction of index, be not only applicable to small sample data, also be applicable to the large system of multiple-unit evaluation and many indexs, be also applicable to dynamic, continuity data; Can directly make full use of primary data information (pdi) and calculate, in computation process, not reduce variable number, the good and bad degree of the different evaluation objects of expression that ranking results can be quantitative; Because can eliminate the impact of different dimensions, therefore can carry out comprehensive assessment to the index of different dimensions simultaneously; Method is simple, directly perceived, reliable, clear and definite, rational in infrastructure, applying flexible sort.

3, Vague collection is considered to be subordinate to and the non-evidence that is subordinate to two aspects simultaneously, in power quality index comprehensive assessment, considers the one side of index of correlation high-quality, can also reflect one side and the uncertain one side of its poor quality simultaneously.In the time considering each evaluation object to the distance of VPIS and VNIS, true, the false subordinate function of Weighted Guidelines and each index and the weight of abstention part have been considered simultaneously, the influence factor that has represented the quality of power supply to different monitoring targets is different, has embodied the relative importance of each point to positive and negative ideal solution distance.The method can be generalized in the power quality index comprehensive assessment of other occasions and goes.

4, the present invention not only obtains the sequence of the power quality index at marine wind electric field points of common connection place, and obtains good and bad grade and the sequence thereof of the quality of power supply.

5, the present invention applies in the electric energy quality synthesis evaluation of marine wind electric field, and gained comprehensive assessment result can help to grasp in time the quality of power supply of wind power plant, improves operation and management level; Be definite foundation that provides of the rate for incorporation into the power network of Oversea wind power generation to a certain extent; Power quality treatment method can reasonably be selected in time according to electric energy quality synthesis evaluation result by generating side.

Brief description of the drawings

Fig. 1 is offshore wind farm grid-connected system structural drawing;

Fig. 2 is the wind speed of two kinds of marine wind electric fields using in emulation;

Fig. 3 is the offshore wind farm grid-connected system structural drawing that adds Static Var Compensator SVC.

Embodiment

Below in conjunction with the drawings and specific embodiments, the invention will be further described.

Comprise the steps: based on Vague collection and the energy quality comprehensive assessment method that improvement approaches ideal solution

(1) on PSCAD/EMTDC software platform, offshore grid-connected wind farm is simulated, obtain the emulated data of 6 power quality indexs at offshore grid-connected wind farm points of common connection place;

(2) judgment matrix that uses analytical hierarchy process in conjunction with the Pair Analysis Construction of A Model of Set Pair Analysis Theory understanding homogeneity matrix and cognitive diversity matrix, consistent matrix method is processed homogeneity matrix and is obtained corresponding weight value, objectively entropy power method is processed otherness matrix and is obtained corresponding weight, finally by Set Pair Analysis Theory, both combinations is obtained to power quality index weight vectors W;

(3) invite expert to evaluate the quality of evaluation object and each power quality index, adopt the scoring to each index of Vague value representation each evaluation object, definite method of the Vague value of each index is: in conjunction with the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of use Vague value representation;

(4) determine definitely positive ideal solution and definitely negative ideal solution, in the time only considering the weight of index, using the ideal value in GB and limit value as absolute ideal solution and definitely negative ideal solution, join together in the middle of evaluation object, when this ideal value refers to that working as 6 adopted power quality index values is tending towards 0, using numerical value 0 as ideal value;

(5) method that adopts improved weighting to approach ideal solution on the basis of Vague collection theory, calculates each evaluation object and the absolute just distance of ideal solution distance with definitely negative ideal solution the weight vectors W of the power quality index that institute's weighted value is each evaluation object;

(6) according to the distance of the definitely positive ideal solution having calculated distance with definitely negative ideal solution calculate the relative approach degree σ of each evaluation object _i,

${\mathrm{\σ}}_i=\frac{{\mathrm{\Γ}}_i^{*}}{{\mathrm{\Γ}}_i^{-}+{\mathrm{\Γ}}_i^{*}},i=\mathrm{1,2},...,n$

The relative approach degree σ of evaluation object _iless, represent that evaluation object approaches positive ideal solution and away from negative ideal solution, by approaching degree principle, the quality of power supply carried out to comprehensive assessment and sequence.

Described step (1) is:

Marine wind electric field emulation adopts the situation of two kinds of capacity: 50MW and 75MW, wherein, 50MW includes the double-fed wind power generator group of 10 unit 5MW, 75MW includes the double-fed wind power generator group of 15 unit 5MW, every typhoon group of motors is elevated to 35kV outlet voltage from 0.69kV through machine end step-up transformer, then pass through current collection network in wind energy turbine set, be connected to the low-pressure side of offshore boosting station, by offshore boosting station, the voltage that collects of 35kV is elevated to 220kV, then by the subsea cable access Infinite bus system of 20km.Offshore wind farm grid-connected system structural drawing as shown in Figure 1.

The power quality data of having chosen the marine wind electric field that in two kinds of situations, electric pressure is 220kV is as evaluation object: 1. in the time that wind speed situation is different, and the power quality data at offshore grid-connected wind farm point place; 2. in the time that the generator installed capacity of marine wind electric field is different, and the power quality data at site place, totally three groups of emulated datas: a, wind speed steadily, when installed capacity 50MW; B, fluctuations in wind speed are large, when installed capacity 50MW; C, fluctuations in wind speed are large, when installed capacity 75MW, two kinds of wind speed in these three groups of data are as shown in Figure 2.

6 power quality index data are all to adopt the large value of 95% probability.The large value of 95% probability is that all measured values are arranged from big to small, gives up front 5% and gets remaining maximal value, obtains the corresponding large value of 95% probability, in table 1 according to the power quality index data in three kinds of situations of following marine wind electric field.

Power quality index data in three kinds of situations of table 1 marine wind electric field

Object

Voltage deviation

Voltage fluctuation

Flickering

Harmonic wave is abnormal

Three-phase

Frequency

	/％	/％		Variability/%	Imbalance/%	Deviation/Hz
							a	2.0942	0.2589	0.5194	1.8577	0.0902	0.0051
b	2.1062	0.2586	0.528	2	0.0899	0.0054
							c	2.1340	0.3113	0.7578	2	0.1212	0.0068
Ideal value in GB	0	0	0	0	0	0
							Limit value	10	3	0.8	2	2	0.2

The method of the definite power quality index weight described in step (2) is:

1) Set Pair Analysis is with Pair Analysis model, the certainty and uncertainty of things to be processed as a system, to the characteristic spread analysis of a set antithetical phrase, sets up these two the I. D.C connexion formula μ that are integrated under relevant issues background

μ＝a+bi+cj (1)

Wherein, a, b, c is respectively institute's analects and is combined in identical degree, diversity factor, the opposition degree under relevant issues background, and i is diversity factor coefficient, is defined in [1,1] in interval, look different situation value, j is opposition degree coefficient, is defined as-1, and a and c determine relatively, b is relatively uncertain, because of changeability, the complicacy of objective objects, be familiar with ambiguity and subjectivity when portraying objective things, can cause this relatively uncertain;

Expert has different views to the same relation in the time determining index weights, exists the otherness in subjective understanding, exists uncertain.Be provided with r position expert, index set X={X _k, k=1,2 ..., n, every expert compares index between two, Judgement Matricies M _zkl, z=1,2 ..., r; K=1,2 ..., n; L=1,2 ..., n, is shown below, and represents the view of z position expert to the relative important relationship between any two indexs, x _z23represent that the 2nd index compare with the 3rd index, the 3rd important degree of index of the 2nd ratio;

$M_{\mathrm{zkl}}=\left[\begin{array}{cccc}{x}_{z11}& x_{z12}& ...& x_{z1n}\\ x_{z21}& x_{z22}& ...& x_{z2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ x_{\mathrm{zn}1}& x_{\mathrm{zn}2}& ...& x_{\mathrm{znn}}\end{array}\right]---\left(2\right)$

Wherein $x_{\mathrm{zkl}}=\frac{1}{x_{\mathrm{zlk}}};$

Utilize the matrix form of Pair Analysis to set up the Pair Analysis model μ of description indexes relative importance relation _qkl, that is:

${\mathrm{\μ}}_{\mathrm{qkl}}=A_{\mathrm{kl}}+B_{\mathrm{kl}}i=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ a_{21}& a_{22}& ...& a_{2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ a_{n1}& a_{n2}& ...& a_{\mathrm{nn}}\end{array}\right]+\left[\begin{array}{cccc}{b}_{11}& b_{12}& ...& b_{1n}\\ b_{21}& b_{22}& ...& b_{2n}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ b_{n1}& b_{n2}& ...& b_{\mathrm{nn}}\end{array}\right]i---\left(3\right)$

Wherein,

$a_{\mathrm{kl}}=\left\{\begin{array}{cc}\underset{z}{\min }\left\{x_{\mathrm{zkl}}\right\}& x_{\mathrm{zkl}}\≥1\\ \underset{z}{\max }\left\{x_{\mathrm{zkl}}\right\}& x_{\mathrm{zkl}}\≤1\end{array}\right.---\left(4\right)$

z＝1,2,…,r；k＝1,2,…,n；l＝1,2,…,n

$b_{\mathrm{kl}}=\left\{\begin{array}{cc}\left|\underset{z}{\max }\right\{x_{\mathrm{zkl}}\}-\underset{z}{\min }\{x_{\mathrm{zkl}}\left\}\right|& x_{\mathrm{zkl}}\≥1\\ (-1)\left|\underset{z}{\max }\right\{x_{\mathrm{zkl}}\}-\underset{z}{\min }\{x_{\mathrm{zkl}}\left\}\right|& x_{\mathrm{zkl}}\≤1\end{array}\right.---\left(5\right)$

z＝1,2,…,r；k＝1,2,…,n；l＝1,2,…,n

A _klfor describing the understanding homogeneity matrix of expert to the relative importance between each index; B _klfor describing the cognitive diversity matrix of expert to the relative importance between each index.For research object herein, the situation of opposition does not have, therefore j=0;

A _klwhat represent is the common understandings of different experts to evaluation index relative importance relation, is complete relation that can be definite; To x _zkl≤ 1 and x _zkl>=1 situation is value respectively, represents different experts' consensus of opinion, works as x _zlk≤ 1 o'clock, a _lk=max{x _zlk, because $x_{\mathrm{zkl}}=\frac{1}{x_{\mathrm{zlk}}},$ So $a_{\mathrm{lk}}=\max \left\{x_{\mathrm{zlk}}\right\}=\frac{1}{\min \left\{x_{\mathrm{zkl}}\right\}};$ Work as x _zkl>=1 o'clock, X _kwith X _lrelation at least will to meet scale value be min{x _zklcorresponding situation, the trend of its variation is scale value y > min{x _zklcorresponding situation but be no more than scale value y=max{x _zklcorresponding state.Obtain thus, adopt tectonic link degree μ in this way _qklin homogeneity matrix A _kl, this matrix makes to represent index X _lwith X _krelation two coefficient correspondences get up;

2) consistent matrix method is processed homogeneity matrix A _kl, obtain basis matrix D _kl, wherein D _kl=(d _kl), d _klbasis matrix D _klelement, and d satisfies condition _kk=1 He homogeneity matrix A _klcorresponding index weights ω _kcalculated by following formula:

${\mathrm{\ω}}_k=\frac{c_k}{\underset{s=1}{\overset{n}{\mathrm{\Σ}}}{c}_s},k=\mathrm{1,2},...,n---\left(6\right)$

Wherein, $c_s=\sqrt[n]{\underset{l=1}{\overset{n}{\mathrm{\Π}}}{d}_{\mathrm{kl}}}---\left(7\right)$

3) entropy power method is processed otherness matrix B _kl, obtain B _klcorresponding otherness weights, these weights have reflected the different opinions of brainstrust, concrete steps are as follows:

A) according to the entropy of the each evaluation index of concept definition of traditional entropy be:

$H_k=-\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{f}_{\mathrm{kl}}\ln f_{\mathrm{kl}}\right)/\ln n,k=\mathrm{1,2},...,n;l=\mathrm{1,2},...,n---\left(8\right)$

In formula: b _klit is otherness matrix B _klelement;

Obviously work as f _kl=0 o'clock, lnf _klmeaningless, therefore to f _klcalculating revised, be defined as

$f_{\mathrm{kl}}=(1+b_{\mathrm{kl}})/\underset{l=1}{\overset{n}{\mathrm{\Σ}}}(1+b_{\mathrm{kl}})---\left(9\right)$

B) entropy that calculates indices is weighed

$\mathrm{\η}={\left({\mathrm{\η}}_k\right)}_{1\×n},{\mathrm{\η}}_k=(1-H_k)/(n-\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{H}_k)---\left(10\right)$

And meet $\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\η}}_k=1;$

Due to diversity factor coefficient i ∈ [0,1], so the index weights obtaining is a value range instead of normal value, along with people are to the more deep research of the quality of power supply and understanding, the scope of diversity factor i can diminish gradually, and then the value range of weight also will diminish thereupon.This just represents that the cognitive diversity of brainstrust and subjective random degree reduce gradually, and index weights is changed to determinacy by uncertainty.When getting while determining i value, obtain determining evaluation index weight, work as n=6, try to achieve index weights W by formula (6) and formula (10);

4) the concrete computation process of weight:

By 9 experts, 6 of the wind energy turbine set quality of power supply indexs are compared between two, use nine scale standards in analytical hierarchy process, obtain evaluating matrix group M _kl=(M _1kl, M _2kl..., M _9kl), evaluate power quality index in matrix and arrange in the following order: supply voltage deviation I ₁/ %, voltage fluctuation I ₂/ %, Short Term Flicker I ₃, percent harmonic distortion I ₄/ %, three-phase imbalance I ₅/ % and frequency departure I ₆/ Hz, provides two expert opinion matrixes below as example:

$M_{1\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/2& 1& 1& 1/3\\ 2& 1& 1& 2& 2& 1/2\\ 2& 1& 1& 2& 2& 1/2\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 3& 2& 2& 3& 3& 1\end{array}\right]$ $M_{2\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/3& 1/2& 1& 1/3\\ 2& 1& 1& 1& 2& 1/2\\ 3& 1& 1& 1& 2& 1/2\\ 2& 1& 1& 1& 2& 1/2\\ 1& 1/2& 1/2& 1/2& 1& 1/3\\ 3& 2& 2& 2& 3& 1\end{array}\right]---\left(11\right)$

Obtain homogeneity matrix by formula (4):

$A_{\mathrm{kl}}=\left[\begin{array}{cccccc}1& 1/2& 1/2& 1& 1& 1/3\\ 2& 1& 1& 1& 2& 1/2\\ 2& 1& 1& 1& 2& 1/2\\ 1& 1& 1& 1& 1& 1/2\\ 1& 1/2& 1/2& 1& 1& 1/3\\ 3& 2& 2& 2& 3& 1\end{array}\right]---\left(12\right)$

Obtain otherness matrix by formula (5):

$B_{\mathrm{kl}}=\left[\begin{array}{cccccc}0& 0& -1/6& -1/2& 0& 0\\ 0& 0& 0& 1& 0& 0\\ 1& 0& 0& 1& 0& 0\\ 1& -1/2& -1/2& 0& 1& -1/6\\ 0& 0& 0& -1/2& 0& 0\\ 0& 0& 0& 1& 0& 0\end{array}\right]---\left(13\right)$

Adopt consistent matrix method to process homogeneity matrix A _kl:

$D_{\mathrm{kl}}=\left[\begin{array}{cccccc}1& 0.5888& 0.5888& 0.7418& 1& 0.3240\\ 1.6984& 1& 1& 1.2599& 1.6984& 0.5503\\ 1.6984& 1& 1& 1.2599& 1.6984& 0.5503\\ 1.3480& 0.7937& 0.7937& 1& 1.3480& 0.4368\\ 1& 0.5888& 0.5888& 0.7418& 1& 0.3240\\ 3.0862& 1.8171& 1.8171& 2.2894& 3.0862& 1\end{array}\right]---\left(14\right)$

According to formula (7), we obtain:

C _s(6×1)＝[0.6608,1.1224,1.1224,0.8908,0.6608,2.0395] ^T

Obtain homogeneity matrix A by formula (6) _klcorresponding weight:

W _k(6×1)＝[ω ₁,ω ₂,ω ₃,ω ₄,ω ₅,ω ₆] ^T＝(0.1017,0.1728,0.1728,0.1371,0.1017,0.3139) ^T (15)

Obtain otherness matrix B by formula (8)～formula (10) _klcorresponding weight:

η _(1×6)＝[η ₁,η ₂,η ₃,η ₄,η ₅,η ₆] ^T＝[0.0701,0.1265,0.1697,0.4379,0.0692,0.1265] ^T (16)

Consider that the suggestion otherness between expert is smaller, therefore coefficient of variation is got i=0.08, by formula (15) and formula (16) must power quality index weight be:

W＝[0.0994,0.1694,0.1726,0.1594,0.0993,0.3000] ^T (17)。

Described step (3) is:

By the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of Vague value representation respectively as shown in table 2 and table 3.According to table 2 and table 3, the power quality index data value table of each evaluation object of table 1 is shown as to Vague value;

21 grades of linguistic variables of Vague value representation for table 2

Grade	Vague value	Abstention situation
			AG	[1,1]	0
AG -	[0.95,0.975]	0.025
			VG	[0.9,0.95]	0.05
VG -	[0.85,0.925]	0.075
			G	[0.8,0.9]	0.1
G -	[0.75,0.875]	0.125
			FG	[0.7,0.85]	0.15
FG -	[0.65,0.825]	0.175
			MG	[0.6,0.8]	0.2
MG -	[0.55,0.75]	0.2
			M	[0.5,0.5]	0
M -	[0.45,0.65]	0.2
			MP	[0.4,0.6]	0.2
MP -	[0.35,0.525]	0.175
			FP	[0.3,0.45]	0.15
FP -	[0.25,0.375]	0.125
			P	[0.2,0.3]	0.1
P -	[0.15,0.225]	0.075
			VP	[0.1,0.15]	0.05
VP -	[0.05,0.075]	0.025
			AP	[0,0]	0

The quality of power supply grade classification of the each single index under table 3 220kV electric pressure

Described step (4) is: form respectively optimal value vector Z by indices optimal value and the most bad value ⁺the most bad value vector Z ^-, i.e. definitely positive ideal solution and definitely negative ideal solution

$Z^{+}=(z_1^{+},z_2^{+},...z_p^{+});$ $Z^{-}=(z_1^{-},z_2^{-},...,z_p^{-}),---\left(18\right)$

Wherein, $z_j^{+}=\max \{z_{1j},z_{2j},...,z_{\mathrm{nj}}\},$ $z_j^{-}=\min \{z_{1j},z_{2j},...,z_{\mathrm{nj}}\},j=\mathrm{1,2},...,p$

In the time only considering the weight of index, consider using the ideal value in GB and limit value as definitely positive ideal solution and definitely negative ideal solution.

Described step (5) is:

If U is a nonempty set, its element represents with x, and a Vague collection A on U refers to a pair of subordinate function t on U _aand f _a,

t _A：U→[0,1]，f _A：U→[0,1]

Meet t _a(x)+f _a(x)≤1, and 0≤t _a(x)≤1,0≤f _a(x)≤1, wherein: t _afor the true subordinate function of Vague collection A, express support for the degree of membership lower bound of the evidence of x ∈ A; f _afor the false subordinate function of Vague collection A, the degree of membership lower bound of the evidence of the x ∈ A that makes difficulties; If x ∈ is U, claim closed interval [t _a(x), 1-f _a(x)] be Vague collection A in the Vague of x value,

Two kinds of Vague value Similarity Measures,

1) Similarity Measures between improved Vague value is as follows:

$M_z(x,y)=1-\frac{|t_x-t_y-(f_x-f_y)|}{8}-\frac{|t_x-t_y+f_x-f_y|}{4}-\frac{|t_x-t_y|+|f_x-f_y|}{8}---\left(19\right)$

M _zthe value of (x, y) is larger, represents that Vague value x is more similar with y,

2) improved weighting Vague value Similarity Measures is as follows:

If x, y is two Vague values, and the weight of x and y is all w, wherein, w=(a, b, c), a represents the weight of true subordinate function part, b represents the weight of false subordinate function part, c represents the weight of unknown portions, considers to affect three factors of Vague value similarity measure simultaneously, and improved weighting Vague value Similarity Measures is:

$\begin{array}{c}{M}_w^z(x,y)=1-\frac{|a*(t_x-t_y)-b*(f_x-f_y)|}{2(a+b+c)}\\ +\frac{a*|t_x-t_y|+b*|f_x-f_y|+c*|t_x-t_y+f_x-f_y|}{2(a+b+c)}\end{array}---\left(20\right)$

In formula (20), a, b, c>=0, and a+b+c > 0, value larger, represent that Vague value x is more similar with y.Work as a=1, b=1, when c=2, formula (20) is exactly (19);

Calculating each evaluation object with the distance method of definitely positive ideal solution and definitely negative ideal solution is:

1. only consider the weight of index, the similarity measure formula of Vague value is formula (19)

${\mathrm{\Γ}}_i^{*}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VPIS}),i=\mathrm{1,2},...,(n+2)---\left(21\right)$

${\mathrm{\Γ}}_i^{-}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VNIS}),i=\mathrm{1,2},...,(n+2)---\left(22\right)$

2. not only consider the weight of index, consider the weight of the true subordinate function of each index, false subordinate function, abstention part simultaneously

${\mathrm{\Γ}}_i^{*}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_w^z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VPIS}),i=\mathrm{1,2},...,(n+2)---\left(23\right)$

${\mathrm{\Γ}}_i^{-}=\underset{j=1}{\overset{p}{\mathrm{\Σ}}}{W}_j{M}_w^z\left(\right[t_{\mathrm{ij}},t_{\mathrm{ij}}^{*}],\mathrm{VNIS}),i=\mathrm{1,2},...,(n+2)---\left(24\right).$

In described step (6), calculate the relative approach degree σ of each evaluation object _imethod be:

${\mathrm{\σ}}_i=\frac{{\mathrm{\Γ}}_i^{*}}{{\mathrm{\Γ}}_i^{-}+{\mathrm{\Γ}}_i^{*}},i=\mathrm{1,2},...,n---\left(25\right)$

σ _iless, represent that evaluation object more approaches ideal solution and away from negative ideal solution.By approaching degree principle, the quality of power supply is carried out to comprehensive assessment and sequence, obtain the good and bad grade of the quality of power supply.

By these 6 power quality indexs by being divided into 5 grades in the acceptability limit of national Specification: speciality, high-quality, good, medium and qualified, as shown in table 4.According to σ _iresiding position in table 4, can obtain the grade result of electricity quality evaluation.

Table 4 quality of power supply table of grading

Good degree	Speciality	High-quality	Well	Medium	Qualified
						Quality grade	0.3333,0.3704	0.4074,0.4444	0.4815,0.5185	0.5556,0.5926	0.6296,0.6667

Below a kind of of proposition is applied in the electric energy quality synthesis evaluation of marine wind electric field based on Vague collection and the energy quality comprehensive assessment method that improvement approaches ideal solution.The data of evaluation object are as shown in table 1.

Concrete appraisal procedure is as follows:

1) please expert be evaluated by the quality of evaluation object and each power quality index.By the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of Vague value representation as shown in table 2 and table 3, adopt the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of Vague value representation that every power quality index data value table of each evaluation object is shown as to Vague value, as shown in table 5.

2) ask the weight vectors W of the power quality index of each evaluation object according to formula (17).

3), while only considering the weight of index, consider, using the ideal value in GB and limit value as absolute ideal solution VPIS and definitely negative ideal solution VNIS, to join together in the middle of evaluation object.Calculate connect recency, sequence and the comprehensive assessment result of each assessment monitoring point to ideal solution according to formula (21) and formula (22), as shown in table 6.

The Vague value of each power quality index of table 5 evaluation object

Object

I 1

I 2

I 3

I 4

I 5

I 6

a

[0.8,0.9]

[0.95,0.975]

[0.35,0.525]

[0.05,0.075]

[1,1]

[1,1]

b

[0.8,0.9]

[0.95,0.975]

[0.35,0.525]

[0,0]

[1,1]

[1,1]

c

[0.8,0.9]

[0.9,0.95]

[0.05,0.075]

[0,0]

[0.95,0.975]

[1,1]

Ideal value in GB

[1,1]

[1,1]

[1,1]

[1,1]

[1,1]

[1,1]

Limit value

[0,0]

[0,0]

[0,0]

[0,0]

[0,0]

[0,0]

W j

0.0994

0.1694

0.1726

0.1594

0.0993

0.3

The recency that connects, sequence and the comprehensive assessment result of the each evaluation object of table 6 and ideal solution

4) not only consider the weight of index, consider the weight of the true subordinate function of each index, false subordinate function, abstention part simultaneously.Vague value, VPIS and the VNIS of each power quality index of evaluation object are as shown in table 7.(a for weight of the true subordinate function of each index, false subordinate function, abstention part _j, b _j, c _j) represent, j=1 ..., 6, this weight is next given by the expert of wind power generation field, supposes that weighted value is as shown in table 8.

Vague value, VPIS and the VNIS of each power quality index of table 7 evaluation object

Object

I 1

I 2

I 3

I 4

I 5

I 6

a

[0.8,0.9]

[0.95,0.975]

[0.35,0.525]

[0.05,0.075]

[1,1]

[1,1]

b

[0.8,0.9]

[0.95,0.975]

[0.35,0.525]

[0,0]

[1,1]

[1,1]

c

[0.8,0.9]

[0.9,0.95]

[0.05,0.075]

[0,0]

[0.95,0.975]

[1,1]

VPIS

[0.8,0.9]

[0.95,0.975]

[0.35,0.525]

[0.05,0.075]

[1,1]

[1,1]

VNIS

[0.8,0.9]

[0.9,0.95]

[0.05,0.075]

[0,0]

[0.95,0.975]

[1,1]

W j

0.0994

0.1694

0.1726

0.1594

0.0993

0.3

The weight of the true subordinate function of the each index of table 8, false subordinate function, abstention part

	I 1	I 2	I 3	I 4	I 5	I 6
							(a j,b j,c j)	(10,1,1)	(3,2,3)	(6,2,2)	(3,2,2)	(2,3,2)	(7,3,1)

5) calculate each evaluation object to the distance of VPIS, VNIS and the recency that connects according to formula (23)～formula (25), result is as shown in table 9.

Connect recency and the sequence of the each evaluation object of table 9 and ideal solution

The electric energy quality synthesis evaluation result that can be obtained the method by table 6 and table 9 is: a > b > c (" > " represent be better than), shows that quality of power supply when wind regime fluctuation is larger is poor when more steady than wind speed; Quality of power supply when blower fan installed capacity is larger is poorer, larger on quality of power supply impact electrical network and site.

Based on Vague collection with improve the energy quality comprehensive assessment method that approaches ideal solution and be applied to the power quality controlling of marine wind electric field: parallel connection type active electric filter APF, improves the harmonic wave of marine wind electric field; Static Var Compensator SVC, STATCOM STATCOM, compensate the reactive power of marine wind electric field.

Wherein, add Static Var Compensator SVC offshore wind farm grid-connected system structural drawing afterwards as shown in Figure 3.

Add that the realistic model of SVC power quality controlling adopts be that fluctuations in wind speed is large, wind energy turbine set model when installed capacity 50MW, power quality index data while not adding SVC are exactly b group data in table 1, add the power quality index data after SVC administers as shown in table 10, through comparative analysis, better while adding the power quality controlling device quality of power supply afterwards not administer, there is improvement significantly through administering indices.

Table 10 adds SVC and administers front and back marine wind electric field power quality index Data Comparison analysis

Claims (7)

1. an energy quality comprehensive assessment method that approaches ideal solution based on Vague collection and improvement, is characterized in that comprising the steps:

(1) on PSCAD/EMTDC software platform, offshore grid-connected wind farm is simulated, obtain the emulated data of 6 power quality indexs at offshore grid-connected wind farm points of common connection place;

(2) judgment matrix that uses analytical hierarchy process in conjunction with the Pair Analysis Construction of A Model of Set Pair Analysis Theory understanding homogeneity matrix and cognitive diversity matrix, consistent matrix method is processed homogeneity matrix and is obtained corresponding weight value, objectively entropy power method is processed otherness matrix and is obtained corresponding weight, finally by Set Pair Analysis Theory, both combinations is obtained to power quality index weight vectors W;

(3) invite expert to evaluate the quality of evaluation object and each power quality index, adopt the scoring to each index of Vague value representation each evaluation object, definite method of the Vague value of each index is: in conjunction with the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of use Vague value representation;

(4) determine definitely positive ideal solution and definitely negative ideal solution, in the time only considering the weight of index, using the ideal value in GB and limit value as absolute ideal solution and definitely negative ideal solution, join together in the middle of evaluation object, when this ideal value refers to that working as 6 adopted power quality index values is tending towards 0, using numerical value 0 as ideal value;

(5) method that adopts improved weighting to approach ideal solution on the basis of Vague collection theory, calculates each evaluation object and the absolute just distance of ideal solution distance with definitely negative ideal solution , the weight vectors W of the power quality index that institute's weighted value is each evaluation object;

(6) according to the distance of the definitely positive ideal solution having calculated distance with definitely negative ideal solution , calculate the relative approach degree σ of each evaluation object _i,

The relative approach degree σ of evaluation object _iless, represent that evaluation object approaches positive ideal solution and away from negative ideal solution, by approaching degree principle, the quality of power supply carried out to comprehensive assessment and sequence.

2. method according to claim 1, is characterized in that, described step (1) is:

Marine wind electric field emulation adopts the situation of two kinds of capacity: 50MW and 75MW, wherein, 50MW includes the double-fed wind power generator group of 10 unit 5MW, 75MW includes the double-fed wind power generator group of 15 unit 5MW, every typhoon group of motors is elevated to 35kV outlet voltage from 0.69kV through machine end step-up transformer, then pass through current collection network in wind energy turbine set, be connected to the low-pressure side of offshore boosting station, by offshore boosting station, the voltage that collects of 35kV is elevated to 220kV, then by the subsea cable access Infinite bus system of 20km;

The power quality data of having chosen the marine wind electric field that in two kinds of situations, electric pressure is 220kV is as evaluation object: 1. in the time that wind speed situation is different, and the power quality data at offshore grid-connected wind farm point place; 2. in the time that the generator installed capacity of marine wind electric field is different, and the power quality data at site place, totally three groups of emulated datas: a, wind speed steadily, when installed capacity 50MW; B, fluctuations in wind speed are large, when installed capacity 50MW; C, fluctuations in wind speed are large, when installed capacity 75MW;

6 power quality index data are all to adopt the large value of 95% probability, the large value of 95% probability is that all measured values are arranged from big to small, give up front 5% and get remaining maximal value, obtain the corresponding large value of 95% probability according to the power quality index data in three kinds of situations of following marine wind electric field

3. method according to claim 1, is characterized in that, the method for the definite power quality index weight described in step (2) is:

1) Set Pair Analysis is with Pair Analysis model, the certainty and uncertainty of things to be processed as a system, to the characteristic spread analysis of a set antithetical phrase, sets up these two the I. D.C connexion formula μ that are integrated under relevant issues background

μ＝a+bi+cj (1)

Wherein, a, b, c is respectively institute's analects and is combined in identical degree, diversity factor, the opposition degree under relevant issues background, and i is diversity factor coefficient, is defined in [1,1] in interval, look different situation value, j is opposition degree coefficient, is defined as-1, and a and c determine relatively, b is relatively uncertain, because of changeability, the complicacy of objective objects, be familiar with ambiguity and subjectivity when portraying objective things, can cause this relatively uncertain;

Be provided with r position expert, index set X={X _k, k=1,2 ..., n, every expert compares index between two, Judgement Matricies M _zkl, z=1,2 ..., r, k=1,2 ..., n, l=1,2 ..., n, is shown below, and represents the view of z position expert to the relative important relationship between any two indexs, x _z23represent that the 2nd index compare with the 3rd index, the 3rd important degree of index of the 2nd ratio;

Wherein

Utilize the matrix form of Pair Analysis to set up the Pair Analysis model μ of description indexes relative importance relation _qkl, that is:

Wherein,

A _klfor describing the understanding homogeneity matrix of expert to the relative importance between each index; B _klfor describing the cognitive diversity matrix of expert to the relative importance between each index.For research object herein, the situation of opposition does not have, therefore j=0;

A _klwhat represent is the common understandings of different experts to evaluation index relative importance relation, is complete relation that can be definite; To x _zkl≤ 1 and x _zkl>=1 situation is value respectively, represents different experts' consensus of opinion, works as x _zlk≤ 1 o'clock, a _lk=max{x _zlk, because so work as x _zkl>=1 o'clock, X _kwith X _lrelation at least will to meet scale value be min{x _zklcorresponding situation, the trend of its variation is scale value y > min{x _zklcorresponding situation but be no more than scale value y=max{x _zklcorresponding state.Obtain thus, adopt tectonic link degree μ in this way _qklin homogeneity matrix A _kl, this matrix makes to represent index X _lwith X _krelation two coefficient correspondences get up;

2) consistent matrix method is processed homogeneity matrix A _kl, obtain basis matrix D _kl, wherein D _kl=(d _kl), d _klbasis matrix D _klelement, and d satisfies condition _kk=1 He homogeneity matrix A _klcorresponding index weights ω _kcalculated by following formula:

Wherein,

3) entropy power method is processed otherness matrix B _kl, obtain B _klcorresponding otherness weights, these weights have reflected the different opinions of brainstrust, concrete steps are as follows:

A) according to the entropy of the each evaluation index of concept definition of traditional entropy be:

In formula: b _klit is otherness matrix B _klelement;

Obviously work as f _kl=0 o'clock, ln f _klmeaningless, therefore to f _klcalculating revised, be defined as

B) entropy that calculates indices is weighed

And meet

When getting while determining i value, obtain determining evaluation index weight, work as n=6, try to achieve index weights W by formula (6) and formula (10);

4) the concrete computation process of weight:

By 9 experts, 6 of the wind energy turbine set quality of power supply indexs are compared between two, use nine scale standards in analytical hierarchy process, obtain evaluating matrix group M _kl=(M _1kl, M _2kl..., M _9kl), evaluate power quality index in matrix and arrange in the following order: supply voltage deviation I ₁/ %, voltage fluctuation I ₂/ %, Short Term Flicker I ₃, percent harmonic distortion I ₄/ %, three-phase imbalance I ₅/ % and frequency departure I ₆/ Hz, provides two expert opinion matrixes below as example:

Obtain homogeneity matrix by formula (4):

Obtain otherness matrix by formula (5):

Adopt consistent matrix method to process homogeneity matrix A _kl:

According to formula (7), we obtain:

C _s(6×1)＝[0.6608,1.1224,1.1224,0.8908,0.6608,2.0395] ^T

Obtain homogeneity matrix A by formula (6) _klcorresponding weight:

W _k(6×1)＝[ω ₁,ω ₂,ω ₃,ω ₄,ω ₅,ω ₆] ^T＝(0.1017,0.1728,0.1728,0.1371,0.1017,0.3139) ^T (15)

Obtain otherness matrix B by formula (8)～formula (10) _klcorresponding weight:

η _(1×6)＝[η ₁,η ₂,η ₃,η ₄,η ₅,η ₆] ^T＝[0.0701,0.1265,0.1697,0.4379,0.0692,0.1265] ^T (16)

Get i=0.08, by formula (15) and formula (16) must power quality index weight be:

W＝[0.0994,0.1694,0.1726,0.1594,0.0993,0.3000] ^T (17)。

4. method according to claim 1, is characterized in that, described step (3) is:

Adopt the quality of power supply grade classification of the each single index under 21 grades of linguistic variables and the 220kV electric pressure of Vague value representation that the power quality index data value table of each evaluation object is shown as to Vague value; By 21 grades of linguistic variables of Vague value representation be:

Grade Vague value Abstention situation AG [1,1] 0 AG ^- [0.95,0.975] 0.025 VG [0.9,0.95] 0.05 VG ^- [0.85,0.925] 0.075 G [0.8,0.9] 0.1 G ^- [0.75,0.875] 0.125 FG [0.7,0.85] 0.15 FG ^- [0.65,0.825] 0.175 MG [0.6,0.8] 0.2 MG ^- [0.55,0.75] 0.2 M [0.5,0.5] 0 M ^- [0.45,0.65] 0.2 MP [0.4,0.6] 0.2 MP ^- [0.35,0.525] 0.175 FP [0.3,0.45] 0.15 FP ^- [0.25,0.375] 0.125 P [0.2,0.3] 0.1 P ^- [0.15,0.225] 0.075 VP [0.1,0.15] 0.05 VP ^- [0.05,0.075] 0.025 AP [0,0] 0

The quality of power supply grade classification of the each single index under 220kV electric pressure is:

5. method according to claim 1, is characterized in that, described step (4) is: form respectively optimal value vector Z by indices optimal value and the most bad value ⁺the most bad value vector Z ^-, i.e. definitely positive ideal solution and definitely negative ideal solution

Wherein,

In the time only considering the weight of index, consider using the ideal value in GB and limit value as definitely positive ideal solution and definitely negative ideal solution.

6. method according to claim 1, is characterized in that, described step (5) is:

If U is a nonempty set, its element represents with x, and a Vague collection A on U refers to a pair of subordinate function t on U _aand f _a,

t _A：U→[0,1]，f _A：U→[0,1]

Meet t _a(x)+f _a(x)≤1, and 0≤t _a(x)≤1,0≤f _a(x)≤1, wherein: t _afor the true subordinate function of Vague collection A, express support for the degree of membership lower bound of the evidence of x ∈ A; f _afor the false subordinate function of Vague collection A, the degree of membership lower bound of the evidence of the x ∈ A that makes difficulties; If x ∈ is U, claim closed interval [t _a(x), 1-f _a(x)] be Vague collection A in the Vague of x value,

Two kinds of Vague value Similarity Measures,

1) Similarity Measures between improved Vague value is as follows:

M _zthe value of (x, y) is larger, represents that Vague value x is more similar with y,

2) improved weighting Vague value Similarity Measures is as follows:

If x, y is two Vague values, and the weight of x and y is all w, wherein, w=(a, b, c), a represents the weight of true subordinate function part, b represents the weight of false subordinate function part, c represents the weight of unknown portions, considers to affect three factors of Vague value similarity measure simultaneously, and improved weighting Vague value Similarity Measures is:

In formula (20), a, b, c>=0, and a+b+c > 0, value larger, represent that Vague value x is more similar with y.Work as a=1, b=1, when c=2, formula (20) is exactly (19);

Calculating each evaluation object with the distance method of definitely positive ideal solution and definitely negative ideal solution is:

1. only consider the weight of index, the similarity measure formula of Vague value is formula (19)

2. not only consider the weight of index, consider the weight of the true subordinate function of each index, false subordinate function, abstention part simultaneously

7. an application that approaches the energy quality comprehensive assessment method of ideal solution based on Vague collection and improvement as claimed in claim 1, it is characterized in that, be applied to the power quality controlling of marine wind electric field: parallel connection type active electric filter APF, improves the harmonic wave of marine wind electric field; Static Var Compensator SVC, STATCOM STATCOM, compensate the reactive power of marine wind electric field.