CN105117847A - Method for evaluating transformer failure importance - Google Patents

Abstract

The present invention discloses a method for evaluating transformer failure importance. The method comprises: firstly, according to various factors influencing failure mode importance, establishing an evaluation index set of transformer failure modes; secondly, calculating weight of each evaluation index in the evaluation index set of the transformer failure modes by using an entropy weight method; and finally, according to the calculated weight of the evaluation index, calculating failure importance of each failure mode of a power transformer. According to the method, based on sufficient consideration of various factors influencing failure mode importance, the evaluation index set of the failure modes is established, and the final failure importance is calculated by integrating each evaluation index and corresponding weight; the failure importance of each transformer failure mode can be accurately determined, thereby providing a scientific basis for making transformer maintenance and repair decisions.

Description

A kind of appraisal procedure of transformer fault importance degree

Technical field

The present invention relates to a kind of can the method for accurate evaluation Power Transformer Faults importance degree, belong to transformer technology field.

Background technology

Large-scale power transformer is as the hub device of electric system, and its operational reliability is related to the safety and stability of whole electric system.Transformer fault importance degree can reflect the size of certain fault mode to electric system entire effect, and guides design optimization and the repair and maintenance of system.Significance Analysis is carried out to Power Transformer Faults, the formulation that can be transformer maintenance decision-making provides scientific basis, human resources allocation optimization, for the operational reliability improving power transformer, and then ensures that the safety and stability tool of whole electric system is of great significance.

In prior art, do not run to transformer the many factors impacted from fault mode when analyzing the fault importance of the various fault modes of power transformer and judge to consider, the not accurate enough science of determination of fault importance, therefore there is a need to study further.

Summary of the invention

The object of the invention is to the drawback for prior art, a kind of appraisal procedure of transformer fault importance degree is provided, for transformer maintenance decision-making provides scientific basis.

Problem of the present invention realizes with following technical proposals:

An appraisal procedure for transformer fault importance degree, described method, first according to the various factors affecting fault mode importance degree, sets up the evaluation indice of transformer fault pattern; Then the evaluation index utilizing entropy assessment to calculate fault mode concentrates the weight of each evaluation index; The evaluation criterion weight that last basis calculates, tries to achieve the fault importance of often kind of fault mode of power transformer.

The appraisal procedure of above-mentioned transformer fault importance degree, said method comprising the steps of:

A. the evaluation indice of transformer fault pattern is set up

If the set participating in the fault mode evaluated is M=(M ₁m ₂m _n), evaluation indice is D=(D ₁d ₂d _m); Evaluate a kind of fault mode M _ian evaluation index D _jvalue be x _ij(i=1,2 ..., n; J=1,2 ..., m), form decision matrix X=(x based on this _ij) _{m × n};

B. the weight of each evaluation index in Calculation Estimation index set

1. standardization is carried out to decision matrix X, obtain normalized matrix V=(v _ij) _{m × n}, concrete grammar is:

All evaluation indexes are divided into two classes, are respectively more little more excellent type and more large more excellent type,

Evaluation index for more little more excellent type:

$v_{ij}=\frac{max\left(x_j\right)-x_{ij}}{max\left(x_j\right)-min\left(x_j\right)}$

Evaluation index for more large more excellent type:

$v_{ij}=\frac{x_{ij}-min\left(x_j\right)}{\max \left(x_j\right)-\min \left(x_j\right)}$

In formula, max (x _j), min (x _j) be respectively evaluation index D _jmaximal value and minimum value;

2. Calculation Estimation index D _jentropy e _j:

$e_j=-\frac{1}{ln\left(m\right)}\underset{i=1}{\overset{m}{\Σ}}{p}_{ij}ln\left(p_{ij}\right);$

Wherein work as p _ij=0 or p _ijwhen=1, p _ijln (p _ij)=0;

3. Calculation Estimation index D _jindex weights w _j:

$w_j=\frac{d_j}{\underset{k=1}{\overset{n}{\Σ}}{d}_k},j=1,2,...,m$

Wherein, d _j=1-e _j, d _krepresent the coefficient of variation of kth item evaluation index, d _jrepresent the coefficient of variation of jth item evaluation index;

C. the fault importance of often kind of fault mode of power transformer is calculated

1. weighted decision matrix R=(r is built _ij) _{m × n}, wherein, r _ij=w _jv _ij;

2. positive ideal solution is calculated with minus ideal result

3. fault mode M to be evaluated is calculated _iwith the distance of positive ideal solution and with the distance of minus ideal result

${\mathrm{Sd}}_i^{+}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{+}-r_{ij})}^2},i=1,2,...,m$

${\mathrm{Sd}}_i^{-}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{-}-r_{ij})}^2},i=1,2,...,m$

4. object M to be evaluated is calculated _iwith the relative similarity degree η of ideal solution _i:

${\mathrm{\η}}_i=\frac{{\mathrm{Sd}}_i^{-}}{{\mathrm{Sd}}_i^{+}+{\mathrm{Sd}}_i^{-}},i=1,2,...,m;$

By η _icharacterize transformer fault importance degree, η _ilarger, fault importance is larger, and vice versa.

The appraisal procedure of above-mentioned transformer fault importance degree, the evaluation index that the evaluation index of described transformer fault pattern is concentrated comprises fault rate, impact on systemic-function, maintenance cost, maintenance difficulty, detectability and loss and judges.

The present invention is on the basis taking into full account the various factors affecting fault mode importance degree, set up the evaluation indice of fault mode, and comprehensively each evaluation index and its respective weights calculate final fault importance, the method can determine the fault importance of the various fault modes of transformer exactly, for the formulation of transformer maintenance decision-making provides scientific basis.

Accompanying drawing explanation

Below in conjunction with accompanying drawing, the invention will be further described.

Fig. 1 is the transformer fault importance degree appraisal procedure process flow diagram of the embodiment of the present invention;

Fig. 2 is the transformer typical fault pattern diagram in the embodiment of the present invention.

In figure neutralization literary composition, each symbol inventory is: M=(M ₁m ₂m _n) be the set of fault mode, D=(D ₁d ₂d _m) be evaluation indice, X=(x _ij) _{m × n}for decision matrix, V=(v _ij) _{m × n}for normalized matrix, max (x _j), min (x _j) be respectively evaluation index D _jmaximal value and minimum value, e _jfor evaluation index D _jentropy, w _jfor evaluation index D _jindex weights, d _krepresent the coefficient of variation of kth item evaluation index, d _jrepresent the coefficient of variation of jth item evaluation index, R=(r _ij) _{m × n}for weighted decision matrix, for positive ideal solution, for minus ideal result, for M _iwith the distance of positive ideal solution, for M _iwith the distance of minus ideal result, η _ifor object M to be evaluated _iwith the relative similarity degree of ideal solution.

Embodiment

Below in conjunction with embodiment, the present invention is described in further detail.

Referring to Fig. 1, the present invention includes following steps:

S1, set up the evaluation indice of at least one fault mode of transformer, described evaluation index is concentrated and is comprised at least one evaluation index, and what each evaluation index evaluated that a kind of fault mode produces transformer respectively from Different factor aspect affects size.

By the Different factor of analyzing influence fault mode importance degree, set up importance degree assessment indicator system and evaluation indice, the assessment indicator system set up in the present embodiment is as table 1.

Table 1 importance degree assessment indicator system

The evaluation index of S2, calculating fault mode concentrates the index weights of each evaluation index.

If the object set participating in the described fault mode evaluated is M=(M ₁m ₂m _n), evaluation indice is D=(D ₁d ₂d _m), give 6 evaluation indexes in the present embodiment table 1, it is not intended to limit the present invention, and those skilled in the art can add other more evaluation indexes as the case may be.Evaluate a kind of fault mode and object M _ian evaluation index be D _j, D _jvalue be x _ij(i=1,2 ..., n; J=1,2 ..., m), standard one column data namely in corresponding table 1, forms decision matrix X=(x based on this _ij) _{m × n};

For eliminating each scalar dimension to the impact of program decisions, standardization being carried out to decision matrix X, forming normalized matrix V=(v _ij) _{m × n}, concrete grammar is:

All evaluation indexes are divided into two classes: the evaluation index of more little more excellent type and more large more excellent type;

To evaluation index D _jwhen carrying out standardization, process respectively by following process formula:

Evaluation index for more little more excellent type:

$v_{ij}=\frac{max\left(x_j\right)-x_{ij}}{max\left(x_j\right)-min\left(x_j\right)}$

Evaluation index for more large more excellent type:

$v_{ij}=\frac{x_{ij}-min\left(x_j\right)}{\max \left(x_j\right)-\min \left(x_j\right)}$

V in formula _ijbe the value after normalization; Max (x _j), min (x _j) be respectively evaluation index D _jmaximal value and minimum value.The more little more excellent type index of what is called refers to that the scoring of this index is lower, and the performance of system is more excellent, also namely less on the impact of transformer, and more large more excellent type index refers to that the scoring of this index is higher, and the performance of system is more excellent, also namely less on the impact of transformer.The evaluation index that in the present embodiment, table 1 provides is all more little more excellent type index.

After standardization completes, according to entropy assessment parameter weight, specific as follows:

According to following formula Calculation Estimation index D _jentropy e _j:

$e_j=-\frac{1}{ln\left(m\right)}\underset{i=1}{\overset{m}{\Σ}}{p}_{ij}ln\left(p_{ij}\right)---\left(1\right)$

Wherein work as p _ij=0 or p _ijwhen=1, p _ijln (p _ij)=0;

According to following formula Calculation Estimation index D _jindex weights w _j:

$w_j=\frac{d_j}{\underset{k=1}{\overset{n}{\Σ}}{d}_k},j=1,2,...,m---\left(2\right)$

Wherein, d _j=1-e _j, d _krepresent the coefficient of variation of kth item evaluation index, d _jrepresent the coefficient of variation of jth item evaluation index, d _jlarger, corresponding index weights is also larger.For a certain index D _j, v _ijdifference is less, e _jlarger.When being respectively evaluated the evaluation index D of object _jwhole equal time, e _j=e _max=1.By the definition of entropy, the evaluation index difference of object is larger, and the quantity of information of this index reflection is larger.Definition coefficient of variation d _j=1-e _j.

The evaluation index that S3, basis calculate concentrates the index weights of each evaluation index, calculates the fault importance of often kind of fault mode of power transformer.Concrete steps are as follows:

Build weighted decision matrix: be multiplied with normalized matrix V by the index weights of each evaluation index calculated, obtain weighted decision matrix R=(r _ij) _{m × n}: wherein, r _ij=w _jv _ij, w _jfor the index weights of each evaluation index;

Calculate positive ideal solution with minus ideal result

" more large more excellent type " and " more little more excellent type " the corresponding more large evaluation index of more excellent type and evaluation index of more little more excellent type respectively in above formula.

Calculate object M to be evaluated _iwith the distance of plus-minus ideal solutions:

${\mathrm{Sd}}_i^{+}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{+}-r_{ij})}^2},i=1,2,...,m$

${\mathrm{Sd}}_i^{-}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{-}-r_{ij})}^2},i=1,2,...,m$

Calculate object M to be evaluated _iwith the relative similarity degree of ideal solution:

Object M to be evaluated _iwith the relative similarity degree η of positive ideal solution _ibe expressed as:

${\mathrm{\η}}_i=\frac{{\mathrm{Sd}}_i^{-}}{{\mathrm{Sd}}_i^{+}+{\mathrm{Sd}}_i^{-}},i=1,2,...,m;$

According to η _icharacterize described fault importance, η _ilarger, required fault importance is larger, and vice versa.

η _ilarger, required fault importance is larger, and namely the entire effect that causes transformer of this kind of fault is more severe.

In the present invention, the relative Link Importance of each evaluation index is different, and the index weights of trying to achieve according to entropy assessment is equivalent to the relative Link Importance of evaluation index.The value of each evaluation index after normalized matrix V standardization and the scoring of fault, be weighted and form the impact that weighted decision matrix not only considers fault scoring, have also contemplated that the impact of index relative Link Importance, more science is comprehensive, and it is more accurate to calculate.

The fault importance of the often kind of fault mode calculated is carried out sorting and showing.Like this can in transformer life cycle management, the fault mode high for importance degree carries out important management and control, promotes the parallel-adder settle-out time of transformer.

Below in conjunction with a specific embodiment, the present invention will be described.

The typical fault pattern of transformer is referring to Fig. 2.Be described for bushing shell for transformer fault in the present embodiment, bushing shell for transformer fault mainly contains leakage of oil (M ₁), humidified insulation (M ₂), insulator flashover (M ₃), aging (M ₄), electric discharge (M ₅), leakage field heating (M ₆), puncture (M ₇) seven typical fault patterns.

According to aforementioned built assessment indicator system.So can form following decision matrix:

$X=\left[\begin{array}{ccccccc}& D_1& D_2& D_3& D_4& D_5& D_6\\ M_1& 0.4621& 0.4& 0.3& 0.3& 0.2& 1\\ M_2& 0.0313& 0.5& 0.4& 0.2& 0.5& 3\\ M_3& 0.0273& 0.5& 1& 0.6& 0.3& 4\\ M_4& 0.1128& 0.3& 1.2& 0.1& 0.6& 3\\ M_5& 0.1227& 0.6& 1.5& 0.5& 0.3& 5\\ M_6& 0.2124& 0.5& 1& 0.7& 0.6& 7\\ M_7& 0.0314& 0.7& 3& 0.8& 0.2& 10\end{array}\right];$

For eliminating each scalar dimension to the impact of program decisions, standardization is carried out to decision matrix X.Make its dimension unified, and then form normalized matrix V=(v _ij) _{m × n}:

As follows for more little more excellent type index calculate

$v_{ij}=\frac{max\left(x_j\right)-x_{ij}}{max\left(x_j\right)-\min \left(x_j\right)}$

As follows for more large more excellent type index calculate:

$v_{ij}=\frac{x_{ij}-min\left(x_j\right)}{\max \left(x_j\right)-\min \left(x_j\right)}$

According to the appraisal procedure of entropy assessment, obtain the entropy weight of each evaluation index, assessment result is as shown in table 2:

Table 2 entropy assessment assessment result

As can be seen from Table 2, probability of malfunction, upkeep cost, breakdown loss is had to judge and maintenance difficulty for the larger index of Power Transformer Faults importance degree evaluation impact.The evaluation index that wherein weight is maximum is probability of malfunction.But when considering fault importance, probability of malfunction is the one side of assessment, and affects other factors that also have of fault importance sequence, and therefore other evaluation indexes also need to take into account.

Six evaluation indexes selected in the present embodiment are all more little more excellent type indexs.Evaluation criterion weight is obtained in conjunction with entropy assessment.First decision matrix is normalized and obtains V.According to the index weights that entropy assessment obtains, matrix V is weighted, then obtains out plus-minus ideal solutions.Finally obtain bushing shell for transformer fault plus-minus ideal solutions distance and approach degree as shown in table 3:

Table 3

An each fault approach degree from small to large minor sort is: humidified insulation, insulator flashover, aging, discharge, puncture, leakage field heating, leakage of oil.Sort more forward, situation is more excellent, and namely comprehensive examination and evaluation importance degree is lower, and approach degree is larger, and importance degree is higher, more needs guard key.

The present invention is on the basis taking into full account the various factors affecting fault mode importance degree, set up the evaluation indice of a fault mode, each evaluation index comprehensive and its respective weights (i.e. relative Link Importance) calculate final fault importance, the fault importance of the various different faults patterns of transformer can be determined so exactly, thus can in transformer life cycle management, the fault mode high for importance degree carries out important management and control, promotes the parallel-adder settle-out time of transformer.The present invention is applicable to any transformer, especially in network transformer.

The present invention constructs comparatively complete transformer fault importance degree assessment indicator system, and comprehensive many factors makes the evaluation of transformer fault importance degree more scientific and reasonable.

Utilize the evaluation criterion weight of entropy assessment determination evaluation object in the present invention, consider evaluation object association attributes, utilize information entropy to determine each index weights and then to calculate fault importance, decrease interference from human factor, calculate accuracy high.

The present invention can fast accurate ground calculating transformer certain fault mode relative to the Euclidean distance of optimum state, realize accurately determining sequence to the fault importance of transformer, calculate simple, thus can in transformer life cycle management, the fault mode high for importance degree carries out important management and control, promotes the parallel-adder settle-out time of transformer.

By reference to the accompanying drawings the specific embodiment of the present invention is described in detail above, but the present invention is not restricted to above-mentioned embodiment, in the spirit and scope situation of claim not departing from the application, those skilled in the art can make various amendment or remodeling.

Claims (3)

1. an appraisal procedure for transformer fault importance degree, is characterized in that, described method, first according to the various factors affecting fault mode importance degree, sets up the evaluation indice of transformer fault pattern; Then the evaluation index utilizing entropy assessment to calculate fault mode concentrates the weight of each evaluation index; The evaluation criterion weight that last basis calculates, calculates the fault importance of often kind of fault mode of power transformer.

2. the appraisal procedure of transformer fault importance degree according to claim 1, is characterized in that, said method comprising the steps of:

A. the evaluation indice of transformer fault pattern is set up

If the set participating in the fault mode evaluated is M=(M ₁m ₂m _n), evaluation indice is D=(D ₁d ₂d _m); Evaluate a kind of fault mode M _ian evaluation index D _jvalue be x _ij(i=1,2 ..., n; J=1,2 ..., m), form decision matrix X=(x based on this _ij) _{m × n};

B. the weight of each evaluation index in Calculation Estimation index set

1. standardization is carried out to decision matrix X, obtain normalized matrix V=(v _ij) _{m × n}, concrete grammar is:

All evaluation indexes are divided into two classes, are respectively more little more excellent type and more large more excellent type,

Evaluation index for more little more excellent type:

$v_{ij}=\frac{max\left(x_j\right)-x_{ij}}{max\left(x_j\right)-min\left(x_j\right)};$

Evaluation index for more large more excellent type:

$v_{ij}=\frac{x_{ij}-min\left(x_j\right)}{\max \left(x_j\right)-\min \left(x_j\right)};$

In formula, max (x _j), min (x _j) be respectively evaluation index D _jmaximal value and minimum value;

2. Calculation Estimation index D _jentropy e _j:

$e_j=-\frac{1}{ln\left(m\right)}\underset{i=1}{\overset{m}{\Σ}}{p}_{ij}ln\left(p_{ij}\right);$

Wherein work as p _ij=0 or p _ijwhen=1, p _ijln (p _ij)=0;

3. Calculation Estimation index D _jindex weights w _j:

$w_j=\frac{d_j}{\underset{k=1}{\overset{n}{\Σ}}{d}_k},j=1,2,...,m;$

Wherein, d _j=1-e _j, d _krepresent the coefficient of variation of kth item evaluation index, d _jrepresent the coefficient of variation of jth item evaluation index;

C. the fault importance of often kind of fault mode of power transformer is calculated

1. weighted decision matrix R=(r is built _ij) _{m × n}, wherein, r _ij=w _jv _ij;

2. positive ideal solution is calculated with minus ideal result

3. fault mode M to be evaluated is calculated _iwith the distance of positive ideal solution and with the distance of minus ideal result

${\mathrm{Sd}}_i^{+}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{+}-r_{ij})}^2},i=1,2,...,m;$

${\mathrm{Sd}}_i^{-}=\sqrt{\underset{j=1}{\overset{n}{\Σ}}{(S_j^{-}-r_{ij})}^2},i=1,2,...,m;$

4. object M to be evaluated is calculated _iwith the relative similarity degree η of ideal solution _i:

${\mathrm{\η}}_i=\frac{{\mathrm{Sd}}_i^{-}}{{\mathrm{Sd}}_i^{+}+{\mathrm{Sd}}_i^{-}},i=1,2,...,m;$

By η _icharacterize transformer fault importance degree, η _ilarger, fault importance is larger, and vice versa.

3. the appraisal procedure of transformer fault importance degree according to claim 1 and 2, it is characterized in that, the evaluation index that the evaluation index of described transformer fault pattern is concentrated comprises fault rate, impact on systemic-function, maintenance cost, maintenance difficulty, detectability and loss and judges.