Live-line detection method for transformer winding deformation in GIS substations

Abstract

This application relates to the field of live winding deformation detection technology, specifically to a method for live transformer winding deformation detection applicable to GIS substations. The method includes: generating amplitude-frequency response curves of the three-phase transformer by calculating a transfer function; obtaining the response amplitude at each frequency point in the amplitude-frequency response curves of phases A, B, and C of the transformer within each measurement cycle; analyzing the sharpness of local frequency domain waveform characteristics in the amplitude-frequency response curves and the differences between amplitude-frequency responses of each phase to obtain the complexity of sharpness differences among the transformer phases; evaluating the significant degree of abnormal changes in the waveform sharpness characteristics of the amplitude-frequency response curves of each phase in the time domain; and determining the winding deformation monitoring value for each measurement cycle based on the average level of the complexity of sharpness differences among the three phases of the transformer; using the winding deformation monitoring value to obtain a deformation judgment threshold, thereby detecting live transformer winding deformation. This application can improve the accuracy of live transformer winding deformation detection.

Description

Technical Field

[0001] This application relates to the field of live detection technology for winding deformation, specifically to a live detection method for transformer winding deformation applicable to GIS substations. Background Technology

[0002] Transformers are one of the key pieces of equipment in the power system of GIS (Gas Insulated Substation). When a transformer is subjected to a short-circuit current from the power grid, the strong electromagnetic force can cause deformation of the transformer windings, which can easily lead to damage to the winding turns, a reduction in local insulation distance, and partial discharge in the damaged area, accelerating insulation aging and causing the damaged area to expand continuously. Currently, transformer winding deformation is one of the common hidden dangers of transformer failure. Transformer winding deformation detection is mainly divided into offline and live detection. Offline detection of transformer winding deformation requires shutting down the transformer. However, due to the heavy power transmission load of the power grid, there is a lack of time for offline detection of transformer winding deformation, which can easily lead to a large number of transformers operating in a damaged state for a long time, posing a high safety hazard.

[0003] In existing methods for detecting transformer winding deformation under live conditions, the Frequency Response Analysis (FRA) method is used for online detection. This typically follows the relevant standards in the power industry standard DL / T911-2016, "Frequency Response Analysis Method for Power Transformer Winding Deformation," using correlation coefficients to determine the deformation. Currently, existing FRA methods generally combine horizontal and vertical comparison methods to perform correlation analysis on the amplitude-frequency response curves of the transformer's three-phase power, thereby detecting winding deformation in GIS substations. However, because transformer winding deformation is generally complex, and existing technologies do not fully consider abnormal changes in the frequency domain waveform characteristics of the amplitude-frequency response curve, they cannot accurately measure the degree of change in the amplitude-frequency response characteristics of the transformer's three-phase power, resulting in poor accuracy in detecting transformer winding deformation under live conditions. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a method for detecting transformer winding deformation under live conditions in GIS substations, thereby resolving the existing issues.

[0005] The live-line detection method for transformer winding deformation in GIS substations proposed in this application adopts the following technical solution:

[0006] One embodiment of this application provides a method for live-line detection of transformer winding deformation in GIS substations, comprising the following steps:

[0007] The amplitude-frequency response curves of the three-phase electricity of the transformer are generated by calculating the transfer function, and the response amplitudes of each frequency point in the amplitude-frequency response curves of phases A, B, and C of the transformer are obtained in each measurement cycle.

[0008] By analyzing the sharpness of local frequency domain waveform characteristics in the amplitude-frequency response curve and the differences between amplitude-frequency responses of each phase, the sharpness difference complexity of each phase of the transformer is obtained.

[0009] The significance of the abnormal changes in the waveform sharpness of each phase amplitude-frequency response curve of the transformer in the time domain is evaluated. Combined with the average level of the sharpness difference complexity of the three phases of the transformer, the winding deformation monitoring value of each measurement cycle is determined.

[0010] The deformation judgment threshold is obtained by using the winding deformation monitoring value, and then the transformer winding deformation is detected to be energized.

[0011] Preferably, before obtaining the sharp difference complexity of each phase of the transformer, multiple local frequency points of each frequency point are extracted, and the response amplitudes of each frequency point and all its local frequency points are arranged in ascending order of frequency to form a local amplitude sequence of each frequency point. The kurtosis of each local amplitude sequence is calculated, and the kurtosis corresponding to all frequency points in the amplitude-frequency response curve of each phase is arranged in ascending order of frequency to form a sharp change sequence of each phase.

[0012] Preferably, the formula for obtaining the complexity of the sharp differences between each phase of the transformer is:

[0013] In the formula, The sharp difference complexity of phase A of the transformer. and These are the permutation entropies of the first and second sharp difference sequences of phase A of the transformer, respectively. and These are the dispersion of the first sharp difference sequence and the second sharp difference sequence of transformer phase A, respectively. The absolute difference sequences between the sharp change sequence of transformer phase A and the sharp change sequences of phases B and C are denoted as the first sharp difference sequence and the second sharp difference sequence of transformer phase A.

[0014] Preferably, before acquiring the winding deformation monitoring value for each measurement cycle, multiple adjacent measurement cycles of each measurement cycle are extracted, and the normalized result of the similarity between the sharp change sequence of transformer phase A in each measurement cycle and the sharp change sequence of transformer phase A in each adjacent measurement cycle is calculated. The normalized results of all similarities are then combined into a similarity set of transformer phase A in each measurement cycle.

[0015] Preferably, the multiple frequency points that are closest to each frequency point in the amplitude-frequency response curve of each phase of the transformer are recorded as each local frequency point of each frequency point, and the multiple measurement cycles that are closest to the time interval of each measurement cycle are recorded as each adjacent measurement cycle of each measurement cycle.

[0016] Preferably, for each measurement cycle, the abnormal change evaluation value of the waveform sharpness characteristics of each phase amplitude-frequency response curve of the transformer in the time domain angle is calculated: In the formula, This is the evaluation value for the abnormal change in the time-domain angle of the waveform sharpness characteristic of the transformer's A-phase amplitude-frequency response curve. It is an exponential function with the natural constant as its base. Let be the mean of the set of similarities for phase A of the transformer. Let represent the degree of discretization of the similarity set of phase A of the transformer. For preset amplification parameters, To avoid extremely small positive numbers with a denominator of zero.

[0017] Preferably, for each measurement cycle, the mean values ​​of the sharp difference complexity of phases A, B, and C of the transformer and the mean values ​​of the abnormal change assessment values ​​corresponding to phases A, B, and C are obtained, and the two mean values ​​are normalized respectively. The sum of the normalized results of the two mean values ​​is used as the winding deformation monitoring value for each measurement cycle.

[0018] Preferably, the first T measurement cycles after the initial operation of the transformer are pre-set to be normal operation in the absence of winding deformation defects. After the transformer starts operating for T measurement cycles, the maximum value of the winding deformation monitoring value of the T most recent measurement cycles before each measurement cycle is taken as the deformation judgment threshold, and the measurement cycle corresponding to the winding deformation monitoring value that is higher than the deformation judgment threshold is taken as the measurement cycle with winding deformation defects. Here, T is a preset value.

[0019] Preferably, the deformation judgment threshold is modified by eliminating measurement cycles with winding deformation defects before the current measurement cycle, and using the maximum value of the winding deformation monitoring value of the most recent T measurement cycles without winding deformation defects before the current measurement cycle as the deformation judgment threshold of the current measurement cycle.

[0020] Preferably, if the winding deformation monitoring value of the current measurement cycle is greater than the deformation judgment threshold, then the transformer in the GIS substation has a winding deformation defect in the current measurement cycle; otherwise, the transformer in the GIS substation does not have a winding deformation defect.

[0021] This application has at least the following beneficial effects:

[0022] This application analyzes the sharpness of local frequency domain waveform characteristics in the amplitude-frequency response curve, and combines the complexity and fluctuation of the difference in waveform sharpness to more accurately measure and analyze the complexity of the waveform sharpness difference between the amplitude-frequency response curves of each phase of the transformer and the amplitude-frequency response curves of other phases, and more clearly reflects the difference in amplitude-frequency response characteristics between phases when there are winding deformation defects in the transformer.

[0023] In this application, by extracting the similarity set of each phase of the transformer, the abnormal changes of the waveform sharpness of each phase of the transformer amplitude-frequency response curve in the time domain angle can be more accurately identified. This clearly illustrates the abnormal changes of the frequency domain waveform characteristics in the amplitude-frequency response curve in the time domain angle, which is beneficial to improving the accuracy of subsequent live detection of transformer winding deformation.

[0024] This application measures the overall abnormal characteristics of transformer winding deformation more accurately by considering the abnormal changes in the waveform sharpness of each phase amplitude-frequency response curve in the time domain angle, and by simultaneously considering the complexity of the waveform sharpness differences between each phase amplitude-frequency response curve and the amplitude-frequency response curves of other phases. Attached Figure Description

[0025] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating the steps of the live-line detection method for transformer winding deformation in GIS substations provided in this application;

[0027] Figure 2 This is a schematic diagram of the structure for frequency response voltage signal acquisition provided in this application;

[0028] Figure 3 This is a schematic diagram of the detection of winding deformation monitoring values ​​provided in this application. Detailed Implementation

[0029] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the transformer winding deformation live-line detection method applicable to GIS substations proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0030] Unless otherwise defined, terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the article or device that includes said element. Furthermore, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0031] The following description, in conjunction with the accompanying drawings, details the specific scheme of the live-line detection method for transformer winding deformation applicable to GIS substations provided in this application.

[0032] This application provides an embodiment of a live-line detection method for transformer winding deformation in GIS substations. For details, please refer to [link to relevant documentation]. Figure 1 This includes the following steps:

[0033] Step 1: Calculate the amplitude-frequency response curve of the three-phase power of the transformer using the transfer function, and obtain the response amplitude of each frequency point in the amplitude-frequency response curve of phase A, phase B, and phase C of the transformer in each measurement cycle.

[0034] In order to accurately detect the deformation of transformer windings in GIS substations under live conditions, it is necessary to fully analyze the abnormal changes in the frequency domain waveform characteristics in the amplitude-frequency response curve. This will allow for a more accurate measurement of the degree of change in the amplitude-frequency response characteristics of the transformer's three-phase electricity, thus avoiding affecting the accuracy of live detection of transformer winding deformation.

[0035] The FRA frequency response analysis method was used to detect the deformation of the transformer windings of a transformer in a GIS substation. Excitation sensors were installed on the high-voltage side core grounding wire of the transformer, and three response sensors were installed at the ends of the cable terminals of each phase on the high-voltage side. A 100V sweep voltage signal was injected into the excitation sensors by a signal source. According to the requirements of the power industry standard DL / T911-2016 "Frequency Response Analysis Method for Winding Deformation of Power Transformers", the sweep frequency range was set to 1kHz-1MHz, and the number of sweep frequency points was ≥2000. In this embodiment, the number of sweep frequency points was 2048, and the measurement cycle for one sweep was 13 seconds. Based on the principle of magnetic field coupling, the excitation sensors coupled the sweep voltage signal into the high-voltage windings. The response sensors installed at the ends of the A, B, and C phase cable terminals on the high-voltage side output the frequency response voltage signals generated in the transformer windings.

[0036] Furthermore, the frequency response voltage signal output by the response sensor is first filtered by a high-pass filter to remove power frequency noise, and then by a low-pass filter to remove high-frequency noise above 1MHz. After processing by the low-pass filter, the frequency response voltage signal enters the signal amplifier, which amplifies the frequency response voltage signal from tens to hundreds of microvolts to the millivolt level. The response sensor then collects the frequency response voltage signal of the three-phase transformer after processing by the signal amplifier in each measurement cycle. , , , respectively represent the frequency response voltage signals of phase A, phase B, and phase C on the high-voltage side of the transformer at the k-th frequency point within one measurement cycle.

[0037] A transfer function is constructed based on the swept-frequency voltage signal injected into the excitation sensor by the signal source and the frequency response voltage signal of the three-phase electricity of the transformer. The amplitude-frequency response curve of the three-phase electricity of the transformer is then calculated using this transfer function. Specifically, , , ,in, , , These represent the response amplitudes at the k-th frequency point in the amplitude-frequency response curves of phases A, B, and C of the transformer within one measurement cycle. The signal source injects a sweep voltage signal at the k-th frequency point into the excitation sensor.

[0038] Therefore, the response amplitude at the k-th frequency point in the amplitude-frequency response curves of phases A, B, and C of the transformer within each measurement cycle is calculated using the transfer function.

[0039] Specifically, the structural schematic diagram for frequency response voltage signal acquisition is as follows: Figure 2 As shown.

[0040] Step 2: Analyze the sharpness of the local frequency domain waveform characteristics in the amplitude-frequency response curve and the differences between the amplitude-frequency responses of each phase to obtain the sharpness difference complexity of each phase of the transformer.

[0041] Existing technologies combine horizontal and vertical comparison methods to perform correlation analysis on the amplitude-frequency response curves of three-phase transformers. However, since transformer winding deformation is generally complex, and existing technologies do not fully consider abnormal changes in the frequency domain waveform characteristics of the amplitude-frequency response curve, they cannot accurately measure the degree of change in the amplitude-frequency response characteristics of the three-phase transformer, thus affecting the accuracy of live-line detection of transformer winding deformation in GIS substations. Therefore, to improve the accuracy of live-line detection of transformer winding deformation, it is necessary to perform feature analysis on abnormal changes in the frequency domain waveform characteristics of the amplitude-frequency response curve.

[0042] Therefore, taking the amplitude-frequency response curve of transformer phase A within each measurement cycle as an example for analysis, the M frequency points in the amplitude-frequency response curve of transformer phase A that are closest to the k-th frequency point in terms of Euclidean distance are denoted as the M local frequency points of the k-th frequency point. The value of M can be in the range of 20-30, and in this embodiment, it is taken as 20. In order to analyze the local waveform characteristics in the amplitude-frequency response curve, the sequence of response amplitudes of the k-th frequency point and its M local frequency points in ascending order of frequency is denoted as the local amplitude sequence of each frequency point in the amplitude-frequency response curve, reflecting the local frequency domain waveform characteristics of each frequency point in the amplitude-frequency response curve.

[0043] Furthermore, the kurtosis of the local amplitude sequence at each frequency point is calculated using the formula for sample kurtosis. This reflects the sharpness of the local frequency domain waveform characteristics in the amplitude-frequency response curve. The greater the kurtosis, the sharper the local waveform change at that frequency point in the amplitude-frequency response curve. To analyze the differences in amplitude-frequency response characteristics among the transformer phases, the kurtosis corresponding to all frequency points in the amplitude-frequency response curves of each phase of the transformer is arranged in ascending order of frequency. This sequence is denoted as the sharpness change sequence of each phase of the transformer, reflecting the changing characteristics of the waveform sharpness in the amplitude-frequency response curves of each phase of the transformer.

[0044] Under normal circumstances, when there are no winding deformation defects in the transformer in a GIS substation, the differences between the three-phase amplitude-frequency response curves are small. However, when there are winding deformation defects in the transformer, the distributed inductance and distributed capacitance of the three-phase coils of the transformer will show significant differences, resulting in significant differences in the amplitude-frequency response characteristics between the phases of the transformer.

[0045] Therefore, for the sharp change sequence of each phase of the transformer in each measurement cycle, the absolute difference sequence between the sharp change sequence of each phase of the transformer and the sharp change sequences of the other two phases is calculated and recorded as the first sharp difference sequence and the second sharp difference sequence of each phase of the transformer, respectively. These reflect the degree of difference in amplitude-frequency response characteristics between each phase of the transformer. For example, taking phase A of the transformer as an example, the absolute difference sequence between the sharp change sequence of phase A of the transformer and the sharp change sequences of phases B and C are recorded as the first sharp difference sequence and the second sharp difference sequence of phase A of the transformer, respectively.

[0046] Generally, if a transformer in a GIS substation has winding deformation defects, it will cause complex differences in the amplitude-frequency response characteristics of different frequency bands. Therefore, in order to accurately reflect the complexity of the first sharp difference sequence and the second sharp difference sequence, the preset embedding dimension parameter is set to 5 and the preset delay step size parameter is set to 2. Based on the preset embedding dimension parameter and the preset delay step size parameter, the permutation entropy of the first sharp difference sequence and the second sharp difference sequence is calculated respectively. The larger the permutation entropy, the more significant the complex differences in the amplitude-frequency response characteristics of different frequency bands.

[0047] Based on the above analysis, for each measurement cycle, the complexity of the sharp differences between each phase of the transformer is calculated. In this embodiment, the specific calculation formula is as follows:

[0048] In the formula, The sharp difference complexity of phase A of the transformer. and These are the permutation entropies of the first and second sharp difference sequences of phase A of the transformer, respectively. and The dispersion of the first sharp difference sequence and the second sharp difference sequence of transformer phase A are respectively. The absolute difference sequences between the sharp change sequence of transformer phase A and the sharp change sequences of phase B and phase C are denoted as the first sharp difference sequence and the second sharp difference sequence of transformer phase A. The dispersion can be measured by variance or standard deviation. In this embodiment, variance is used for measurement.

[0049] According to the above formula, since the sharp difference sequence can reflect the degree of difference in amplitude-frequency response characteristics between different phases of the transformer, the discreteness of the sharp difference sequence characterizes the variation and fluctuation of the waveform sharpness, and the permutation entropy of the sharp difference sequence characterizes the complexity of the variation of amplitude-frequency response characteristics in different frequency bands, the discreteness and permutation entropy are fused by multiplication, and the fusion results of the first sharp difference sequence and the second sharp difference sequence are added together to measure the complexity of the sharp difference in amplitude-frequency response curves of each phase of the transformer.

[0050] Among them, the sharp difference complexity reflects the complexity of the sharp difference between the amplitude-frequency response curves of each phase of the transformer and the amplitude-frequency response curves of the other phases. The greater the sharp difference complexity, the greater the complexity of the sharp difference between the amplitude-frequency response curves of each phase of the transformer and the amplitude-frequency response curves of the other phases. It can more clearly reflect the difference characteristics between the amplitude-frequency response curves of the three phases of the transformer when the transformer winding is deformed, which is conducive to improving the accuracy of subsequent detection of transformer winding deformation under live conditions.

[0051] Step 3: Evaluate the degree of abnormal change in the waveform sharpness of each phase amplitude-frequency response curve of the transformer in the time domain, and determine the winding deformation monitoring value for each measurement cycle by combining the average level of the sharpness difference complexity of the three phases of the transformer.

[0052] To further analyze the abnormal changes in the frequency domain waveform characteristics of the amplitude-frequency response curve from a time domain perspective, the N measurement cycles with the closest time interval between each measurement cycle are taken as the N nearest neighbor measurement cycles for each measurement cycle. The value of N can range from 2 to 4; in this embodiment, it is set to 3.

[0053] Furthermore, taking the sharp change sequence of transformer phase A as an example, the cosine similarity between the sharp change sequence of transformer phase A in each measurement cycle and the sharp change sequence of transformer phase A in each of its neighboring measurement cycles is calculated, and the cosine similarity is normalized and mapped to... Within the interval, the set of normalized results of all cosine similarities is denoted as the similarity set of transformer phase A within each measurement period. In this embodiment, the maximum value normalization method is used for normalization. In practical applications, implementers can choose other existing normalization methods; no special restrictions are imposed in this embodiment.

[0054] Similarly, by calculating the cosine similarity of the sharp change sequence of each phase of the transformer in each measurement cycle, the similarity set of each phase of the transformer in each measurement cycle can be obtained. If the average level of all cosine similarities in the similarity set of each phase of the transformer in a certain measurement cycle is smaller, and the dispersion of all cosine similarities in the similarity set is greater, it can better reflect the abnormal change of the waveform sharp characteristics in the amplitude-frequency response curve of each phase of the transformer in the time domain angle within that measurement cycle. At this time, the transformer is more likely to have winding deformation defects.

[0055] Therefore, based on the above analysis, for each measurement cycle, the abnormal change evaluation value of the waveform sharpness characteristics of each phase amplitude-frequency response curve of the transformer in the time domain angle is calculated. In this embodiment, the specific formula is as follows:

[0056] In the formula, This is the evaluation value for the abnormal change in the time-domain angle of the waveform sharpness characteristic of the transformer's A-phase amplitude-frequency response curve. It is an exponential function with the natural constant as its base. Let be the mean of the set of similarities for phase A of the transformer. The degree of dispersion of the similarity set of transformer phase A can be measured by variance or standard deviation. In this embodiment, variance is used for measurement. To avoid extremely small positive numbers with a denominator of zero, this embodiment uses a value of 0.001. This is intended to prevent calculation errors caused by a zero denominator. Implementers can set the value according to their actual needs. The preset amplification parameter is used to avoid the difference in the result of the exponential negative mapping being too small to clearly show the abnormal changes in the waveform's sharp features in the time domain angle. The value range is 10-15, and in this embodiment, the value is 10.

[0057] In the above formula, when the winding deforms, the mean of the similarity set decreases, while the dispersion of the similarity set increases. Therefore, a division method is used to fuse the mean and dispersion of the similarity set. Based on the above fusion result, an exponential function is used for negative mapping processing, combined with preset amplification parameters, to evaluate the abnormal changes of sharp waveform features in the time domain.

[0058] It is understandable that the abnormal change assessment value reflects the abnormal change of the waveform sharpness of each phase amplitude-frequency response curve of the transformer in the time domain angle. The more significant the abnormal change of the waveform sharpness in the time domain angle, the clearer it can explain the abnormal change of the frequency domain waveform characteristics in the amplitude-frequency response curve in the time domain angle. At this time, the transformer is more likely to have winding deformation defects.

[0059] Because of winding deformation defects in the transformer, significant differences in the amplitude-frequency response characteristics between different phases can occur, resulting in significant abnormal changes in the frequency domain waveform characteristics of the amplitude-frequency response curve in the time domain. Therefore, to more accurately perform live-line testing of the transformer windings, the winding deformation monitoring value for each measurement cycle is measured by combining the sharp difference complexity and the abnormal change evaluation value. Specifically, for each measurement cycle, the mean of the sharp difference complexity of phases A, B, and C of the transformer is calculated, as well as the mean of the abnormal change evaluation value corresponding to phases A, B, and C. These mean values ​​are then normalized. In this embodiment, the maximum value normalization method is used, and the specific process is known to those skilled in the art and will not be elaborated upon in this embodiment. Furthermore, the summation between the normalized result of the mean of the sharp difference complexity and the normalized result of the mean of the abnormal change assessment value is recorded as the winding deformation monitoring value for each measurement cycle, reflecting the overall abnormal characteristics of the transformer winding deformation. The larger the winding deformation monitoring value, the more significant the overall abnormal characteristics of the transformer winding deformation. At this time, the transformer in the GIS substation is more likely to have transformer winding deformation defects.

[0060] Step 4: Use the winding deformation monitoring values ​​to obtain the deformation judgment threshold, and then detect the transformer winding deformation and energization.

[0061] Furthermore, in order to achieve live detection of transformer winding deformation in GIS substations, a cover inspection is conducted on the transformer before it starts operation to confirm that there are no winding deformation defects before operation. The transformer operates normally in the early stages after operation without winding deformation defects. Therefore, the monitoring results and upper limit of the fluctuation range of winding deformation monitoring values ​​in multiple measurement cycles during the early stages of operation are within the normal fluctuation range. However, if the winding deformation monitoring value of the current measurement cycle exceeds the upper limit of the normal fluctuation range of winding deformation monitoring values, there is a high probability that the transformer winding deformation defect has occurred in the GIS substation during the current measurement cycle.

[0062] Therefore, the transformer is pre-set to operate normally for the first T measurement cycles after initial operation, assuming no winding deformation defects. Here, T is a preset value ranging from 80 to 100; in this embodiment, T is 80. After the transformer has operated for T measurement cycles, the maximum value of the winding deformation monitoring values ​​from the T most recent measurement cycles preceding each measurement cycle is used as the deformation judgment threshold. Measurement cycles with winding deformation monitoring values ​​exceeding the deformation judgment threshold are considered to have winding deformation defects. Furthermore, the deformation judgment threshold is updated and corrected. Using one measurement cycle as a step, measurement cycles with winding deformation defects preceding the current measurement cycle are eliminated. The maximum value of the winding deformation monitoring values ​​from the T most recent measurement cycles without winding deformation defects preceding the current measurement cycle is recorded as the deformation judgment threshold for the current measurement cycle. The maximum value of the winding deformation monitoring values ​​from the latest measurement cycle without winding deformation defects is used to correct the deformation judgment threshold, avoiding long-term slow reference drift caused by equipment aging or environmental factors.

[0063] If the winding deformation monitoring value of the current measurement cycle is greater than the deformation judgment threshold, it is considered that the transformer in the GIS substation has a winding deformation defect in the current measurement cycle; if the winding deformation monitoring value of the current measurement cycle is less than or equal to the deformation judgment threshold, it is considered that the transformer in the GIS substation does not have a winding deformation defect in the current measurement cycle.

[0064] Specifically, in this embodiment, the detection diagram of the winding deformation monitoring value is as follows: Figure 3 As shown, the horizontal axis represents the measurement cycle number, and the vertical axis represents the winding deformation monitoring value. The deformation judgment threshold obtained during the detection process is 9.5153. The specific changes in the winding deformation monitoring value are as follows: Figure 3 As shown.

[0065] It is understood that references to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include the specific features, structures, or characteristics described in connection with that embodiment. Therefore, the appearance of phrases such as "in one embodiment," "in some embodiments," "in other embodiments," or "in still other embodiments" in different parts of this specification does not necessarily refer to the same embodiment, but rather means "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0066] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous. Moreover, the sequence numbers of the steps in the embodiments do not imply a specific order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments in this specification.

[0067] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A live-line detection method for transformer winding deformation in GIS substations, characterized in that, Includes the following steps: The amplitude-frequency response curves of the three-phase electricity of the transformer are generated by calculating the transfer function, and the response amplitudes of each frequency point in the amplitude-frequency response curves of phases A, B, and C of the transformer are obtained in each measurement cycle. By analyzing the sharpness of local frequency domain waveform characteristics in the amplitude-frequency response curve and the differences between amplitude-frequency responses of each phase, the sharpness difference complexity of each phase of the transformer is obtained. The significance of the abnormal changes in the waveform sharpness of each phase amplitude-frequency response curve of the transformer in the time domain is evaluated. Combined with the average level of the sharpness difference complexity of the three phases of the transformer, the winding deformation monitoring value of each measurement cycle is determined. The deformation judgment threshold is obtained by using the winding deformation monitoring value, and then the transformer winding deformation is detected to be energized.

2. The method for detecting live deformation of transformer windings in GIS substations as described in claim 1, characterized in that, Before obtaining the sharp difference complexity of each phase of the transformer, multiple local frequency points are extracted from each frequency point. The response amplitudes of each frequency point and all its local frequency points are arranged in ascending order of frequency to form a local amplitude sequence of each frequency point. The kurtosis of each local amplitude sequence is calculated. The kurtosis corresponding to all frequency points in the amplitude-frequency response curve of each phase is arranged in ascending order of frequency to form a sharp change sequence of each phase.

3. The method for detecting transformer winding deformation under live conditions in GIS substations as described in claim 2, characterized in that, The formula for obtaining the complexity of the sharp differences between each phase of the transformer is: In the formula, The sharp difference complexity of phase A of the transformer. and These are the permutation entropies of the first and second sharp difference sequences of phase A of the transformer, respectively. and These are the dispersion of the first sharp difference sequence and the second sharp difference sequence of transformer phase A, respectively. The absolute difference sequences between the sharp change sequence of transformer phase A and the sharp change sequences of phases B and C are denoted as the first sharp difference sequence and the second sharp difference sequence of transformer phase A.

4. The method for detecting live deformation of transformer windings in GIS substations as described in claim 2, characterized in that, Before acquiring the winding deformation monitoring values ​​for each measurement cycle, multiple adjacent measurement cycles are extracted for each measurement cycle. The normalized result of the similarity between the sharp change sequence of transformer phase A in each measurement cycle and the sharp change sequence of transformer phase A in each adjacent measurement cycle is calculated. All the normalized results of the similarity are combined into a similarity set of transformer phase A in each measurement cycle.

5. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 4, characterized in that, The frequency points closest to each frequency point in the amplitude-frequency response curve of each phase of the transformer are recorded as each local frequency point of each frequency point, and the measurement cycles closest to the time interval of each measurement cycle are recorded as each neighboring measurement cycle of each measurement cycle.

6. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 4, characterized in that, For each measurement cycle, calculate the evaluation value of the abnormal change in the waveform sharpness of each phase amplitude-frequency response curve of the transformer in the time domain angle: In the formula, This is the evaluation value for the abnormal change in the time-domain angle of the waveform sharpness characteristic of the transformer's A-phase amplitude-frequency response curve. It is an exponential function with the natural constant as its base. Let be the mean of the set of similarities for phase A of the transformer. Let represent the degree of discretization of the similarity set of phase A of the transformer. For preset amplification parameters, To avoid extremely small positive numbers with a denominator of zero.

7. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 6, characterized in that, For each measurement cycle, the mean values ​​of the sharp difference complexity of phases A, B, and C of the transformer and the mean values ​​of the abnormal change assessment values ​​corresponding to phases A, B, and C are obtained. The two mean values ​​are normalized respectively, and the sum of the normalized results of the two mean values ​​is used as the winding deformation monitoring value for each measurement cycle.

8. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 1, characterized in that, The transformer is pre-set to operate normally for the first T measurement cycles after initial operation, under conditions where there are no winding deformation defects. After the transformer has been running for T measurement cycles, the maximum value of the winding deformation monitoring value of the T most recent measurement cycles before each measurement cycle is taken as the deformation judgment threshold, and the measurement cycle corresponding to the winding deformation monitoring value that is higher than the deformation judgment threshold is taken as the measurement cycle where there is a winding deformation defect. Here, T is a preset value.

9. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 8, characterized in that, The deformation judgment threshold is corrected by eliminating measurement cycles with winding deformation defects before the current measurement cycle. The maximum value of the winding deformation monitoring value of the most recent T measurement cycles without winding deformation defects before the current measurement cycle is taken as the deformation judgment threshold of the current measurement cycle.

10. The method for live-line detection of transformer winding deformation in GIS substations as described in claim 9, characterized in that, If the winding deformation monitoring value of the current measurement cycle is greater than the deformation judgment threshold, then the transformer in the GIS substation has a winding deformation defect in the current measurement cycle; otherwise, the transformer in the GIS substation does not have a winding deformation defect.

Images