CN103513138B - A kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults - Google Patents

Abstract

The present invention relates to a kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults, the method comprises the time domain sequences gathering original vibration signal and vibration signal to be measured; Described time domain sequences is changed into respective frequency domain sequence respectively; Determine original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured; Described original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured are carried out data processing; Fault diagnosis is carried out according to described data processed result.The inventive method achieves being separated of iron core and basket vibration signal, can the vibrational state of more effecting reaction Transformer Winding and iron core, monitoring result is accurate, practical.

Description

A kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults

Technical field:

The present invention relates to a kind of Transformer Winding and iron core method for diagnosing faults, more specifically relate to a kind of Transformer Winding in vibration performance and iron core method for diagnosing faults.

Background technology:

Power transformer is one of most important equipment in electric system, and its reliability service is for safeguarding that the stability of electric system is significant.In recent years, along with the increase of transformer load capacity, cutting-out of voltage changer causes the dash current flowing through winding coil to increase, and causes winding deformation; In addition, due to the long-time running of transformer, iron core snap-in force weakens, and siliconized plate loosens Frequent Accidents, and serious causes iron core flexural deformation, destroys the mechanical stability of transformer, causes potential potential safety hazard to the stable operation of transformer.

The method for diagnosing faults of current Transformer Winding and iron core has: short circuit impedance method, frequency response method, Low Voltage Impulse Method, dissolved gas analysis method.But all there is the defects such as sensitivity is lower, poor anti jamming capability, repeatability are not strong in varying degrees in these methods.Meanwhile, these methods all belong to off-line monitoring method, can not real time on-line monitoring Transformer Winding and iron core operation conditions Timeliness coverage fault.

Patent publication No. is " CN101782426A ", the Chinese patent that name is called " detection method of looseness fault vibration of power transformer winding " by calculate the Fourier spectrum of vibration signal, kurtosis value, 6 time scale sampling entropys with etc. parameter, to the special vibrational state of Transformer Winding, slightly to loosen, moderate loosens and serious loosening judges.But the method just loosens the vibration parameters analysis under situation for the specific fault of transformer, winding and core vibration signal are not effectively separated, also just cannot do effective monitoring to other malfunctions such as winding deformation and iron core grounding, all there is certain limitation in the accuracy of its status monitoring, validity and practicality.

Summary of the invention:

The object of this invention is to provide a kind of Transformer Winding in vibration performance and iron core method for diagnosing faults, the method can the vibrational state of more effecting reaction Transformer Winding and iron core, and monitoring result is more accurate, and practicality is stronger.

For achieving the above object, the present invention by the following technical solutions: a kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults, the method comprises the following steps:

(1) time domain sequences of original vibration signal and vibration signal to be measured is gathered;

(2) described time domain sequences is changed into respective frequency domain sequence respectively;

(3) original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured is determined;

(4) described original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured are carried out data processing;

(5) fault diagnosis is carried out according to the data processed result of described step (4).

A kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, the time domain of the original vibration signal in described step (1) is the original unloaded vibration data of transformer and original load vibration data; The time domain of described vibration signal to be measured is transformer unloaded vibration data to be measured and transformer load vibration data to be measured.

A kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, the original unloaded vibration data of described transformer be after transformer field Installation and Debugging, put into operation before light condition under, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Under the rated load condition that described original load vibration data is after described transformer puts into operation, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Described transformer unloaded vibration data to be measured is under the postrun light condition of described transformer, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Described transformer load to be measured vibration data is under the postrun rated load condition of described transformer, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured.

Another preferably a kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults provided by the invention, convert the time domain sequences of original vibration signal in described step (2) and the time domain sequences of vibration signal to be measured to respective frequency domain sequence respectively by fast Fourier change;

Described original unloaded vibration data is converted to its frequency domain sequence X (k) by fast Fourier change;

Described original load vibration data is converted to its frequency domain sequence Y (k) by fast Fourier change;

Described unloaded vibration data to be measured is converted to its frequency domain sequence A (k) by fast Fourier change;

Described load vibration data to be measured is converted to its frequency domain sequence B (k) by fast Fourier change.

Preferably a kind of Transformer Winding based on vibration performance and an iron core method for diagnosing faults more provided by the invention, in described step (3), original core vibration signal frequency domain sequence is X (k), core vibration signal frequency domain sequence to be measured is A (k);

Described original basket vibration signal frequency domain sequence is Z(k)=Y(k)-X(k), wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Described basket vibration signal frequency domain sequence C (k) to be measured=B(k)-A(k), wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Be A (k) by described core vibration signal frequency domain sequence to be measured with original core vibration signal frequency domain sequence be that X (k) carries out data processing;

Described basket vibration signal frequency domain sequence C (k) to be measured and original basket vibration signal frequency domain sequence Z (k) are carried out data processing.

Another preferably a kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, described data processing step is:

(4-1) the related coefficient C of described frequency domain sequence A (k) and described frequency domain sequence X (k) is determined ₁with amplitude coefficient F ₁:

(4-2) the related coefficient C of described sequence of differences C (k) and described sequence of differences Z (k) is determined ₂with amplitude coefficient F ₂.

Another preferably a kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, described related coefficient determining step is:

$C=\frac{{\mathrm{COV}}_{\mathrm{pq}}}{\sqrt{D_p}\sqrt{D_q}}---\left(1\right)$

${\mathrm{COV}}_{\mathrm{pq}}=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}P\left(k\right)][Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)]---\left(2\right)$

$\left\{\begin{array}{c}{D}_p=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{[P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}P\left(k\right)]}^2\\ D_q=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{[Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)]}^2\end{array}\right.---\left(3\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k);

C is described related coefficient C1 or C2;

COV _pqfor described frequency domain sequence A (k) and the covariance of described frequency domain sequence X (k), or

The covariance of described sequence of differences C (k) and described sequence of differences Z (k);

D _pfor the variance of described frequency domain sequence A (k), or

The variance of described sequence of differences C (k)

D _qfor the variance of described frequency domain sequence X (k), or

The variance of described sequence of differences Z (k).

Another preferably a kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, described amplitude coefficient determining step is:

$F=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[P\left(k\right)-Q\left(k\right)]}{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)}---\left(4\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

F is described amplitude coefficient F ₁or F ₂;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k).

Another preferably a kind of Transformer Winding based on vibration performance provided by the invention and iron core method for diagnosing faults, described step (5) is according to the related coefficient C in table 1 ₁with amplitude coefficient F ₁span and table 2 in related coefficient C ₂with amplitude coefficient F ₂span fault diagnosis is carried out to transformer core and winding.

Table 1

Table 2

With immediate prior art ratio, the invention provides technical scheme and there is following excellent effect

1, the mode that in the present invention, method adopts no-load transformer and load signal to gather respectively, achieves being separated of iron core and basket vibration signal, can the vibrational state of more effecting reaction Transformer Winding and iron core, and more accurately, practicality is stronger for monitoring result;

2, the present invention is directed to Transformer Winding and core vibration signal difference, set different related coefficients, amplitude coefficient threshold value respectively, propose corresponding Quantitative Diagnosis judgment criteria, there is site operative strong, accuracy of judgement, be quick on the draw, advantages such as judging, easy to implement can be quantized;

3, in the present invention, the present invention can be widely used in the fault diagnosis of Transformer Winding and iron core.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of the inventive method;

Fig. 2 is the schematic flow sheet of the inventive method;

Fig. 3 is data flow diagram of the present invention.

Embodiment

Below in conjunction with embodiment, the invention will be described in further detail.

Embodiment 1:

As Figure 1-3, the invention of this example is based on the Transformer Winding of vibration performance and iron core method for diagnosing faults, and the method comprises the following steps:

(1) time domain sequences of original vibration signal and vibration signal to be measured is gathered;

(2) time domain sequences of described original vibration signal and vibration signal to be measured is changed into respective frequency domain sequence respectively;

(3) original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured is determined;

(4) described original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured are carried out data processing;

(5) fault diagnosis is carried out according to the data processed result of described step (4).

Described original vibration signal acquisition step refers to and gathers the original unloaded vibration data of transformer and original load vibration data, then described original vibration signal is delivered to original vibration signal fast Fourier change FFT time-frequency convert step.The original unloaded vibration data of transformer of described collection refers to after described transformer field Installation and Debugging, before putting into operation, under described no-load transformer state, measures the vibration signal time domain sequences of oil tank of transformer surface fixed position.The transformer original load vibration data of described collection refers to after described transformer puts into operation, and under described transformer rated load condition, measures the vibration signal time domain sequences of oil tank of transformer surface fixed position.

Described vibration signals collecting step to be measured refers to described collection transformer unloaded vibration data to be measured and transformer load vibration data to be measured, then described vibration signal to be measured is delivered to vibration signal fast Fourier to be measured change FFT time-frequency convert step.The to be measured unloaded vibration data of described collection transformer refers to after described transformer runs a period of time, under described no-load transformer state, measures the vibration signal time domain sequences of oil tank of transformer surface fixed position.The transformer load vibration data to be measured of described collection refers to after transformer runs a period of time, under transformer rated load condition, measures the vibration signal time domain sequences of oil tank of transformer surface fixed position.

Described original vibration signal fast Fourier change FFT time-frequency convert step is that the original unloaded vibration data of described transformer of described original vibration signal acquisition step transmission and the original load vibration data of transformer are carried out fast Fourier FFT conversion respectively, obtain original unloaded vibration data frequency domain sequence X (k) of described transformer and original load vibration data frequency domain sequence Y (k) of transformer, deliver to original winding signal and be separated.

Described vibration signal FFT time-frequency convert step to be measured is that the transformer unloaded vibration data to be measured of described vibration signals collecting step transmission to be measured and transformer load vibration data to be measured are carried out fast Fourier FFT conversion respectively, obtain described transformer unloaded vibration data frequency domain sequence A (k) to be measured and transformer load vibration data frequency domain sequence B (k) to be measured, deliver to winding signal to be measured and be separated.

Wherein, described original core vibration signal frequency domain sequence is X (k) and is A (k) with core vibration signal frequency domain sequence to be measured;

Described original basket vibration signal frequency domain sequence is Z(k)=Y(k)-X(k), wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Described basket vibration signal frequency domain sequence C (k) to be measured=B(k)-A(k), wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Be A (k) by described core vibration signal frequency domain sequence to be measured with original core vibration signal frequency domain sequence be that X (k) carries out data processing;

Described basket vibration signal frequency domain sequence C (k) to be measured and original basket vibration signal frequency domain sequence Z (k) are carried out data processing.

Described data processing step be according to process and Signal separator after original iron core and winding signal and iron core to be measured and winding signal frequency domain sequence, to calculate described transformer unloaded sequence A (k) to be measured be described core vibration signal frequency domain sequence to be measured and described original unloaded sequence X (k) is the related coefficient C of original core vibration signal frequency domain sequence ₁with amplitude coefficient F ₁, calculate described transformer sequence of differences C (k) i.e. to be measured basket vibration signal frequency domain sequence to be measured and described original difference sequence Z (k) the i.e. related coefficient C of original basket vibration signal frequency domain sequence ₂with amplitude coefficient F ₂.

Described troubleshooting step is by described related coefficient C ₁, C ₂with amplitude coefficient F ₁, F ₂according to Quantitative Diagnosis judgment criteria, the size of value, judges whether Transformer Winding and iron core exist fault, obtain last diagnostic conclusion.

The computing method of related coefficient C described in described data processing step are:

$C=\frac{{\mathrm{COV}}_{\mathrm{pq}}}{\sqrt{D_p}\sqrt{D_q}}---\left(1\right)$

${\mathrm{COV}}_{\mathrm{pq}}=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}P\left(k\right)][Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)]---\left(2\right)$

$\left\{\begin{array}{c}{D}_p=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{[P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}P\left(k\right)]}^2\\ D_q=\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{[Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)]}^2\end{array}\right.---\left(3\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k);

C is described related coefficient C1 or C2;

COV _pqfor described frequency domain sequence A (k) and the covariance of described frequency domain sequence X (k), or

The covariance of described sequence of differences C (k) and described sequence of differences Z (k);

D _pfor the variance of described frequency domain sequence A (k), or

The variance of described sequence of differences C (k);

D _qfor the variance of described frequency domain sequence X (k), or

The variance of described sequence of differences Z (k).

Described amplitude coefficient F determining step is:

$F=\frac{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[P\left(k\right)-Q\left(k\right)]}{\underset{k=1}{\overset{N}{\mathrm{\Σ}}}Q\left(k\right)}---\left(4\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

F is described amplitude coefficient F ₁or F ₂;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k).

In described troubleshooting step, iron core fault diagnosis judgment criteria: work as C ₁> 0.92 and F ₁during < 0.07, show that iron core is in good condition; As 0.88 < C ₁< 0.92 and 0.07 < F ₁during < 0.11, show that slight loosening may occur iron core silicon-steel sheet; As 0.81 < C ₁< 0.88 and 0.11 < F ₁during < 0.16, show that iron core silicon-steel sheet may seriously loosen; As 0.75 < C ₁< 0.81 and 0.16 < F ₁during < 0.2, show that iron core may flexural deformation; Work as C ₁< 0.75 and 0.2 < F ₁time, showing may iron core grounding.

Described winding failure diagnosis judgment criteria is for work as C ₂> 0.9 and F ₂during < 0.09, show that winding state is good; As 0.85 < C ₂< 0.9 and 0.09 < F ₂during < 0.14, show that slight loosening may occur winding; As 0.77 < C ₂< 0.85 and 0.14 < F ₂during < 0.2, show that winding may seriously loosen; As 0.7 < C ₂< 0.77 and 0.2 < F ₂during < 0.25, show that winding may torsional deformation; Work as C ₂< 0.7 and 0.25 < F ₂time, show that transformer may exist short circuit.

Described transformer due to run time produce vibration mainly comprise basket vibration and core vibration.Under idle condition, described transformer load electric current is zero, and the vibration of described winding can be ignored, and described transformer body vibrates that the vibration of mainly exciting curent effect lower core causes.Under that loading condition, have load current to flow through in the coil of described transformer, except described core vibration, the vibration of described transformer body also comprises the vibration of winding coil under load current effect.Therefore, utilize vibration transducer, measure the vibration signal of described transformer body under zero load and loading condition respectively, what record time unloaded is exactly described transformer core vibration signal, is the superposition of described iron core and basket vibration signal during load.Vibration signal when isolating zero load from vibration signal during described transformer load just can obtain the vibration signal of winding.

With Chinese DianKeYuan 110kv extra-high voltage experiment base transformer for experimental subjects, test according to method of the present invention, when diagnosing, for the ease of the vibration signal of the different amplitude of distortionless continuous acquisition, vibration transducer is fixed on fuel tank sidewall, adopt strong magnetic suck or fasten with glue fixing, described vibration signal sample frequency is set to 8192HZ.

Gather 10 groups, fixed position, oil tank of transformer surface running status vibration data in this test respectively, specifically comprise: 1. iron core is in good condition; 2. iron core silicon-steel sheet slightly loosens; 3. iron core silicon-steel sheet seriously loosens; 4. iron core flexural deformation; 5. iron core grounding; 6. good 7. windings of winding state slightly loosen; 8. winding seriously loosens 9. winding torsional deformations; 10. transformer short-circuit.Through calculating, under the various state of transformer, related coefficient and amplitude coefficient are for shown in table 3 and table 4:

Coefficient table under the various state of table 3 transformer core

Coefficient table under the various state of table 4 Transformer Winding

The related coefficient of table 3 and table 4 and amplitude coefficient binding tests error draw judges standard as the transformer core of table 1 and table 2 and winding failure diagnosis:

Table 1 transformer core fault diagnosis judgment criteria table

Table 2 transformer winding fault diagnosis judgment criteria table

From theoretical analysis, described transformer winding vibration signal concentrates on fundamental frequency 100Hz place substantially, and the size of vibration signal is directly proportional to load current I; Described core vibration signal mainly concentrates on 100 ~ 400Hz, substantially decays to 0 after 1000Hz.When described Transformer Winding and iron core exist potential faults, its vibrational energy curve, amplitude characteristic parameter will change, and be in particular in that the feature on vibration signal is: the harmonic components adding more high order in vibration frequency, and vibration amplitude becomes large.

In view of above-mentioned feature, calculate related coefficient and amplitude coefficient that described transformer waits to diagnose vibration frequency domain sequence and original vibration data frequency domain sequence two curves by experiment respectively, by drawing the experimental data of transformer under 10 kinds of different running statuses: transformer core and winding state good time, both being calculated by the method, gained related coefficient is tending towards 1, and amplitude coefficient is tending towards 0.Illustrate that vibration signals spectrograph sequence to be measured is compared with frequency spectrum transformer normal operating condition from the frequency spectrum angle analysis of vibration signal, crest frequency and higher harmonic component basically identical, in vibration signal to be measured there is not obvious increase and decrease in amplitude simultaneously, and the curve of the two correspondence is basically identical.Occur slight to loosen at transformer core silicon-steel and winding, seriously loosen when even there is serious flexural deformation, related coefficient is significantly successively decreased, and amplitude coefficient increases progressively simultaneously.Illustrate that now vibration signals spectrograph sequence to be measured and frequency spectrum phase ratio error under transformer normal operating condition are comparatively large, may have new higher hamonic wave to produce, and the vibration amplitude increase that each frequency is corresponding, finally cause related coefficient to be successively decreased, amplitude coefficient increases progressively.Size accordingly by two coefficients judges Transformer Winding and iron core fault severity level.

Finally should be noted that: above embodiment is only in order to illustrate that technical scheme of the present invention is not intended to limit, although with reference to above-described embodiment to invention has been detailed description, those of ordinary skill in the field are to be understood that: still can modify to the specific embodiment of the present invention or equivalent replacement, and not departing from any amendment of spirit and scope of the invention or equivalent replacement, it all should be encompassed in the middle of this right.

Claims (2)

1., based on Transformer Winding and the iron core method for diagnosing faults of vibration performance, it is characterized in that: the method comprises the following steps:

(1) time domain sequences of original vibration signal and vibration signal to be measured is gathered;

(2) described time domain sequences is changed into respective frequency domain sequence respectively;

(3) original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured is determined;

(4) described original core vibration signal frequency domain sequence, original basket vibration signal frequency domain sequence and core vibration signal frequency domain sequence to be measured, basket vibration signal frequency domain sequence to be measured are carried out data processing;

(5) fault diagnosis is carried out according to the data processed result of described step (4);

The time domain of the original vibration signal in described step (1) is the original unloaded vibration data of transformer and original load vibration data; The time domain of described vibration signal to be measured is transformer unloaded vibration data to be measured and transformer load vibration data to be measured;

The original unloaded vibration data of described transformer be after transformer field Installation and Debugging, put into operation before light condition under, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Under the rated load condition that described original load vibration data is after described transformer puts into operation, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Described transformer unloaded vibration data to be measured is under the postrun light condition of described transformer, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

Described transformer load to be measured vibration data is under the postrun rated load condition of described transformer, the vibration signal time domain sequences of the described oil tank of transformer surface fixed position measured;

The time domain sequences of original vibration signal in described step (2) and the time domain sequences of vibration signal to be measured is converted to respective frequency domain sequence respectively by fast Fourier change;

Described original unloaded vibration data is converted to its frequency domain sequence X (k) by fast Fourier change;

Described original load vibration data is converted to its frequency domain sequence Y (k) by fast Fourier change;

Described unloaded vibration data to be measured is converted to its frequency domain sequence A (k) by fast Fourier change;

Described load vibration data to be measured is converted to its frequency domain sequence B (k) by fast Fourier change;

In described step (3), original core vibration signal frequency domain sequence is X (k), core vibration signal frequency domain sequence to be measured is A (k);

Described original basket vibration signal frequency domain sequence is Z (k)=Y (k)-X (k), and wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Described basket vibration signal frequency domain sequence C (k) to be measured=B (k)-A (k), wherein k is in data sequence 1,2 ... N arbitrary data, N is the length of data sequence;

Be A (k) by described core vibration signal frequency domain sequence to be measured with original core vibration signal frequency domain sequence be that X (k) carries out data processing;

Described basket vibration signal frequency domain sequence C (k) to be measured and original basket vibration signal frequency domain sequence Z (k) are carried out data processing;

Described data processing step is:

(4-1) the related coefficient C of described frequency domain sequence A (k) and described frequency domain sequence X (k) is determined ₁with amplitude coefficient F ₁:

(4-2) the related coefficient C of described frequency domain sequence C (k) and described frequency domain sequence Z (k) is determined ₂with amplitude coefficient F ₂;

Described related coefficient determining step is:

$C=\frac{{\mathrm{COV}}_{pq}}{\sqrt{D_p}\sqrt{D_q}}---\left(1\right)$

${\mathrm{COV}}_{pq}=\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}\&lsqb;P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}P\left(k\right)\&rsqb;\&lsqb;Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}Q\left(k\right)\&rsqb;---\left(2\right)$

$\left\{\begin{array}{c}{D}_p=\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}{\&lsqb;P\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}P\left(k\right)\&rsqb;}^2\\ D_q=\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}{\&lsqb;Q\left(k\right)-\frac{1}{N}\underset{k=1}{\overset{N}{\&Sigma;}}Q\left(k\right)\&rsqb;}^2\end{array}\right.---\left(3\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k);

C is described related coefficient C1 or C2;

COV _pqfor described frequency domain sequence A (k) and the covariance of described frequency domain sequence X (k), or

The covariance of described sequence of differences C (k) and described sequence of differences Z (k);

D _pfor the variance of described frequency domain sequence A (k), or

The variance of described sequence of differences C (k)

D _qfor the variance of described frequency domain sequence X (k), or

The variance of described sequence of differences Z (k);

Described amplitude coefficient determining step is:

$F=\frac{\underset{k=1}{\overset{N}{\&Sigma;}}\&lsqb;P\left(k\right)-Q\left(k\right)\&rsqb;}{\underset{k=1}{\overset{N}{\&Sigma;}}Q\left(k\right)}---\left(4\right)$

Wherein, N is the length of data sequence, and k is in data sequence 1,2 ... N arbitrary data;

F is described amplitude coefficient F ₁or F ₂;

P (k) and Q (k) are respectively transformer zero load to be measured and vibrate frequency domain sequence A (k) and frequency domain sequence X (k) is vibrated in original zero load, or

P (k) and Q (k) are respectively transformer sequence of differences C (k) to be measured and original difference sequence Z (k).

2. a kind of Transformer Winding based on vibration performance and iron core method for diagnosing faults as claimed in claim 1, is characterized in that: described step (5) is according to the related coefficient C in table 1 ₁with amplitude coefficient F ₁span and table 2 in related coefficient C ₂with amplitude coefficient F ₂span fault diagnosis is carried out to transformer core and winding.

Table 1

Table 2