CN104091066A - Condition evaluation method for high-voltage circuit breaker - Google Patents

Abstract

A condition evaluation method for a high-voltage circuit breaker comprises the steps of establishing a hierarchical level evaluation model of the high-voltage circuit breaker, establishing judgment grade sets for various factors, determining weight sets of the various factors, establishing a weight set matrix according to the model parts and the grey parts of the weight sets, establishing the grey fuzzy judgment matrixes of all hierarchical levels, synthesizing a grey fuzzy comprehensive judgment result under a parent layer, and obtaining the operation state of the high-voltage circuit breaker. The condition evaluation method for the high-voltage circuit breaker has the advantages that the judgment result is more effectively and reliably close to the practical operation state, and the system problems of being fuzzy and incomplete in information can be more effectively solved.

Description

A kind of state evaluating method of primary cut-out

Technical field

The present invention relates to a kind of state evaluating method of primary cut-out.

Background technology

Primary cut-out plays protection and control double action in electrical network; its protective effect is mainly reflected in the time of power circuit or device fails; by coordinating with protective relaying device; by faulty component excision fast from electrical network; ensure the normal operation of non-fault part in electrical network; its control action is mainly reflected in the time that electrical network normally moves, and according to operation of power networks demand, uses primary cut-out a part of power equipment or circuit input or out of service.Therefore,, as one of core switching equipment of electric system, the quality of its operation conditions has great impact to the safety and stablization operation of electric system.At present, to the maintenance of isolating switch also in the prophylactic repair stage, except maintenance amount is large, economic cost high, also easily cause maintenance deficiency or maintenance excessively, therefore, the State Maintenance that isolating switch is carried out to " working as Xiu Zexiu " is significant, is the key prerequisite of implementing State Maintenance and isolating switch is carried out to state estimation accurately.

At present, Chinese scholars has proposed some different high-voltage circuit-breaker status appraisal procedures, comprise fuzzy comprehensive evaluation method and improve one's methods, matter-element and Method of Evidence Theory and artificial neural network etc., wherein most methods is based on Fuzzy comprehensive evaluation theory.Though matter-element and Evidence theory model are obtained certain achievement, it is only the different evaluation indexes that matter-element theory and evidence theory are applied to respectively to primary cut-out, does not also consider the grey of evaluation index.Although it is more extensive that artificial neural network is applied aspect data prediction, but it needs a large amount of data as basis, and primary cut-out data itself are relatively deficient, and disappearance is serious, therefore artificial neural network be not suitable for the state estimation of primary cut-out.Although Fuzzy comprehensive evaluation and improve one's methods and can process well complicated fuzzy system, it does not consider the grey of assessment factor, thereby makes the confidence level of assessment result lower.

Summary of the invention

Object of the present invention is exactly in order to address the above problem, a kind of state evaluating method of primary cut-out has been proposed, the method is described the relation between assessment factor and evaluation object by fuzzy relation, and introduce some gray scale and describe the credibility of corresponding fuzzy relation, can effectively solve system problem fuzzy, that information is incomplete.

For achieving the above object, the present invention adopts following technical scheme:

A state evaluating method for primary cut-out, is characterized in that, comprises the following steps:

(1) according to the mechanical property, electrical specification, insulation characterisitic and other characteristic factors that affect high-voltage circuit-breaker status, choose the representative index under various factors, set up primary cut-out Recurison order hierarchy assessment models;

(2) set up respectively the evaluation rank collection of various factors: according to the complexity of passing judgment on precision and computing, the running status under primary cut-out various factors is divided into 4 grades, be respectively excellent, good, in, poor, the corresponding V={V that gathers ₁, V ₂, V ₃, V ₄;

(3) determine the weight sets of various assessment factors, according to the mould portion of described weight sets and grey cage structure weight sets matrix;

(4) construct the grey fuzzy discrimination matrix of each Recurison order hierarchy: described grey fuzzy discrimination matrix comprises fuzzy part and grey color part; Fuzzy part characterizes the fuzzy relation between each assessment factor and evaluation status with fuzzy membership, and grey color part characterizes the confidence level of corresponding fuzzy membership with a gray scale;

(5) the Grey Fuzzy Comprehensive Evaluation result under synthetic this father's level; By forming the grey fuzzy discrimination matrix of high-order level assessment factor with the comprehensive evaluation result of the assessment factor of this father's level same level, the comprehensive evaluation result of primary cut-out running status is tried to achieve in successively computing, draws the running status of primary cut-out.

Described step (1) mesohigh isolating switch Recurison order hierarchy assessment models is divided into destination layer, rule layer, sub-rule layer and indicator layer,

Described destination layer is primary cut-out running status;

Described rule layer comprises: mechanical property, electrical specification, insulation characterisitic and other characteristics;

It between each level, is the relation of subdivision step by step.

In described step (3), the mould portion of weight sets determines by analytical hierarchy process; Under the identical prerequisite of the weight sets ash portion of supposing the each assessment factor under same father's level, the method that the grey portion of weight sets adopts expert's marking to ask for average is determined.

In described step (3), definite method of weight sets ash portion is:

To n sub-assessment factor a ₁, a ₂..., a _n, m expert comprehensively gives a mark respectively, and corresponding score value is h ₁, h ₂..., h _m; Remove a maximal value and a minimum value, then average, corresponding to the sub-assessment factor a under this father's level ₁, a ₂..., a _nweight sets ash portion be:

$v_i=\frac{\underset{j=1}{\overset{m}{\mathrm{\Σ}}}{h}_j-\max \left(h_j\right)-\min \left(h_j\right)}{m-2}---\left(1\right)$

I ∈ in formula 1,2,3 ..., l}, l represents the assessment factor number with this father's level same level.

In described step (3), weight sets matrix is:

$\underset{{\⊗}_i}{\overset{~}{W}}=[(w_1,v_i),(w_2,v_i),\·\·\·,(w_n,v_i)]---\left(2\right)$

Wherein, { w ₁, w ₂..., w _nto the assessment factor a under should father's level ₁, a ₂..., a _nthe mould portion of corresponding weight sets, v _ithe grey portion of respective weights collection.

In described step (4), definite method of fuzzy part is:

For the quantitative target of high-voltage circuit-breaker status, adopt respectively half trapezoidal profile function and trigonometric function as membership function, determine that according to pointer type corresponding membership function distributes:

For more little more excellent type index, evaluation rank V ₁membership function be formula (3), evaluation rank V ₂, V ₃membership function be formula (4), evaluation rank V ₄membership function be formula (5);

For more excellent more greatly type index, evaluation rank V ₁membership function be formula (5), evaluation rank V ₂, V ₃membership function be formula (4), evaluation rank V ₄membership function be formula (3);

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}1& x\&le;m_1\\ \frac{m_2-x}{m_2-m_1}& m_1<x\&le;m_2\\ 0& m_2<x\end{array}\right.---\left(3\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}\frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ \frac{m_3-x}{m_3-m_2}& m_2\&le;x<m_3\\ 0& x\&le;m_1,m_3\&le;x\end{array}\right.---\left(4\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}0& x\&le;m_1\\ \frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ 1& m_2<x\end{array}\right.---\left(5\right)$

For the qualitative index of high-voltage circuit-breaker status, adopt fuzzy statistics test method(s) to ask for its degree of membership, concrete grammar is: by expert, each assessment factor is rule of thumb judged, then result of determination is gathered to obtain to the corresponding grade frequency of each assessment factor, do normalized and obtain corresponding membership function, its formula is:

In described step (4), definite method of grey color part is:

By the next corresponding certain tonal range of certain descriptive language, carry out choosing value according to the abundant degree of actual information, determine corresponding gray-scale value.

In described step (4), under this father's level, the grey fuzzy discrimination matrix of sub-assessment factor is:

$\underset{{\&CircleTimes;}_i}{\overset{~}{R}}=\left[\begin{array}{cccc}({\mathrm{\&mu;}}_{11},v_{11})& ({\mathrm{\&mu;}}_{12},v_{12})& ({\mathrm{\&mu;}}_{13},v_{13})& ({\mathrm{\&mu;}}_{14},v_{14})\\ ({\mathrm{\&mu;}}_{21},v_{21})& ({\mathrm{\&mu;}}_{22},v_{22})& ({\mathrm{\&mu;}}_{23},v_{23})& ({\mathrm{\&mu;}}_{24},v_{24})\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ ({\mathrm{\&mu;}}_{n1},v_{n1})& ({\mathrm{\&mu;}}_{n2},v_{n2})& ({\mathrm{\&mu;}}_{n3},v_{n3})& ({\mathrm{\&mu;}}_{n4},v_{n4})\end{array}\right]---\left(7\right)$

Wherein, μ _ijrepresent assessment factor a under this father's layer _ithe degree of membership of phase In Grade j, corresponding v _ijrepresent the confidence level of this membership, i ∈ 1,2,3 ..., n}, j ∈ { 1,2,3,4}.

In described step (5), the result of Grey Fuzzy Comprehensive Evaluation is:

Wherein, j ∈ 1,2,3,4}, i ∈ 1,2 ..., l}, l represents the assessment factor number with this father's level same level, represent weight sets matrix; represent corresponding with it grey fuzzy discrimination matrix; w _k, v _ifor weight and the corresponding grey scale of each evaluation index; μ _kj, v _kjfor degree of membership value and the corresponding point gray-scale value of corresponding assessment factor.

The method of the described running status that draws primary cut-out is:

Suppose b _ibe i vector, make d _i=1-v _i, wherein, v _irepresent gray-scale value, d _irepresent b _iconfidence level; Make b _i=(μ _i, d _i), the comprehensive evaluation result of primary cut-out running status is reduced to:

$\left|\right|b_i\left|\right|=\sqrt{[b_i,b_i]}---\left(10\right)$

In formula, [b _i, b _i] be vectorial b _iinner product, utilize maximum membership grade principle to draw the running status of primary cut-out.

Principle of the present invention is as follows:

(1) grey fuzzy mathematics

If be space X=fuzzy subset on x}, if x for degree of membership μ _a(x) be a grey number on [0,1], its gray scale is v _a(x), claim for the gray fuzzy sets on X, be denoted as

$\underset{\&CircleTimes;}{\overset{~}{A}}=\left\{(x,{\mathrm{\&mu;}}_A\left(x\right),v_A\left(x\right))\right|x\&Element;X\}---\left(11\right)$

Given space X={ x}, Y={y}, if x and y are to fuzzy relation degree of membership be μ _r(x, y), the some gray scale v of its correspondence _r(x, y), claims set for the Theory of Grey Fuzzy Relation on direct product space X × Y, also can represent by the form of grey fuzzy matrix, as follows:

$\underset{\&CircleTimes;}{\overset{~}{R}}=\left[\begin{array}{cccc}({\mathrm{\&mu;}}_{11},v_{11})& ({\mathrm{\&mu;}}_{12},v_{12})& \&CenterDot;\&CenterDot;\&CenterDot;& ({\mathrm{\&mu;}}_{1n},v_{1n})\\ ({\mathrm{\&mu;}}_{21},v_{21})& ({\mathrm{\&mu;}}_{22},v_{22})& \&CenterDot;\&CenterDot;\&CenterDot;& ({\mathrm{\&mu;}}_{2n},v_{2n})\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ ({\mathrm{\&mu;}}_{m1},v_{m1})& ({\mathrm{\&mu;}}_{m2},v_{m2})& \&CenterDot;\&CenterDot;\&CenterDot;& ({\mathrm{\&mu;}}_{\mathrm{mn}},v_{\mathrm{mn}})\end{array}\right]={\left[({\mathrm{\&mu;}}_{\mathrm{ij}},v_{\mathrm{ij}})\right]}_{m\&times;n}--\left(12\right)$

Can also be expressed as with collection is even $\underset{\&CircleTimes;}{\overset{~}{R}}=(\tilde{R},\underset{\&CircleTimes;}{R}),$ Wherein, $\tilde{R}=\left\{((x,y),{\mathrm{\&mu;}}_R(x,y))\right|x\&Element;X,y\&Element;Y\}$ Fuzzy relation on representation space X × Y, represent the gray relation on direct product space X × Y.

Be provided with grey fuzzy matrix $\underset{\&CircleTimes;}{\overset{~}{W}}={\left[(w_{\mathrm{ij}}^W,v_{\mathrm{ij}}^W)\right]}_{m\&times;s}$ With $\underset{\&CircleTimes;}{\overset{~}{R}}={\left[({\mathrm{\&mu;}}_{\mathrm{ij}}^R,v_{\mathrm{ij}}^R)\right]}_{s\&times;n},$ ? with compositive relation be:

In formula: for fuzzy part; for grey color part; F ,+F are respectively mould portion computing broad sense "AND" and "or" operator; G ,+G are respectively grey portion computing broad sense "AND" and "or" operator.

(2) analytical hierarchy process (Analytical Hierarchy Process, AHP)

Analytical hierarchy process (Analytical Hierarchy Process, AHP) is a kind of qualitative analysis being combined with quantitative test for analyzing the system analysis method of multiple goal, multiple criteria, complication system, is widely used in a lot of fields.

First,, according to the general objective of problem, set up the Recurison order hierarchy model of evaluating system, in table 1.Secondly, by expert, evaluation index is compared between two structure discrimination matrix.Discrimination matrix represents for last layer factor, the relative importance between the associated element of this level, and it was determined by mutually comparing between two between element, adopted 1～9 scale to mark, mark implication is in table 3.Then,, to above-mentioned gained discrimination matrix, solve its maximum characteristic root λ _maxand corresponding characteristic vector W _i'=(w' ₁, w' ₂..., w' _n), required proper vector is each assessment factor importance ranking, i.e. weight allocation is carried out standardization and is obtained the weight sets mould W of portion _i=(w ₁, w ₂..., w _n).

Finally, carry out consistency check.In actual decision analysis, because complexity and the people of studied problem are familiar with upper issuable one-sidedness, make the discrimination matrix constructing often not there is consistance, therefore need to carry out consistency checking, can complete by consistance formula, as follows:

$\mathrm{CR}=\frac{\mathrm{CI}}{\mathrm{RI}}---\left(14\right)$

In formula: CR---the random Consistency Ratio of discrimination matrix;

The general coincident indicator of CI---discrimination matrix, obtained by following formula:

$\mathrm{CI}=\frac{1}{n-1}({\mathrm{\&lambda;}}_{\max }-n)---\left(15\right)$

The mean random coincident indicator of RI---discrimination matrix, for 1～9 rank discrimination matrix, the value of RI is listed in table 4.

In the time of CR<0.1, think that discrimination matrix has satisfied consistance, illustrate that weight allocation is rational; Otherwise need to adjust discrimination matrix, until obtain satisfied consistance.

Beneficial effect of the present invention:

Compare with traditional fuzzy comprehensive evoluation, the present invention has taken into full account that primary cut-out assessment factor has the feature of ambiguity and grey, by in the theoretical grey fuzzy comprehensive evaluation state estimation of introducing primary cut-out, make evaluation result closing to reality running status more effectively reliably, can effectively solve system problem fuzzy, that information is incomplete.

Brief description of the drawings:

Fig. 1 is grey fuzzy comprehensive evaluation process flow diagram;

Embodiment:

Below in conjunction with accompanying drawing and embodiment, the present invention is further detailed.

As shown in Figure 1, a kind of state evaluating method of primary cut-out, comprises the following steps:

Step 1: set up evaluation factor collection

Primary cut-out is as a complicated multifactor system, and factor mechanical, electric, insulation aspect is the principal element that affects circuit-breaker status, in addition, also has some factors to should give consideration, as working environment, state of appearance, its maintenance situation etc.Viewpoint at this based on systems engineering, considers operation mechanism and the correlation test project of isolating switch, chooses representative index, takes into account the rationality of modeling and comprehensive, comprehensively sets up primary cut-out scoring model, as shown in table 1.

Table 1 primary cut-out scoring model

Step 2: set up evaluation rank collection

Passing judgment on the grade of collection need to divide according to actual conditions, need consider the complexity of passing judgment on precision and computing.Primary cut-out running status is divided into 4 grades by this model, be respectively excellent, good, in, poor, corresponding set V={V ₁, V ₂, V ₃, V ₄.

Step 3: determine assessment factor weight sets

Weight sets can be considered the Theory of Grey Fuzzy Relation of passing judgment between object and evaluation factor, and the significance level of each assessment factor in net assessment reflected with the form quantizing in the mould portion of weight sets, and the confidence level of this assessment has been reflected in its grey portion.This model is by separately definite to the mould portion of weight sets and grey portion, and the mould portion of weight sets determines by the analytical hierarchy process with universality; Ensureing, under the prerequisite of Evaluation accuracy, to reduce operand, the weight sets ash portion of the assessment factor under the same father's level of this model assumption is identical as far as possible, and giving a mark by expert, to get the method for average definite.Suppose to there is n sub-assessment factor under certain father's level, be respectively a ₁, a ₂..., a _n.

(1) determining of weight sets mould portion

Analytical hierarchy process (Analytical Hierarchy Process, AHP) be a kind of qualitative analysis being combined with quantitative test for analyzing the system analysis method of multiple goal, multiple criteria, complication system, be widely used in a lot of fields, utilize the method to determine the weight sets mould W of portion _i=(w ₁, w ₂..., w _n).

(2) determining of weight sets ash portion

Being more or less the same based on the relatively abundant degree of assessment factor risk data under same level, is simplified operation amount, and under the same father's level of this model assumption, the weight sets of assessment factor has identical grey portion.In conjunction with expertise, the grey portion of weight sets adopts expert's marking method of averaging to determine, marking standard is in table 2,

Table 2 gray scale marking standard

Concrete grammar is as follows:

To n sub-assessment factor a ₁, a ₂..., a _n, m expert comprehensively gives a mark respectively, and corresponding score value is h ₁, h ₂..., h _m.In order to alleviate expert's subjective bias, remove a maximal value and a minimum value, then average, corresponding to the sub-assessment factor a under this father's level ₁, a ₂..., a _nweight sets ash portion be:

$v_i=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{h}_j-\max \left(h_j\right)-\min \left(h_j\right)}{m-2}---\left(1\right)$

I ∈ in formula 1,2,3 ..., n}, the assessment factor number of l representative and this father's level same level.

To sum up, structure weight sets matrix, the assessment factor a under this father's level ₁, a ₂..., a _ncorresponding weight sets mould portion is { w ₁, w ₂..., w _n, weight sets ash portion is v _i, the matrix representation forms of weight sets is as follows:

$\underset{{\&CircleTimes;}_i}{\overset{~}{W}}=[(w_1,v_i),(w_2,v_i),\&CenterDot;\&CenterDot;\&CenterDot;,(w_n,v_i)]---\left(2\right)$

Step 4: determine grey fuzzy discrimination matrix

Grey fuzzy discrimination matrix is made up of fuzzy part and grey color part.Fuzzy part characterizes the fuzzy relation between each assessment factor and evaluation status with fuzzy membership, and grey color part characterizes the confidence level of corresponding fuzzy membership with a gray scale.The concrete grammar of determining grey fuzzy discrimination matrix is as follows:

(1) determining of fuzzy part

In high-voltage circuit-breaker status index, comprise quantitative target and qualitative index, for quantitative target, the triangular membership functions that this model selection is simple and applicability is strong is determined degree of membership; For qualitative index, take fuzzy statistics test method(s) to ask for degree of membership herein.

For quantitative target, consider computational accuracy and operand, first raw data associated is done to corresponding pre-service herein.Then, half trapezoidal profile function of employing widespread use and trigonometric function, as membership function, determine that according to pointer type corresponding membership function distributes.For more little more excellent type index, evaluation rank V ₁distribution function adopt type less than normal half trapezoidal profile (formula 3), evaluation rank V fall ₂, V ₃distribution function adopt osculant triangle distribution (formula 4) without loss of generality, evaluation rank V ₄distribution function adopt liter half trapezoidal profile (formula 5) of type bigger than normal.For more excellent more greatly type index, evaluation rank V ₁distribution function adopt liter half trapezoidal profile (formula 5) of type bigger than normal, evaluation rank V ₂, V ₃distribution function still adopt osculant triangle distribution (formula 4) without loss of generality, evaluation rank V ₄distribution function adopt type less than normal half trapezoidal profile (formula 3) falls.Finally, determine according to correlation test specification and expertise the parameter m comprising in the corresponding distribution of different evaluation ranks ₁, m ₂, m ₃, and will bring corresponding membership function into through pretreated data and ask for fuzzy membership μ _ij.

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}1& x\&le;m_1\\ \frac{m_2-x}{m_2-m_1}& m_1<x\&le;m_2\\ 0& m_2<x\end{array}\right.---\left(3\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}\frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ \frac{m_3-x}{m_3-m_2}& m_2\&le;x<m_3\\ 0& x\&le;m_1,m_3\&le;x\end{array}\right.---\left(4\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}0& x\&le;m_1\\ \frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ 1& m_2<x\end{array}\right.---\left(5\right)$

For the qualitative index that is difficult to quantize, this model adopts fuzzy statistics test method(s) to ask for its degree of membership, specific practice is: make expert's questionnaire of giving a mark, by expert, each assessment factor is rule of thumb judged, corresponding grade place at marking table ticks, then gather to obtain the corresponding grade frequency of each assessment factor, do normalized and obtain corresponding membership function, its formula is:

(2) determining of grey color part

Determining when fuzzy part, the quantity of information difference that each assessment factor is collected, can cause definite fuzzy relation to exist can not reliability, and has certain difference because the abundant degree of information is different.Consider its impact on net assessment, in Theory of Grey Fuzzy Relation matrix, introduce grey color part, and by the next corresponding certain tonal range of certain descriptive language, as shown in table 2.Determining of ash color part need to be carried out choosing value according to the abundant degree of actual information, and the thinking that adopts expert to give a mark herein to average for each assessment factor is determined corresponding gray-scale value.

To sum up, the Theory of Grey Fuzzy Relation matrix of determining sub-assessment factor under this father's level is:

$\underset{{\&CircleTimes;}_i}{\overset{~}{R}}=\left[\begin{array}{cccc}({\mathrm{\&mu;}}_{11},v_{11})& ({\mathrm{\&mu;}}_{12},v_{12})& ({\mathrm{\&mu;}}_{13},v_{13})& ({\mathrm{\&mu;}}_{14},v_{14})\\ ({\mathrm{\&mu;}}_{21},v_{21})& ({\mathrm{\&mu;}}_{22},v_{22})& ({\mathrm{\&mu;}}_{23},v_{23})& ({\mathrm{\&mu;}}_{24},v_{24})\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ ({\mathrm{\&mu;}}_{n1},v_{n1})& ({\mathrm{\&mu;}}_{n2},v_{n2})& ({\mathrm{\&mu;}}_{n3},v_{n3})& ({\mathrm{\&mu;}}_{n4},v_{n4})\end{array}\right]---\left(7\right)$

Step 5: Grey Fuzzy Comprehensive Evaluation

The state estimation of primary cut-out is the analysis to circuit-breaker status variation tendency, and state estimation factor in fact not can accurately be grasped.In order to retain judge information as much as possible, the computing employing M of mould portion (,+) and operator, the computing employing M of grey portion (⊙ ,+) operator, the result of synthetic Grey Fuzzy Comprehensive Evaluation is:

J ∈ in formula 1,2,3,4}, i ∈ 1,2 ..., l}, the assessment factor number of l representative and this father's level same level, represent weight sets matrix; represent corresponding with it grey fuzzy discrimination matrix; w _k, v _ifor weight and the corresponding grey scale of each evaluation index; μ _kj, v _kjfor degree of membership value and the corresponding point gray-scale value of corresponding assessment factor.

By with the comprehensive evaluation result of the assessment factor of this father's level same level the grey fuzzy discrimination matrix of composition high-order level assessment factor, shown in (9).Carry out grey fuzzy computing in conjunction with respective weights collection, by that analogy, the comprehensive evaluation result of primary cut-out running status is tried to achieve in successively computing

To the processing of evaluation result, this model is quoted the method that Law of Inner Product and maximum membership grade principle combine and is processed.Suppose b _ibe i vector, make d _i=1-v _i, wherein v _irepresent gray-scale value, d _irepresent b _iconfidence level.Make b _i=(μ _i, d _i), comprehensively pass judgment on can be by calculating b _isize determine, be reduced to and solve b _inorm, as follows.

$\left|\right|b_i\left|\right|=\sqrt{[b_i,b_i]}---\left(10\right)$

[b in formula _i, b _i] be vectorial b _iinner product.Finally, utilize maximum membership grade principle to draw the running status of primary cut-out.

In order to verify validity and the feasibility of the application in high-voltage circuit-breaker status assessment of this algorithm, herein to certain SF ₆isolating switch has carried out grey fuzzy comprehensive evaluation.This isolating switch important technological parameters is as shown in table 3, table 4.

Table 3 3AP1FG SF ₆isolating switch important technological parameters

Table 4 SF ₆circuit breaker gas tensimeter (20 DEG C)

Splitting or integrating lock by preventive trial gained is not respectively 1.65ms, 1.73ms the same period; Just-off speed is respectively 2.4m/s, 2.1m/s with firm resultant velocity; Point, combined floodgate direct current resistance is respectively 90 Ω, 125 Ω.This isolating switch puts into operation 13 years, and accumulative total open and close times is 472 times, point, closing coil minimum voltage action is respectively 63V, 70V; Major loop resistance is 27 μ Ω.In insulating medium, gaseous tension is 0.55MPa, and humidity is 163ppm, and insulation against ground resistance in primary circuit is 8000M Ω.

As shown in Table 1, the main factor of considering four aspects herein: mechanical property, electrical specification, insulation characterisitic and other factors (environment, outward appearance etc.), concrete proof procedure is as follows:

Step 1: determining of the sub-factorial power sets of mechanical property and grey fuzzy discrimination matrix, describes as an example of time parameter example.

(1) determining of time parameter weight sets

First, determine its mould portion, the time parameter under mechanical property comprises the not same period and other factors of the same period, separating brake of closing a floodgate, and rule of thumb carries out relatively obtaining between two discrimination matrix [1 1/2 3 by expert; 215; 1/31/5 1], try to achieve maximum characteristic root λ _max=3.0037, and corresponding characteristic vector W _i'=(0.463,0.871,0.164), carry out standardization and obtain W _i=(0.309,0.582,0.109), further can obtain the random Consistency Ratio CR=0.0032<0.1 of discrimination matrix, has satisfied consistance.So, the time parameter weight sets mould W of portion under mechanical property _i=(0.309,0.582,0.109); Secondly, for the confidence level of its mould portion, 7 experts rule of thumb give a mark respectively, and { 0.2,0.4,0.3,0.3,0.1,0.3,0.5} tries to achieve the grey V of portion based on expert's marking method of averaging to obtain marking value _i=0.3; To sum up in like manner can obtain the weight sets of speed parameter and divide-shut brake coil direct current resistance, in table 5.

The weight sets of the sub-factor of table 5 mechanical property

(2) determining of time parameter grey fuzzy discrimination matrix

First, determine its fuzzy part, for closing a floodgate, separating brake not same period, adopt triangular membership functions to ask degree of membership.Because combined floodgate, the different issue data bulk of separating brake level are 1, meet computational accuracy, directly adopt raw data at this, i.e. pre-service function f (x)=x.Belong to more excellent more greatly type index according to it, select corresponding distribution function, and determine according to correlation test specification and expertise the parameter m comprising in the corresponding distribution of different evaluation ranks ₁, m ₂, m ₃, must close a floodgate, separating brake not the same period membership function, as shown in table 6,

Table 6 closes a floodgate, separating brake not the same period subordinate function

And given data is brought into and tried to achieve corresponding degree of membership.For other factors, adopt fuzzy statistics test method(s) to ask for its degree of membership for [0.3 0.5 0.2 0].To sum up trying to achieve its fuzzy part is $\left[\begin{array}{cccc}0.35& 0.65& 0.325& 0.217\\ 0& 0.54& 0.73& 0.487\\ 0.3& 0.5& 0.2& 0\end{array}\right].$

Secondly, determine its grey color part.Do not belong to excellent the same period for closing a floodgate under time parameter, 7 experts, according to abundant degree and the experience of data, give a mark to such an extent that { 0.6,0.5,0.4,0.4,0.4,0.2,0.3}, trying to achieve corresponding point gray-scale value is 0.4.In like manner, try to achieve the corresponding some gray-scale value that is subordinate to grade of each index factor.

To sum up obtain the grey fuzzy discrimination matrix of time parameter, in table 7.

The sub-factors distinguishing matrix of table 7 mechanical property

In like manner can obtain the grey fuzzy discrimination matrix of speed parameter and divide-shut brake coil direct current resistance.

Utilize grey fuzzy comprehensive evaluation to carry out correlator assessment factor and comprehensively pass judgment on, as follows by formula (8) gained assessment result:

Step 2: formed the grey fuzzy discrimination matrix of mechanical property by the sub-combined factors evaluation result that calculates gained in step 1, as follows:

In like manner can obtain the sub-assessment factor weight sets of the High Voltage Circuit Breaker Condition determine as shown in step 1, the weight sets of the sub-assessment factor of the High Voltage Circuit Breaker Condition and discrimination matrix are as shown in table 8, table 9.

The sub-factorial power sets of table 8 the High Voltage Circuit Breaker Condition

The sub-factors distinguishing matrix of table 9 the High Voltage Circuit Breaker Condition

Dependent evaluation factor is carried out to grey fuzzy comprehensive evaluation, and assessment result is as follows:

Step 3: primary cut-out running status is carried out to Grey Fuzzy Comprehensive Evaluation, and evaluation result is processed.

The grey fuzzy discrimination matrix that is made up of primary cut-out running status the sub-combined factors evaluation result that calculates gained in step 2 is as follows:

In conjunction with carry out grey fuzzy comprehensive evaluation, try to achieve grey fuzzy evaluation result and be

The Grey Fuzzy Comprehensive Evaluation result obtaining is got to norm to be obtained

||b ₁||＝0.9879,||b ₂||＝1.0405,||b ₃||＝1.0475,||b ₄||＝1.0139。In conjunction with maximum membership grade principle, the running status of this primary cut-out of final decision be " in ".Can be found out by actual operating data, some numerical value has departed from factory-said value or optimal value, has deteriorated trend, and actual operating state is: gear train bite, causes closing speed reduce or refuse to close.If only judge according to degree of membership, this High Voltage Circuit Breaker Condition belongs to " good ", and more biased with the existence of actual motion state, reason is that the corresponding gray scale of this degree of membership is larger, and the confidence level of this degree of membership is lower.As can be seen here, the conclusion drawing based on grey fuzzy comprehensive evaluation model is more effectively credible.

By reference to the accompanying drawings the specific embodiment of the present invention is described although above-mentioned; but not limiting the scope of the invention; one of ordinary skill in the art should be understood that; on the basis of technical scheme of the present invention, those skilled in the art do not need to pay various amendments that creative work can make or distortion still in protection scope of the present invention.

Claims (10)

1. a state evaluating method for primary cut-out, is characterized in that, comprises the following steps:

(1) according to the mechanical property, electrical specification, insulation characterisitic and other characteristic factors that affect high-voltage circuit-breaker status, choose the representative index under various factors, set up primary cut-out Recurison order hierarchy assessment models;

(2) set up respectively the evaluation rank collection of various factors: according to the complexity of passing judgment on precision and computing, the running status under primary cut-out various factors is divided into 4 grades, be respectively excellent, good, in, poor, the corresponding V={V that gathers ₁, V ₂, V ₃, V ₄;

(3) determine the weight sets of various assessment factors, according to the mould portion of described weight sets and grey cage structure weight sets matrix;

(4) construct the grey fuzzy discrimination matrix of each Recurison order hierarchy: described grey fuzzy discrimination matrix comprises fuzzy part and grey color part; Fuzzy part characterizes the fuzzy relation between each assessment factor and evaluation status with fuzzy membership, and grey color part characterizes the confidence level of corresponding fuzzy membership with a gray scale;

(5) the Grey Fuzzy Comprehensive Evaluation result under synthetic this father's level; By forming the grey fuzzy discrimination matrix of high-order level assessment factor with the comprehensive evaluation result of the assessment factor of this father's level same level, the comprehensive evaluation result of primary cut-out running status is tried to achieve in successively computing, draws the running status of primary cut-out.

2. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, described step (1) mesohigh isolating switch Recurison order hierarchy assessment models is divided into destination layer, rule layer, sub-rule layer and indicator layer,

Described destination layer is primary cut-out running status;

Described rule layer comprises: mechanical property, electrical specification, insulation characterisitic and other characteristics;

It between each level, is the relation of subdivision step by step.

3. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (3), the mould portion of weight sets determines by analytical hierarchy process; Under the identical prerequisite of the weight sets ash portion of supposing the each assessment factor under same father's level, the method that the grey portion of weight sets adopts expert's marking to ask for average is determined.

4. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (3), definite method of weight sets ash portion is:

To n sub-assessment factor a ₁, a ₂..., a _n, m expert comprehensively gives a mark respectively, and corresponding score value is h ₁, h ₂..., h _m; Remove a maximal value and a minimum value, then average, corresponding to the sub-assessment factor a under this father's level ₁, a ₂..., a _nweight sets ash portion be:

$v_i=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{h}_j-\max \left(h_j\right)-\min \left(h_j\right)}{m-2}---\left(1\right)$

I ∈ in formula 1,2,3 ..., l}, l represents the assessment factor number with this father's level same level.

5. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (3), weight sets matrix is:

$\underset{{\&CircleTimes;}_i}{\overset{~}{W}}=[(w_1,v_i),(w_2,v_i),\&CenterDot;\&CenterDot;\&CenterDot;,(w_n,v_i)]---\left(2\right)$

Wherein, { w ₁, w ₂..., w _nto the assessment factor a under should father's level ₁, a ₂..., a _nthe mould portion of corresponding weight sets, v _ithe grey portion of respective weights collection.

6. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (4), definite method of fuzzy part is:

For the quantitative target of high-voltage circuit-breaker status, adopt respectively half trapezoidal profile function and trigonometric function as membership function, determine that according to pointer type corresponding membership function distributes:

For more little more excellent type index, evaluation rank V ₁membership function be formula (3), evaluation rank V ₂, V ₃membership function be formula (4), evaluation rank V ₄membership function be formula (5);

For more excellent more greatly type index, evaluation rank V ₁membership function be formula (5), evaluation rank V ₂, V ₃membership function be formula (4), evaluation rank V ₄membership function be formula (3);

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}1& x\&le;m_1\\ \frac{m_2-x}{m_2-m_1}& m_1<x\&le;m_2\\ 0& m_2<x\end{array}\right.---\left(3\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}\frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ \frac{m_3-x}{m_3-m_2}& m_2\&le;x<m_3\\ 0& x\&le;m_1,m_3\&le;x\end{array}\right.---\left(4\right)$

$\mathrm{\&mu;}\left(x\right)=\left\{\begin{array}{cc}0& x\&le;m_1\\ \frac{x-m_1}{m_2-m_1}& m_1<x\&le;m_2\\ 1& m_2<x\end{array}\right.---\left(5\right)$

For the qualitative index of high-voltage circuit-breaker status, adopt fuzzy statistics test method(s) to ask for its degree of membership, concrete grammar is: by expert, each assessment factor is rule of thumb judged, then result of determination is gathered to obtain to the corresponding grade frequency of each assessment factor, do normalized and obtain corresponding membership function, its formula is:

7. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (4), definite method of grey color part is:

By the next corresponding certain tonal range of certain descriptive language, carry out choosing value according to the abundant degree of actual information, determine corresponding gray-scale value.

8. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (4), under this father's level, the grey fuzzy discrimination matrix of sub-assessment factor is:

$\underset{{\&CircleTimes;}_i}{\overset{~}{R}}=\left[\begin{array}{cccc}({\mathrm{\&mu;}}_{11},v_{11})& ({\mathrm{\&mu;}}_{12},v_{12})& ({\mathrm{\&mu;}}_{13},v_{13})& ({\mathrm{\&mu;}}_{14},v_{14})\\ ({\mathrm{\&mu;}}_{21},v_{21})& ({\mathrm{\&mu;}}_{22},v_{22})& ({\mathrm{\&mu;}}_{23},v_{23})& ({\mathrm{\&mu;}}_{24},v_{24})\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ ({\mathrm{\&mu;}}_{n1},v_{n1})& ({\mathrm{\&mu;}}_{n2},v_{n2})& ({\mathrm{\&mu;}}_{n3},v_{n3})& ({\mathrm{\&mu;}}_{n4},v_{n4})\end{array}\right]---\left(7\right)$

Wherein, μ _ijrepresent assessment factor a under this father's layer _ithe degree of membership of phase In Grade j, corresponding v _ijrepresent the confidence level of this membership, i ∈ 1,2,3 ..., n}, j ∈ { 1,2,3,4}.

9. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, in described step (5), the result of Grey Fuzzy Comprehensive Evaluation is:

Wherein, j ∈ 1,2,3,4}, i ∈ 1,2 ..., l}, l represents the assessment factor number with this father's level same level, represent weight sets matrix; represent corresponding with it grey fuzzy discrimination matrix; w _k, v _ifor weight and the corresponding grey scale of each evaluation index; μ _kj, v _kjfor degree of membership value and the corresponding point gray-scale value of corresponding assessment factor.

10. the state evaluating method of a kind of primary cut-out as claimed in claim 1, is characterized in that, described in show that the method for the running status of primary cut-out is:

Suppose b _ibe i vector, make d _i=1-v _i, wherein, v _irepresent gray-scale value, d _irepresent b _iconfidence level; Make b _i=(μ _i, d _i), the comprehensive evaluation result of primary cut-out running status is reduced to:

$\left|\right|b_i\left|\right|=\sqrt{[b_i,b_i]}---\left(10\right)$

In formula, [b _i, b _i] be vectorial b _iinner product, utilize maximum membership grade principle to draw the running status of primary cut-out.