A method and system for detecting deformation state of power transformer winding

Abstract

The present invention relates to the field of transformer detection, and in particular to a method and system for detecting the deformation state of a power transformer winding. The method comprises: obtaining a data set, the data set comprising a three-phase winding abnormality index and a harmonic difference score of a frequency domain signal output by each phase winding, and inputting the data set into a preset prediction model to obtain the deformation state of the transformer winding. The present invention improves the accuracy of the detection result by calculating the three-phase winding abnormality index of a linear sweep frequency domain signal, using the three-phase winding abnormality index to measure the abnormality degree of the winding, and then using the prediction model to detect the deformation state of the winding.

Description

Technical Field

[0001] The present invention relates to the field of transformer detection, and in particular to a method and system for detecting the deformation state of a power transformer winding. Background Art

[0002] Power transformers are key equipment in power systems, responsible for the conversion and analysis of electrical energy. During long-term operation, transformer windings undergo mechanical deformation due to frequent temperature fluctuations, improper transportation and installation, and long-term overload operation. Traditional transformer status detection methods such as DC resistance testing and insulation resistance testing mainly focus on the electrical characteristics of the transformer and cannot effectively detect the mechanical deformation of the windings. With the development of technology, three transformer winding detection methods have emerged in the field of winding detection: harmonic analysis, frequency response analysis, and impedance measurement. The three detection methods are also used in combination to detect winding deformation.

[0003] A Chinese patent document with announcement number CN106152980B discloses a transformer winding deformation detection method and system. The method obtains the winding oil pressure data after a transformer fault occurs, and performs a function transformation on the data to obtain the harmonic component of the oil pressure data after the fault. The ratio of the harmonic component of the oil pressure data after the fault to the harmonic component of the oil pressure data before the fault is calculated to obtain the harmonic component amplitude ratio. The amplitude ratio is judged. If it exceeds a threshold, it is judged that the winding is deformed.

[0004] The above technology uses the harmonic analysis method to determine whether the winding is deformed. During the harmonic analysis process, a single fixed frequency scanning method will cause part of the collected electrical signal to be missing, affecting the detection of the winding deformation state, causing misjudgment and distortion of the deformation degree, resulting in poor accuracy in determining the deformation. Summary of the invention

[0005] In order to improve the accuracy of detecting transformer winding deformation, the present invention provides a method and system for detecting the deformation state of a power transformer winding.

[0006] In a first aspect, the present invention provides a method for detecting the deformation state of a power transformer winding, which adopts the following technical solution:

[0007] Acquire a data set, the data set includes three-phase winding abnormality indicators, and input the data set into a preset prediction model to obtain a deformation state of the transformer winding;

[0008] Among them, the calculation method of the three-phase winding abnormality index is:

[0009] Input a linear frequency sweep signal to the first end of each winding of the transformer to be tested, collect an output signal from the second end, convert the linear frequency sweep signal to obtain a linear frequency sweep frequency domain signal, convert the output signal to obtain an output frequency domain signal, and use the amplitude ratio of the linear frequency sweep frequency domain signal to the output frequency domain signal as the amplitude ratio;

[0010] In each phase winding, the abnormality degree of each frequency segment in each linear sweep frequency domain signal is calculated and corrected. The abnormality degree is positively correlated with the difference in amplitude ratio between the transformer to be tested and the corresponding winding of the normal transformer. The corrected abnormality degree is used as the optimal abnormality degree to calculate the abnormality index of the three-phase winding of the linear sweep frequency domain signal. The expression is:

[0011]

[0012] In the formula, Indicates the abnormal index of the three-phase winding of the power transformer. It represents the absolute value of the optimal abnormal degree difference between any two phase windings in the Ith frequency segment in the linear frequency sweep frequency domain signal. Represents the total number of frequency bins in the linear swept frequency domain signal, is the correlation coefficient, Q represents two-phase windings in the three-phase winding, A, B, C represent the A-phase winding, B-phase winding and C-phase winding in the three-phase winding respectively.

[0013] The effect is that by calculating the three-phase winding abnormality index of the linear swept frequency domain signal, the difference in the abnormality degree of any two-phase windings in each frequency band is referred to, the abnormality degree of the winding is measured using the three-phase winding abnormality index, and then the deformation state of the winding is detected using the prediction model, thereby improving the accuracy of the detection result.

[0014] Preferably, the expression of the abnormal degree of the frequency segment is:

[0015]

[0016] In the formula, Indicates the linear sweep frequency domain signal The abnormality of each frequency band, It represents the amplitude ratio of the linear swept frequency domain signal when the frequency is i, It represents the amplitude ratio of the normal transformer when the frequency of the linear sweep frequency domain signal is i, It represents the difference between the amplitude ratio of the linear frequency sweep frequency domain signal of the transformer to be tested when the frequency is i and the amplitude ratio of the adjacent frequency, It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency i and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is i. It represents the phase difference of the normal transformer when the frequency of the linear sweep frequency domain signal is i, i represents the frequency of the linear sweep frequency domain signal, ∆i represents the difference between adjacent frequencies of the linear sweep frequency domain signal, and t represents the sweep time of the linear sweep signal at each frequency.

[0017] The effect is that by calculating the abnormality degree of the frequency segment, the abnormal frequency segment can be screened out, which is convenient for adjusting the abnormal frequency segment. At the same time, the abnormality degree of the winding can be preliminarily understood through the abnormality degree of the frequency segment.

[0018] Preferably, the data set also includes the harmonic difference score of the frequency domain signal output by each phase winding, and the expression of the harmonic difference score is:

[0019]

[0020] In the formula, represents the harmonic difference fraction of the A-phase winding, Indicates the number of times the g-th harmonic appears, represents the amplitude of the gth harmonic in the output frequency domain signal, Represents the amplitude of the fundamental wave in the output frequency domain signal, is the number of times odd harmonics appear, is the number of times the even harmonics appear, and y represents the ordinal number.

[0021] The effect is that the harmonic difference score can be used to understand whether the transformer has an electrical fault or a winding abnormality, making it easier to determine the abnormal state of the transformer.

[0022] Preferably, the expression of the correlation coefficient is:

[0023] ,

[0024] In the formula, is the correlation coefficient, represents the current difference between the output ends of any two-phase windings, represents the voltage difference between any two-phase windings, Q represents two-phase windings in the three-phase winding, and A, B, and C represent the A-phase winding, B-phase winding, and C-phase winding in the three-phase winding respectively.

[0025] Preferably, the correction method of the abnormality degree is:

[0026] The linear frequency sweep frequency domain signal is divided into three categories, namely low-frequency signal, medium-frequency signal and high-frequency signal, wherein each category contains multiple frequency segments; the mean value of the abnormality degree of the frequency segment in each category is calculated, and the frequency segment with an abnormality degree greater than the mean value is regarded as an abnormal frequency segment;

[0027] The frequency sweep step length and frequency sweep time of the linear frequency sweep signal corresponding to the abnormal frequency segment are adjusted, and the abnormal degree of the frequency segment of the frequency domain signal corresponding to the adjusted frequency sweep signal is calculated to realize the correction of the abnormal degree.

[0028] The effect is that, by correcting the abnormality degree, the optimal abnormality degree can be obtained, thereby improving the accuracy of calculating the abnormality degree of the frequency segment.

[0029] Preferably, the expression of the frequency sweep step length after adjustment is:

[0030]

[0031] In the formula, Indicates the adjusted sweep frequency step length when the frequency is j in the abnormal frequency band, is the step size of the linear frequency sweep, It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the linear frequency-sweep frequency-domain signal of the transformer to be tested is j. It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the normal transformer linear sweep frequency domain signal is j. It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the frequency difference in a frequency band.

[0032] The effect is that by calculating the frequency sweep step length, the adjusted frequency band can be obtained, which facilitates the correction of the abnormal degree of the frequency band.

[0033] Preferably, the adjusted sweep duration expression is:

[0034]

[0035] In the formula, It indicates the adjusted sweep duration when the frequency is j in the abnormal frequency segment, t is the duration of the linear sweep, It represents the difference between the amplitude ratio of the linear frequency-sweep frequency domain signal of the transformer to be tested when the frequency is j and the amplitude ratio of the adjacent frequency, where the adjacent frequency is the frequency before or after the frequency j. It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency j and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the signal-to-noise ratio of the abnormal frequency band, Indicates the frequency difference in a frequency band.

[0036] In a second aspect, the present invention provides a system for detecting the deformation state of a power transformer winding, which adopts the following technical solution:

[0037] A power transformer winding deformation state detection system, a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned power transformer winding deformation state detection method is implemented.

[0038] The beneficial effect is that the above-mentioned method for detecting the deformation state of the winding of a power transformer is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a system based on the memory and the processor for easy use.

[0039] The present invention has the following technical effects:

[0040] 1. The state of the transformer winding is detected by multiple dimensions of three-phase winding abnormality indicators and harmonic difference scores. Compared with the traditional detection of winding deformation state using a single factor, the accuracy of detecting transformer winding deformation is improved.

[0041] 2. By calculating the abnormality degree of the frequency segment, the abnormal frequency segment is screened out, and then the abnormality degree of the abnormal frequency segment is corrected, and finally the optimal abnormality degree is obtained, which reduces the interference caused by missing data and improves the accuracy of the calculation results of the three-phase winding abnormality index. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0043] Figure 1 The present invention is a flow chart of a method for detecting a deformation state of a power transformer winding.

[0044] Figure 2 The present invention is a structural block diagram of a power transformer winding deformation state detection system. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0046] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0047] The embodiment of the present invention discloses a method for detecting the deformation state of a power transformer winding. Figure 1 , including the following steps, as follows:

[0048] S1: Input a linear frequency sweep signal to the first end of each winding of the transformer to be tested, collect an output signal from the second end, convert the linear frequency sweep signal to obtain a linear frequency sweep frequency domain signal, convert the output signal to obtain an output frequency domain signal, and use the amplitude ratio of the linear frequency sweep frequency domain signal to the output frequency domain signal as the amplitude ratio.

[0049] The linear swept frequency signal is a voltage signal and / or a current signal. The linear swept frequency signal is divided into three categories according to the frequency, namely, low-frequency signal, intermediate-frequency signal and high-frequency signal. The frequency of the low-frequency signal is (0-10kHz), the frequency of the intermediate-frequency signal is (10-100kHz), and the frequency of the high-frequency signal is (100kHz-2MHz). Each category contains multiple frequency bands, and the frequency step size in each frequency band is the same, and the sweep time of each frequency is also equal.

[0050] Exemplarily, for low-frequency signals, the frequency segments are divided into (0, 1, 2, 3, 4, 5), (6, 7, 8, 9, 10), the unit is kHz, when the frequency is 1 kHz, the sweep time is 1 second; when the frequency is 2 kHz, the sweep time is 1 second;…, when the frequency is 5 kHz, the sweep time is 1 second;….

[0051] The transformer to be tested is divided into A-phase winding, B-phase winding and C-phase winding, the first end of which is the input end and the second end is the output end. A signal generator is used to input linear frequency sweep signals to the A-phase winding, the B-phase winding and the C-phase winding respectively, and then an output signal is obtained from the corresponding output end, the linear frequency sweep signal is converted to obtain a linear frequency sweep frequency domain signal, and the output signal is converted to obtain an output frequency domain signal, so as to analyze and calculate the linear frequency sweep frequency domain signal and the output frequency domain signal.

[0052] S2: In each phase winding, the abnormality degree of each frequency segment in each linear frequency sweep frequency domain signal is calculated. The abnormality degree is positively correlated with the difference in amplitude ratio between the transformer under test and the corresponding winding of the normal transformer.

[0053] In one embodiment, the expression for the abnormality degree of a frequency segment is:

[0054]

[0055] In the formula, Indicates the linear sweep frequency domain signal The abnormality of each frequency band, It represents the amplitude ratio of the linear swept frequency domain signal when the frequency is i, It represents the amplitude ratio of the normal transformer when the frequency of the linear sweep frequency domain signal is i, It represents the difference between the amplitude ratio of the linear frequency sweep frequency domain signal of the transformer to be tested when the frequency is i and the amplitude ratio of the adjacent frequency, It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency i and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is i. It represents the phase difference of a normal transformer when the linear sweep frequency domain signal frequency is i, It represents the difference between two adjacent frequencies of the linear swept frequency domain signal, and t represents the sweeping time of the linear swept frequency signal at each frequency.

[0056] It indicates the response difference of the amplitude ratio between the transformer under test and the normal transformer. The larger the difference, the more serious the deformation of the transformer winding. It indicates the difference in the frequency domain signal amplitude-frequency response variation of the transformer under test and the normal transformer at different frequencies. The greater the difference, the more serious the transformer winding deformation. It indicates the difference in phase-frequency response between the transformer under test and the normal transformer. The greater the difference, the greater the difference in the impedance characteristics of the winding, and the greater the degree of abnormality of the winding. Indicates the linear sweep frequency domain signal The larger the value of the abnormality degree of each frequency band, the greater the possibility that the abnormality of the transformer winding is more serious.

[0057] In another embodiment, the expression of the abnormality degree of the frequency segment is:

[0058]

[0059] In the formula, Indicates the linear sweep frequency domain signal The abnormality of each frequency band, It represents the amplitude ratio of the linear swept frequency domain signal when the frequency is i, It represents the amplitude ratio of the normal transformer when the frequency of the linear sweep frequency domain signal is i, It represents the difference between the amplitude ratio of the linear frequency sweep frequency domain signal of the transformer to be tested when the frequency is i and the amplitude ratio of the adjacent frequency, It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency i and the amplitude ratio of the adjacent frequency, It represents the difference between two adjacent frequencies of the linear swept frequency domain signal, and t represents the sweeping time of the linear swept frequency signal at each frequency.

[0060] In yet another embodiment, the expression for the abnormality degree of a frequency segment is:

[0061]

[0062] In the formula, Indicates the linear sweep frequency domain signal The abnormality of each frequency band, It represents the amplitude ratio of the linear swept frequency domain signal when the frequency is i, It represents the amplitude ratio of the normal transformer when the frequency of the linear sweep frequency domain signal is i, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is i. It represents the phase difference of a normal transformer when the frequency of the linear swept frequency domain signal is i.

[0063] S3: Correct the abnormality level to obtain the optimal abnormality level.

[0064] In the linear frequency sweep measurement method, the interval between each frequency point and the sweep duration are equal. For some specific frequency bands, the sweep speed is too fast to respond in time, resulting in data loss, which further distorts the calculated abnormality. Therefore, it is necessary to correct the abnormality of the corresponding frequency band to improve the accuracy of the abnormality.

[0065] The correction method is: calculate the mean value of the abnormal degree of the frequency segment in each category, and take the frequency segment with an abnormal degree greater than the mean value as the abnormal frequency segment, and vice versa as the normal frequency segment; adjust the sweep step and sweep time of the linear sweep signal corresponding to the abnormal frequency segment, and calculate the abnormal degree of the frequency segment of the frequency domain signal corresponding to the adjusted sweep signal to achieve the correction of the abnormal degree. The abnormal degree greater than the mean value indicates that the calculated abnormal degree deviates greatly from the normal data, and the data of the corresponding frequency segment has a large error during the sweep measurement process.

[0066] The expression of the frequency sweep step after adjustment is: ;

[0067] In the formula, Indicates the adjusted sweep frequency step length when the frequency is j in the abnormal frequency band, is the step size of the linear frequency sweep, It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the linear frequency-sweep frequency-domain signal of the transformer to be tested is j. It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the normal transformer linear sweep frequency domain signal is j. It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the frequency difference in a frequency band.

[0068] The expression of the adjusted sweep duration is:

[0069]

[0070] In the formula, It indicates the adjusted sweep duration when the frequency is j in the abnormal frequency segment, t is the duration of the linear sweep, It represents the difference between the amplitude ratio of the linear frequency-sweep frequency domain signal of the transformer to be tested when the frequency is j and the amplitude ratio of the adjacent frequency, where the adjacent frequency is the frequency before or after the frequency j. It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency j and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the signal-to-noise ratio of the abnormal frequency band. The calculation method of the signal-to-noise ratio is the existing technology. It represents the frequency difference in the frequency band. The specific calculation steps will not be repeated here.

[0071] For example, assuming that the abnormal frequency range is (5, 10, 15, 20, 25), Here it is 5, after adjustment the first frequency is 5, and the calculation is , then the second frequency after adjustment is 5+6=11, and the calculation is (It should be noted that in the calculation 11 is between 10 and 15. is the difference between the amplitude ratios of frequency 10 and frequency 15. 11 is closest to 10. The phase difference is 10), then the third frequency after adjustment is 17; the calculation shows , then the third frequency after adjustment is 24, so the adjusted frequency range is (5, 11, 17, 24).

[0072] Similarly, we can calculate , calculated , calculated , , which means that when the three-phase winding is tested using the swept frequency signal of the adjusted frequency range (5, 11, 17, 24), the frequency is 5 and the sweep time is 2s; the frequency is 11 and the sweep time is 3s; the frequency is 17 and the sweep time is 2s; the frequency is 24 and the sweep time is 4s.

[0073] Then, the winding of the transformer to be measured is measured using the corrected frequency segment, and then the abnormality degree of the corrected frequency segment is calculated, and the corrected abnormality degree and the abnormality degree of the normal frequency segment are taken as the optimal abnormality degree.

[0074] S4: Calculate the three-phase winding abnormality index of the linear sweep frequency domain signal.

[0075] The three-phase winding abnormality index expression is:

[0076]

[0077] ,

[0078] In the formula, Indicates the abnormal index of the three-phase winding of the power transformer. It represents the absolute value of the optimal abnormal degree difference between any two phase windings in the Ith frequency segment in the linear frequency sweep frequency domain signal. Represents the total number of frequency bins in the linear swept frequency domain signal, is the correlation coefficient, Q represents two-phase windings in the three-phase winding, A represents A-phase winding in the three-phase winding, B represents B-phase winding in the three-phase winding, and C represents C-phase winding in the three-phase winding. The larger the value, the greater the abnormality of the transformer winding. represents the current difference between the output ends of any two-phase windings, Indicates the voltage difference between any two phase windings. The three-phase winding abnormality index is used to measure the degree of abnormality of the three-phase winding. The larger the value, the greater the possibility that the three-phase winding is in an abnormal state.

[0079] S5: Calculate the harmonic difference score of the frequency domain signal output by each phase winding.

[0080] The expression for the harmonic difference fraction is:

[0081]

[0082] In the formula, represents the harmonic difference fraction of the A-phase winding, Indicates the number of times the g-th harmonic appears, represents the amplitude of the gth harmonic in the output frequency domain signal, Represents the amplitude of the fundamental wave in the output frequency domain signal, is the number of times odd harmonics appear, is the number of times the even harmonics appear, and y represents the ordinal number. The larger the value, the greater the probability that the abnormality of phase A winding is an electrical fault such as turn-to-turn short circuit or ground fault. The smaller the value, the greater the probability that the abnormality of phase A winding is mechanical deformation. The calculation method of harmonic difference score of phase B winding and phase C winding is the same as that of phase A winding.

[0083] S6: Obtain a data set, and input the data set into a preset prediction model to obtain a deformation state of the transformer winding.

[0084] The data set includes the three-phase winding abnormality index and the harmonic difference score of the frequency domain signal output by each phase winding. A training set is constructed, which includes multiple data sets. A corresponding label is set for each data set. The label represents the deformation state of the transformer winding, which is divided into normal winding, mild deformation, moderate deformation and severe deformation. An SVM model is constructed, and the SVM model is trained using the training set to obtain a prediction model. The training method of the SVM model is a prior art and will not be repeated here. The data set obtained during the measurement process is input into the prediction model to obtain the deformation state of the transformer winding.

[0085] In other embodiments, the prediction model may also be a random forest model, a BP neural network model, or a convolutional neural network model.

[0086] Combination Figure 2 As shown, an embodiment of the present invention also discloses a system for detecting the deformation state of a power transformer winding, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for detecting the deformation state of a power transformer winding according to the present invention is implemented.

[0087] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0088] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high bandwidth memory HBM (High Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0089] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0090] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting deformation state of a power transformer winding, characterized in that: Includes steps: Acquire a data set, the data set includes three-phase winding abnormality indicators, and input the data set into a preset prediction model to obtain a deformation state of the transformer winding; The calculation method of the three-phase winding abnormality index is as follows: inputting a linear frequency sweep signal to the first end of each winding of the transformer to be tested, collecting an output signal from the second end, converting the linear frequency sweep signal to obtain a linear frequency sweep frequency domain signal, converting the output signal to obtain an output frequency domain signal, and taking the amplitude ratio of the linear frequency sweep frequency domain signal to the output frequency domain signal as the amplitude ratio; In each phase winding, the abnormality degree of each frequency segment in each linear sweep frequency domain signal is calculated and corrected. The abnormality degree is positively correlated with the difference in amplitude ratio between the transformer to be tested and the corresponding winding of the normal transformer. The corrected abnormality degree is used as the optimal abnormality degree to calculate the abnormality index of the three-phase winding of the linear sweep frequency domain signal. The expression is: In the formula, Indicates the abnormal index of the three-phase winding of the power transformer. It represents the absolute value of the optimal abnormal degree difference between any two phase windings in the Ith frequency segment in the linear frequency sweep frequency domain signal. Represents the total number of frequency bins in the linear swept frequency domain signal, is the correlation coefficient, Q represents two-phase windings in the three-phase winding, A, B, C represent the A-phase winding, B-phase winding and C-phase winding in the three-phase winding respectively.

2. A method for detecting deformation state of a power transformer winding according to claim 1, characterized in that: The expression of the abnormal degree of the frequency band is: In the formula, Indicates the linear sweep frequency domain signal The abnormality of each frequency band, It represents the amplitude ratio of the linear swept frequency domain signal when the frequency is i, It represents the amplitude ratio of the normal transformer when the frequency of the linear sweep frequency domain signal is i, It represents the difference between the amplitude ratio of the linear frequency sweep frequency domain signal of the transformer to be tested when the frequency is i and the amplitude ratio of the adjacent frequency, It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency i and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is i. It represents the phase difference of the normal transformer when the frequency of the linear sweep frequency domain signal is i, i represents the frequency of the linear sweep frequency domain signal, It represents the difference between adjacent frequencies of the linear swept frequency domain signal, and t represents the sweeping time of the linear swept frequency signal at each frequency.

3. A method for detecting deformation state of a power transformer winding according to claim 1, characterized in that: The data set also includes the harmonic difference score of the frequency domain signal output by each phase winding. The expression of the harmonic difference score is: In the formula, represents the harmonic difference fraction of the A-phase winding, Indicates the number of times the g-th harmonic appears, represents the amplitude of the gth harmonic in the output frequency domain signal, Represents the amplitude of the fundamental wave in the output frequency domain signal, is the number of times odd harmonics appear, is the number of times the even harmonics appear, and y represents the ordinal number.

4. A method for detecting deformation state of a power transformer winding according to claim 1, characterized in that: The expression of the correlation coefficient is: ， In the formula, is the correlation coefficient, represents the current difference between the output ends of any two-phase windings, represents the voltage difference between any two-phase windings, Q represents two-phase windings in a three-phase winding, A represents the A-phase winding in a three-phase winding, B represents the B-phase winding in a three-phase winding, and C represents the C-phase winding in a three-phase winding.

5. A method for detecting deformation state of a power transformer winding according to claim 1, characterized in that: The correction method for the abnormality degree is: The linear frequency sweep frequency domain signal is divided into three categories, namely low-frequency signal, medium-frequency signal and high-frequency signal, wherein each category contains multiple frequency segments; the mean value of the abnormality degree of the frequency segment in each category is calculated, and the frequency segment with an abnormality degree greater than the mean value is regarded as an abnormal frequency segment; The frequency sweep step length and frequency sweep time of the linear frequency sweep signal corresponding to the abnormal frequency segment are adjusted, and the abnormal degree of the frequency segment of the frequency domain signal corresponding to the adjusted frequency sweep signal is calculated to realize the correction of the abnormal degree.

6. A method for detecting deformation state of a power transformer winding according to claim 5, characterized in that: The expression of the frequency sweep step after adjustment is: In the formula, Indicates the adjusted sweep frequency step length when the frequency is j in the abnormal frequency band, is the step size of the linear frequency sweep, It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the linear frequency-sweep frequency-domain signal of the transformer to be tested is j. It represents the difference between the amplitude ratios of two adjacent frequencies when the frequency of the normal transformer linear sweep frequency domain signal is j. It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the frequency difference in a frequency band.

7. A method for detecting deformation state of a power transformer winding according to claim 5, characterized in that: The expression of the adjusted sweep duration is: In the formula, It indicates the adjusted sweep duration when the frequency is j in the abnormal frequency segment, t is the duration of the linear sweep, It represents the difference between the amplitude ratio of the linear frequency-sweep frequency domain signal of the transformer to be tested when the frequency is j and the amplitude ratio of the adjacent frequency, where the adjacent frequency is the frequency before or after the frequency j. It represents the difference between the amplitude ratio of the normal transformer linear sweep frequency domain signal at frequency j and the amplitude ratio of the adjacent frequency, It represents the phase difference of the transformer to be tested when the frequency of the linear sweep frequency domain signal is j and the frequency is the closest to k. It represents the phase difference of the normal transformer with the nearest frequency k when the linear frequency sweep signal frequency is j, Indicates the signal-to-noise ratio of the abnormal frequency band, Indicates the frequency difference in a frequency band.

8. A power transformer winding deformation state detection system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for detecting the deformation state of a power transformer winding according to any one of claims 1 to 7 is implemented.

Images