ONLINE DIAGNOSIS METHOD FOR DEFORMATION POSITION ON TRASNFORMATION WINDING

Abstract

The invention discloses an online diagnosis method for transformer winding deformation position, including: (1) collecting a transformer with known winding state and decomposing into several position sub-samples; (2) performing feature extraction on each position sub-sample with information entropy, adding with label indicating deformation and inputting into support vector machine to train diagnosis model; (3) decomposing a transformer under diagnosis into 9 position subsamples in the way of step (1), performing feature extraction of step (2) and inputting into the diagnostic model trained in step (2); (4) outputting diagnosis result from the support vector machine about whether the position sub-samples of the transformer is deformed. The invention can achieve intelligent diagnosis of winding deformation by comprehensively considering variations of monitoring indicators of the transformer in complexity, time-frequency domain and other aspects and automatically learning diagnostic logic from fault features through machine learning algorithms, thereby reducing labor costs and improving diagnostic efficiency.

Claims

1. An online diagnostic method for deformation position on transformer winding, comprising: (1) taking current, voltage, current difference and voltage difference of each phase of each winding in a transformer for which winding state is known as online monitoring indicators, and grouping the online monitoring indicators into several position sub-samples according to positions; (2) obtaining two non-dimensional online monitoring data sequences by dividing online monitoring data recorded according to the online monitoring indicators into two sequences according to time and normalizing the two sequences; (3) calculating permutation entropy, wavelet energy and arithmetic mean of each of the two non-dimensional online monitoring data sequences, and calculating root mean square errors of the permutation entropies, the wavelet energies and the arithmetic means, respectively; (4) constructing a four-dimensional feature set by using the three root mean square errors obtained in step (3) and a cumulative short-circuit current of a corresponding position sub-sample as features, wherein the cumulative short-circuit current is a sum of short-circuit currents cumulatively suffered at a position corresponding to the position sub-sample; (5) adding the four-dimensional feature set with a label and inputting the four-dimensional feature set into a support vector machine for diagnostic model training, wherein the label is used to display winding deformation status of the position corresponding to the position subsample; and (6) obtaining a four-dimensional feature set by performing feature extraction on a transformer under diagnosis with steps (1)-(4), inputting the obtained four-dimensional feature set into a diagnostic model trained in step (5), and performing diagnose on the positions corresponding to respective position subsamples to determine whether there is a winding deformation.

2. The online diagnostic method for deformation position on transformer winding according to claim 1, wherein there are 9 of the position subsamples: high-voltage phase-A, high-voltage phase-B, high-voltage phase-C, medium-voltage phase-A, medium-voltage phase-B, medium-voltage phase-C, low-voltage phase-A, low-voltage phase-B, and low-voltage phase-C.

3. The online diagnostic method for deformation position on transformer winding according to claim 2, wherein the online monitoring indicators of the high-voltage phase-A are high-voltage phase-A current, high-voltage phase-A voltage, high-voltage phases A/B current difference, and high-voltage phases A/B voltage difference; the online monitoring indicators of high-voltage phase-B are high-voltage phase-B current, high-voltage phase-B voltage, high-voltage phases A/B current difference, and high-voltage phases A/B voltage difference; the online monitoring indicators of high-voltage phase-C are high-voltage phase-B current, high-voltage phase-B voltage, high-voltage phases B/C current difference, and high-voltage phases B/C voltage difference; the online monitoring indicators of medium-voltage phase-A are medium-voltage phase-A current, medium-voltage phase-A voltage, medium-voltage phases A/B current difference and medium-voltage phases A/B voltage difference; the online monitoring indicators of medium-voltage phase-B are medium-voltage phase-B current, medium-voltage phase-B voltage, medium-voltage phases A/B current difference and medium-voltage phases A/B voltage difference; the online monitoring indicators of medium-voltage phase-C are medium-voltage phase-B current, medium-voltage phase-B voltage, medium-voltage phases B/C current difference and medium-voltage phases B/C voltage difference; the online monitoring indicators of low-voltage phase-A are low-voltage phase-A current, low-voltage phase-A voltage, low-voltage phases A/B current difference, and low-voltage phases A/B voltage difference; the on-line monitoring indicators of low-voltage phase-B are low-voltage phase-B current, low-voltage phase-B voltage, low-voltage phases A/B current difference, and low-voltage phases A/B voltage difference; and the on-line monitoring indicators of the low-voltage phase-C are low-voltage phase-B current, low-voltage phase-B voltage, low-voltage phases B/C current difference, and low-voltage phases B/C voltage difference.

4. The online diagnostic method for deformation position on transformer winding according to claim 1, wherein the dividing online monitoring data recorded according to the online monitoring indicators into two sequences according to time comprises: for the transformer that has subjected to short circuit, dividing, according to occurrence time of a latest short circuit, the online monitoring data into a front-segment sequence before the short-circuit and a back-segment sequence after the short-circuit; and for the transformer that has not subjected to short-circuit, dividing, according to time length, the online monitoring data into a front-segment sequence and a back-segment sequence.

5. The online diagnostic method for deformation position on transformer winding according to claim 1, wherein the normalization may be maximum or minimum normalization, and the online monitoring data may be converted into [0, 1] interval by following formula to obtain the non-dimensional online monitoring data x*: $x^{*}=\frac{x-x_{\min }}{x_{\max }-x_{\min }}$ where x is the online monitoring data recorded according to the online monitoring indicators, x.sub.max is a maximum value of the online monitoring data recorded according to a same online monitoring indicator, and x.sub.min is a minimum value of the online monitoring data recorded according to the same online monitoring indicator.

6. The online diagnostic method for deformation position on transformer winding according to claim 1, wherein the diagnostic model training comprises: finding a hyperplane that is able to separate deformation data and normal data and maximizes an interval between these two types of data.

7. The online diagnostic method for deformation position on transformer winding according to claim 6, wherein performing diagnose on the positions corresponding to respective position subsamples to determine whether there is a winding deformation comprises: if a subsample point to be diagnosed is located on a deformation side of the hyperplane, it is determined that a deformation occurs at the position; and if the subsample point to be diagnosed is located on a normal side of the hyperplane, it is determined that no deformation occurs at the position.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0061] FIG. 1 is a flow chart of an online diagnosis method for deformation position on transformer winding according to embodiment 1;

[0062] FIG. 2 is a diagram showing variations of permutation entropies of each online monitoring indicator before and after a short circuit of the transformer a according to embodiment 1; and

[0063] FIG. 3 is a two-dimensional dispersion diagram of root mean square errors of permutation entropies and arithmetic means of front-segment sequence before short circuit and back-segment sequence after the short circuit for 27 position subsamples of three transformers according to embodiment 1.

DETAILED DESCRIPTION OF THE PRESENT INVENTION

[0064] The invention is further illustrated below in conjunction with the drawings and specific embodiments. It is to be understood that the embodiments are illustrative and are not intended to limit the scope of the invention. The experimental methods in the following embodiments for which the specific conditions are not specified are generally in conformity with conventional conditions or in conformity with conditions recommended by the manufacturer.

Embodiment 1

[0065] As shown in FIG. 1, the flow of an online diagnostic method for deformation position on transformer winding is as follows:

[0066] S01, collecting transformers for which winding state is known, decomposing each transformer into 9 position sub-samples according to three phases and three windings: high-voltage phase-A, high-voltage phase-B, high-voltage phase-C, medium-voltage phase-A, medium-voltage phase-B, medium-voltage phase-C, low-voltage phase-A, low-voltage phase-B, and low-voltage phase-C;

[0067] S02, performing feature extraction on each position subsample by using information entropy, adding with a label indicating whether deformation occurs and inputting into a support vector machine to train a diagnosis model;

[0068] S03, decomposing a transformer under diagnosis into 9 position subsamples in the way of S01, performing the feature extraction of S02, and inputting into the diagnostic model trained in S02; and

[0069] S04, outputting diagnosis result from the support vector machine as for whether the position sub-samples of the transformer under diagnosis are deformed.

[0070] (1) Three transformers, for which deformation has happened and specific deformation position is specified, is selected, which are transformer a, transformer b and transformer c, respectively. The voltage level of transformer a is 110 kV, and the voltage level of transformer b and transformer c is 220 kV. The recorded online monitoring indicators that can be directly read are shown in Table 1.

TABLE-US-00001 TABLE 1 Recorded online monitoring indicators that can be directly read High voltage Medium voltage Low voltage winding winding winding Phase A indicator1: indicator2: indicator3: current High voltage Medium voltage Low-voltage winding phase-A winding phase-A winding phase-A current value current value current value Phase A indicator4: indicator5: indicator6: voltage High voltage Medium voltage Low voltage winding phase-A winding phase-A winding phase-A voltage value voltage value voltage value Phase B indicator7: indicator8: indicator9: current High voltage Medium voltage Low-voltage winding phase B winding phase B winding phase B current value current value current value Phase B indicator10: indicator11: indicator12: voltage High voltage Medium voltage Low-voltage winding phase B winding phase B winding phase B voltage value voltage value voltage value Phase C indicator13: indicator14: indicator15: current High voltage Medium voltage Low-voltage winding phase C winding phase C winding phase C current value current value current value Phase C indicator16: indicator17: indicator18: voltage High voltage Medium voltage Low-voltage winding phase C winding phase C winding phase C voltage value voltage value voltage value

[0071] (2) Differences between the on-line monitoring data of the currents and voltages of respective phases and respective windings of each transformer are calculated, to construct the current-phase differences and voltage-phase differences, which are combined with the recorded online monitoring indicators that can be directly read into new complete online monitoring indicators.

[0072] (3) The online monitoring indicators of each transformer are grouped according to three phases and three-windings into 9 position sub-samples. The online monitoring indicators corresponding to each position sub-sample are shown in Table 2. Among the 27 sub-samples of the 3 transformers, 8 sub-samples were found to have undergone winding deformation after offline test and disassemble, which are the medium-voltage phase-A, medium-voltage phase-B, high-voltage phase-A, and high-voltage phase-C of transformer a, medium-voltage phase-B of transformer b, low-voltage phase-A, low-voltage phase-B and low-voltage phase-C of transformer c. The remaining 19 position subsample windings are normal.

TABLE-US-00002 TABLE 2 Grouping results of online monitoring indicators Position subsample Subordinate online monitoring indicators low-voltage Low-voltage winding phase-A current value, phase-A low-voltage winding phase-A voltage value, low-voltage winding phases-AB current difference, low-voltage windingphases-AB voltage value low-voltage Low-voltage winding phase-B current value, phase-B low-voltage winding phase-B voltage value, low-voltage winding phases-AB current difference, low-voltage winding phases-AB voltage value low-voltage Low-voltage winding phase-C current value, phase-C low-voltage winding phase-C voltage value, low-voltage winding phases-BC current difference, low-voltage winding phases-BC voltage value Medium-voltage Medium voltage winding phase-A current value, phase-A medium-voltage winding phase-A voltage value, medium-voltage winding phases-AB current difference, medium-voltage winding phases-AB voltage value Medium voltage Medium voltage winding phase-B current value, phase-B medium-voltage winding phase-B voltage value, medium-voltage winding phases-AB current difference, medium-voltage winding phases-AB voltage value Medium voltage Medium-voltage winding phase-C current value, phase-C medium-voltage winding phase-C voltage value, medium-voltage winding phases-BC current difference, medium-voltage winding phases-BC voltage value High voltage High-voltage winding phase-A current value, phase-A high-voltage winding phase-A voltage value, high-voltage winding phases-AB current difference, high-voltage winding phases-AB voltage value High voltage High-voltage winding phase-B current value, phase-B high-voltage winding phase-B voltage value, high-voltage winding phases-AB current difference, high-voltage winding phases-AB voltage value High voltage High-voltage winding phase-C current value, phase-C high-voltage winding phase-C voltage value, high-voltage winding phases-BC current difference, high-voltage winding phases-BC voltage value

[0073] (4) Feature extraction. The online monitoring data recorded according to the subordinate online monitoring indicators of each position subsample is divided into a front-segment sequence before the short circuit and a back-segment sequence after the short circuit according to the latest short circuit time, and maximum and minimum normalization is performed. The root mean square error of the permutation entropy, the root mean square error of the wavelet energy, and the root mean square error of the arithmetic mean of the processed sequences are calculated. The obtained three root mean square errors and the cumulative short-circuit current of the corresponding position sub-sample are characterized as a four-dimensional feature set.

[0074] Taking the low-voltage phase-A position subsample of transformer a as an example, the feature extraction is performed as follows:

[0075] a. According to the latest short-circuit time, Jan. 24, 2015, the online monitoring data recorded according to the monitoring indicators subordinate to low-voltage phase-A is divided into the front-segment sequence T.sub.before before the short-circuit (2013 Nov. 1-2015 Jan. 24) and the back-segment sequence T.sub.after after the short-circuit (2015 Jan. 24-2015 Aug. 13), and is subjected to the maximum and minimum normalization to be converted to the [0,1] interval for de-dimension.

[0076] The formula for maximum and minimum normalization is:

[00010]$x^{*}=\frac{x-x_{\min }}{x_{\max }-x_{\min }}$

[0077] where x* is non-dimensional online monitoring data, and x is online monitoring data recorded according to online monitoring indicators, where x.sub.max is a maximum value of the online monitoring data recorded according to a same online monitoring indicator, and x.sub.min is a minimum value of the online monitoring data recorded according to the same online monitoring indicator.

[0078] b. The root mean square error of the permutation entropy, the root mean square error of the wavelet energy, and the root mean square error of the arithmetic mean of the two non-dimensional online monitoring data sequences are calculated according to the method described in the Summary of the Invention. The results are shown in Table 3. The root mean square error of the permutation entropy, the root mean square error of the wavelet energy, and the root mean square error of the arithmetic mean for the low-voltage phase-A position subsample are 0.303, 6.9095, and 0.039, respectively.

TABLE-US-00003 TABLE 3 Feature calculation and extraction results of the low-voltage phase-A position subsample of transformer a Permutation Entropy wavelet energy mean value Before After Before After Before After short short Differ- short short Differ- short short Differ- indicator circuit circuit ence circuit circuit ence circuit circuit ence Low 3.0707 2.6097 0.461 17.8113 12.9582 4.8531 0.212 0.1677 0.0443 voltage phase-A current value Low 3.1141 2.6696 0.4445 13.8621 1.2611 12.601 0.9255 0.9793 0.0539 voltage phases-A B current difference Low 3.1121 3.0938 0.0184 20.5114 20.5539 0.0425 0.6004 0.5646 0.0358 voltage phase-A voltage value Low 3.083 3.0868 0.0038 3.6475 0.711 2.9364 0.4967 0.4972 0.0005 voltage phases-A B voltage difference RMSE 0.3203 6.9095 0.0395

[0079] c. The three root mean square errors in Table 3 and the cumulative short circuit current of the low-voltage phase-A position subsample are characterized as a four-dimensional feature set. The cumulative short-circuit current of the low-voltage phase-A position subsample is obtained from the short-circuit record of the ledger information of transformer a, which is 9.2 kA. Therefore, the four-dimensional feature set of the low-voltage phase-A position subsample is [0.33, 6.9095, 0.039, 9.2].

[0080] The feature sets of the other 26 position subsamples are extracted according to steps a to c, and the summary results are shown in Table 4.

TABLE-US-00004 TABLE 4 Summary of feature extraction results for 27 position subsamples Front and Front and back Cumulative back wavelet Front and short trans- Position permutation energy back mean circuit former subsample entropy difference difference current trans- Low 0.320317 6.90946 0.039202 9.2 former voltage a phase-A Low 0.222896 7.233183 0.034076 0 voltage phase-B Low 0.345881 6.941979 0.045849 9.2 voltage phase-C Medium 0.36323 27.80493 0.100947 0 voltage phase-A Medium 0.254158 27.1847 0.068832 0 voltage phase-B Medium 0.363299 27.96553 0.102912 0 voltage phase-C High 0.319302 15.91197 0.050269 0 voltage phase-A High 0.315275 15.91999 0.050043 0 voltage phase-B High 0.325802 15.91999 0.052152 0 voltage phase-C trans- Low 0.020471 36.03064 0.03519 0 former voltage b phase-A Low 0.019591 37.50859 0.032284 0 voltage phase-B Low 0.033643 41.35286 0.038622 0 voltage phase-C Medium 0.078673 30.13068 0.089533 0 voltage phase-A Medium 0.091766 29.02597 0.08934 10.118 voltage phase-B Medium 0.116339 28.367999 0.090171 0 voltage phase-C High 0.054301 32.94126 0.088438 5.884 voltage phase-A High 0.019926 31.57624 0.088071 0 voltage phase-B High 0.108691 31.79485 0.08942 0 voltage phase-C trans- Low 0.182103 34.87654 0.068949 26.81 former voltage c phase-A Low 0.219823 40.41053 0.067634 26.81 voltage phase-B Low 0.219412 40.88089 0.069945 24.38 voltage phase-C Medium 0.08231 17.27661 0.021917 51.68 voltage phase-A Medium 0.07197 20.32252 0.020755 10.056 voltage phase-B Medium 0.107109 38.55591 0.040062 17.256 voltage phase-C High 0.090561 22.76819 0.041354 0 voltage phase-A High 0.094293 22.82805 0.041885 0 voltage phase-B High 0.095809 25.51844 0.043278 0 voltage phase-C

[0081] As shown in FIG. 2, the permutation entropies of the front-segment sequence before the short-circuit and the back-segment sequence after the short-circuit for most of the online monitoring indicators of the transformer a are significantly different, which specifically is: for the low-voltage side current difference, the medium-voltage side voltage difference, the high-voltage side current and the high-voltage side voltage difference after the short-circuit, the permutation entropy after the short circuit is significantly lower than that before the short circuit, and it is inferred that fault occurs to the transformer a after the short circuit, resulting in a change in the operating state.

[0082] As shown in FIG. 3, the horizontal axis is the root mean square errors of the arithmetic means of the front segment sequence before the short circuit and the back segment sequence after the short circuit; the longitudinal axis is the root mean square errors of the permutation entropies of the front segment sequence before the short circuit and the back segment sequence after the short circuit; the hollow circles represent the normal position subsamples, and are mostly concentrated in the lower left corner of the drawing; the solid circles represent the deformed position subsamples, which are mostly concentrated in the upper right corner of the drawing, indicating that, as compared to normal position subsample, the root mean square error of the permutation entropy and the root mean square error of the arithmetic mean of the front segment sequence before the short circuit and the back segment sequence after the short circuit for the deformed position subsample are larger. That's to say, the deformation of the transformer winding leads to change of the permutation entropy and arithmetic mean of the online monitoring data recorded according to the online monitoring indicators, and the role of permutation entropy and arithmetic mean in diagnosing winding deformation is further proved.

[0083] It can also be seen from FIG. 3 that the boundary between the normal position subsamples and the deformed position subsamples exhibits features of a quadratic curve, so using quadratic curve SVM in the training of deformation position diagnostic model is better than the ordinary linear SVM in the aspect of classification effect.

[0084] (5) The four-dimensional feature set obtained in step (4), after being added with labels indicating whether being deformed, is inputted to the SVM for the training and verification of the diagnostic model. Three-fold cross-validation method is used to perform statistics on the determination result of the model. Cross-validation refers to the method of dividing training samples and test samples into multiple sub-samples, dividing these sub-samples according to different proportions, and using a large number of sub-samples to verify a few sub-samples. The results are shown in Tables 5 and 6.

TABLE-US-00005 TABLE 5 Support vector machine cross-validation result statistics Determined by the Determined by the model Cross validation result model as normal as being deformed Actually normal 89.47% 10.53% Actually deformed 12.50% 87.50%

TABLE-US-00006 TABLE 6 support vector machine cross-validation results Actual Model Position operation diagnosis Result of transformer subsample state result determination transformer Low voltage normal normal correct a phase-A Low voltage normal normal correct phase-B Low voltage normal normal correct phase-C Medium deformed deformed correct voltage phase-A Medium deformed deformed correct voltage phase-B Medium normal deformed wrong voltage phase-C High deformed deformed correct voltage phase-A High normal deformed wrong voltage phase-B High deformed deformed correct voltage phase-C transformer Low voltage normal normal correct b phase-A Low voltage normal normal correct phase-B Low voltage normal normal correct phase-C Medium normal normal correct voltage phase-A Medium deformed normal wrong voltage phase-B Medium normal normal correct voltage phase-C High normal normal correct voltage phase-A High normal normal correct voltage phase-B High normal normal correct voltage phase-C transformer Low voltage deformed deformed correct c phase-A Low voltage deformed deformed correct phase-B Low voltage deformed deformed correct phase-C transformer Medium normal normal correct c voltage phase-A Medium normal normal correct voltage phase-B Medium normal normal correct voltage phase-C High normal normal correct voltage phase-A High normal normal correct voltage phase-B High normal normal correct voltage phase-C

[0085] It can be seen from Tables 5 and 6 that the online diagnostic method for deformation position on transformer winding described in this embodiment has high diagnostic accuracy, and the recognition rates for the normal position subsamples and the deformed position subsamples are 89.47% and 87.5%, respectively. An effective diagnosis of whether the winding of the transformer is deformed is possible.

[0086] In addition, it is to be understood that various modifications and changes may be made by those skilled in the art in view of the above description of the present invention, and these equivalents also fall within the scope defined by the appended claims.