Method for evaluating severity of insulation surface defect of GIS (Gas Insulated Switchgear) equipment

Abstract

The invention discloses a GIS equipment insulation surface defect severity assessment method, and relates to the electromechanical equipment discharge fault diagnosis technology field, and the method comprises the following steps: building a simulation circuit model, and obtaining five continuous ultrahigh frequency spectrum data; calculating a forward normalized pulse mean value; calculating four characteristic parameters according to the five continuous ultrahigh frequency spectrum data and the forward normalized pulse mean value, wherein the four characteristic parameters comprise a parameter A, a parameter B, a parameter C and a parameter D; according to the value range of the four characteristic parameters, obtaining a defect severity degree evaluation result; on the basis of a microcosmic discharge mechanism, the relationship among the ultrahigh frequency pulse number, the pulse mean value and the property of the defect is mined, four dimensionless physical quantities are extracted as the basis of severity evaluation, and the universality of the method is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of electromechanical equipment discharge fault diagnosis, and in particular to a method for evaluating the severity of insulation surface defects of GIS equipment. Background Art

[0002] Insulation surface defects are common defects in GIS equipment and pose a serious threat to the insulation performance of the equipment. During the process of discharge along the surface defects until breakdown, the insulating medium deteriorates, and the ultra-high frequency signal will continue to change during this process. As a result, the severity assessment of insulation defects in GIS equipment based on ultra-high frequency signals has always been a difficulty in current research.

[0003] In the existing technology, data-driven evaluation methods based on machine learning or deep learning have been widely used in previous studies, but data-driven evaluation methods based on machine learning or deep learning have certain defects.

[0004] First, if you want to make the algorithm's evaluation results more accurate, a large amount of experimental and field data is needed to train the neural network, which not only increases the difficulty of model training, but also increases the cost of deployment and maintenance; second, the amount of data in the field and experiments is limited, and the trained algorithm is usually only applicable to data from specific detection instruments, which restricts the universality of the method; finally, the calibration of data labels at different degradation stages in the training set is often based on empirical judgment, which makes it difficult to explain the reasons for the signal changes from a mechanistic level, resulting in weak theoretical support for the algorithm.

[0005] Therefore, a method for evaluating the severity of insulation surface defects of GIS equipment is provided to solve the above problems. Summary of the Invention

[0006] The purpose of the present invention is to provide a method for evaluating the severity of insulation defects along the surface of GIS equipment. Based on the current problems faced in defect severity assessment, several ultra-high frequency signal characteristic parameters that best reflect the discharge characteristics at different stages are extracted, so that the characteristic parameters can be integrated with the discharge mechanism and used to formulate defect severity assessment standards.

[0007] To achieve the above object, the present invention provides a method for evaluating the severity of insulation surface defects of GIS equipment, comprising the following steps:

[0008] S1: Build a simulation circuit model and obtain five continuous UHF spectrum data;

[0009] S2: Calculate the forward normalized pulse mean;

[0010] S3: Calculate four characteristic parameters based on five consecutive UHF spectrum data and the forward normalized pulse mean value. The four characteristic parameters include parameter A, parameter B, parameter C, and parameter D;

[0011] S4: Obtain the evaluation result of the defect severity according to the value ranges of the four characteristic parameters.

[0012] Preferably, step S1 specifically includes the following steps:

[0013] S11: Construct a simulation circuit model to simulate the UHF signals generated along the insulation surface of the GIS equipment;

[0014] S12: Confirm the UHF signal phase and the UHF signal amplitude according to the current pulse phase of the simulation circuit model;

[0015] S13: Obtain the UHF spectrum according to the UHF signal phase and the UHF signal amplitude. The UHF spectrum includes five consecutive UHF spectrum data.

[0016] Preferably, in step S13, each piece of consecutive UHF spectrum data includes the UHF signals of the GIS surface defect discharge within 20 power frequency cycles.

[0017] Preferably, step S2 specifically includes the following steps:

[0018] S21: Calculate the work function value Φ according to the positive half-cycle UHF pulse mean value;

[0019] S22: Fit the work function value Φ and the number of discharge pulses within 20 power frequency cycles through a cubic function to obtain the relationship between the work function value Φ and the number of discharge pulses;

[0020] S23: Iteratively find the root of the function according to the measured actual number of discharge pulses to obtain the work function value Φ corresponding to the defect surface when the discharge occurs;

[0021] S24: Eliminate the influence of the work function value Φ on the positive half-cycle UHF pulse mean value according to the relationship between the positive half-cycle UHF pulse mean value and the work function value Φ;

[0022] S25: Obtain the change situation of the UHF pulse mean value caused by the change of the defect severity.

[0023] Preferably, in step S22, the relationship between the work function value Φ and the number of discharge pulses is specifically set as:

[0024] num = a3Φ 3 +a2Φ 2 +a1Φ + a0

[0025] where num represents the number of discharge pulses, a0, a1, a2, and a3 all represent fitting parameters, and a0 is set to -2.915×10 5 , a1 is set to 6.886×10 5 , a2 is set to -5.409×10 5 , a3 is set to 1.413×10 5 .

[0026] Preferably, in step S25, the mean value of the UHF pulses increases with the deepening of the defect severity.

[0027] Preferably, step S4 specifically includes the following steps:

[0028] S41: Encode parameters A, B, C, and D to construct a look-up table of the encoding and the severity;

[0029] S42: Distinguish the data with low severity according to the encodings of parameters A and B;

[0030] S43: If the data does not belong to the low severity, distinguish the medium severity and the high severity according to the encodings of parameters B, C, and D;

[0031] S44: Obtain the severity corresponding to the encoding by looking up the table.

[0032] Preferably, in step S3, parameter A is set to the ratio ampn of the pulse peak value to the sum of the pulse amplitudes, parameter B is set to the ratio aveΦ of the forward normalized pulse mean value to the minimum discharge pulse amplitude, parameter C is set to the standard deviation σamp of ampn, and parameter D is set to the standard deviation σave of aveΦ.

[0033] Therefore, by adopting the above method for evaluating the severity of the insulation surface defect of a GIS device, the present invention has the following beneficial effects:

[0034] (1) The present invention extracts several UHF signal characteristic parameters that can best reflect the discharge characteristics in different stages for formulating the evaluation criteria for the defect severity. Compared with the data-driven method, the characteristic quantities selected by the present invention can be integrated with the discharge mechanism, making the proposed method have both accuracy and a theoretical basis;

[0035] (2) Starting from the microscopic discharge mechanism, the present invention explores the relationship between the number of UHF pulses, the pulse mean value and the nature of the defect itself, and extracts four dimensionless physical quantities as the basis for severity evaluation, enhancing the universality of the method;

[0036] (3) The evaluation method proposed by the present invention has a good classification effect on the data in the initial stage of deterioration. For the development stage and the dangerous stage, it can cope with the randomness of discharge data, with an accuracy rate as high as 85%, and can accurately divide the deterioration stage of the surface discharge defect.

[0037] The method solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Description of the Drawings

[0038] Figure 1 It is a flowchart of an evaluation method for the severity of insulation surface discharge defects of a GIS device according to the present invention;

[0039] Figure 2 It is a simulation circuit model diagram of the UHF spectrum signal of the insulation surface discharge defect according to the present invention;

[0040] Figure 3 It is the corresponding relationship between the average value of the UHF pulse obtained by the circuit model and the work function Φ value according to the present invention;

[0041] Figure 4 It is a flowchart for evaluating the severity of the defect according to the present invention. Detailed Embodiments

[0042] The method solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0043] Unless otherwise defined, the method terms or scientific terms used in the present invention shall have the ordinary meaning as understood by those of ordinary skill in the field to which the present invention belongs.

[0044] The terms such as "including" or "comprising" used in the present invention mean that the elements before this word cover the elements listed after this word, and do not exclude the possibility of also covering other elements. The orientation or positional relationship indicated by the terms "inside", "outside", "above", "below", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. When the absolute position of the described object changes, the relative positional relationship may also change accordingly. In the present invention, unless otherwise clearly defined and limited, terms such as "attached" shall be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can be directly connected, or indirectly connected through an intermediate medium, and can be the internal communication of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0045] Embodiment

[0046] As Figure 1 shown, the present invention provides a method for evaluating the severity of insulation surface defects of GIS equipment, including the following steps:

[0047] S1: As Figure 2 shown, construct a simulation circuit model to obtain five consecutive UHF spectrum data;

[0048] Step S1 specifically includes the following steps:

[0049] S11: Construct a simulation circuit model to simulate the UHF signals generated along the insulation surface of GIS equipment. These UHF data are the basis for judging the severity of defects;

[0050] S12: Confirm the UHF signal phase and UHF signal amplitude according to the current pulse phase of the simulation circuit model;

[0051] S13: Obtain the UHF spectrum according to the UHF signal phase and UHF signal amplitude. The UHF spectrum includes five consecutive UHF spectrum data;

[0052] In step S13, each piece of consecutive UHF spectrum data includes the UHF signals of GIS surface defect discharges within 20 power frequency cycles.

[0053] S2: Calculate the forward normalized pulse mean. Under the power frequency AC voltage, the discharge dangerous state of the surface defect always appears under the positive voltage. Therefore, the forward pulse mean can be selected as an important physical quantity for judging the defect severity. However, in addition to the defect severity, the work function value Φ of the surface will also affect the pulse mean. Therefore, it is necessary to calculate the work function value Φ of the defect surface according to the number of UHF pulses, and then normalize the influence of the work function value Φ through the functional relationship between the surface work function value Φ and the pulse mean, so that the normalized pulse mean can be used as an important reference variable reflecting the defect severity;

[0054] Step S2 specifically includes the following steps:

[0055] S21: Calculate the work function value Φ according to the forward half-cycle UHF pulse mean;

[0056] S22: As Figure 3 shown, the influence of the work function value Φ of the defective surface on the number of multi-cycle discharge pulses hardly changes with the defect severity. As the work function value Φ increases, the number of pulses continuously decreases, and the slope of the decrease first changes from slow to fast and then gradually slows down;

[0057] Control the discharge delay time in the simulation circuit model, and fit the work function value Φ and the number of discharge pulses within 20 power frequency cycles through a cubic function to obtain the relationship between the work function value Φ and the number of discharge pulses;

[0058] In step S22, the relationship between the work function value Φ and the number of discharge pulses is specifically set as follows:

[0059] num = a3Φ 3 + a2Φ 2 + a1Φ + a0

[0060] where num represents the number of discharge pulses, and a0, a1, a2, and a3 all represent fitting parameters. a0 is set to -2.915×10 5 , a1 is set to 6.886×10 5 , a2 is set to -5.409×10 5 , a3 is set to 1.413×10 5 ;

[0061] S23: Perform iterative root finding on the function based on the measured actual number of discharge pulses to obtain the work function value Φ corresponding to the defect surface when discharge occurs;

[0062] S24: According to the relationship between the positive half-cycle UHF pulse mean value and the work function value Φ, eliminate the influence of the work function value Φ on the positive half-cycle UHF pulse mean value. The relationship between the positive half-cycle UHF pulse mean value and the work function value Φ is derived from the operation result of the surface discharge defect circuit model. According to this corresponding relationship, different UHF data collected can be normalized to the same work function value Φ, thereby eliminating the influence of the work function value Φ on the positive half-cycle UHF pulse mean value;

[0063] S25: Obtain the change in the UHF pulse mean value caused by the change in the defect severity;

[0064] In step S25, the UHF pulse mean value increases as the defect severity deepens. In the initial stage of surface defect deterioration, the number of UHF signal discharges is large and the mean value is small, and the change is relatively continuous; in the development stage, the signal peak value is small, the continuity of the characteristic value decreases, but the oscillation amplitude is not large; in the dangerous stage, the signal mean value is high, and there will be an outlier high peak value, and the change degree of the signal is relatively drastic.

[0065] S3: Calculate four characteristic parameters based on five consecutive UHF spectrogram data and the forward normalized pulse mean value to evaluate the defect severity. The four characteristic parameters include parameter A, parameter B, parameter C, and parameter D;

[0066] In step S3, since factors such as the type of detection instrument and the size of the defect may affect the signal characteristic values, it is necessary to further refine more general characteristic parameters based on the three characteristic values measured in the experiment as the basis for evaluating the defect severity;

[0067] In addition, due to the strong randomness of signal changes during discharge, the data fluctuations between signals are also an important indicator for judging the severity of defects. Therefore, when evaluating, the severity cannot be simply judged by a single piece of data;

[0068] Set parameter A as the ratio ampn of the pulse peak value to the sum of pulse amplitudes. Among them, the sum of pulse amplitudes is the sum of all pulse amplitudes in a set of 20-cycle ultra-high frequency data. This data is a dimensionless data, which can reflect the energy proportion of the pulse peak value in the overall discharge pattern and is closely related to both the ultra-high frequency signal peak value and the number of pulses;

[0069] Set parameter B as the ratio aveΦ of the forward normalized pulse mean value to the minimum discharge pulse amplitude. The normalized pulse mean value shows a continuous upward trend during the deterioration process. Considering that this numerical quantity cannot be directly used, this paper chooses to calculate the ratio of the normalized mean value to the minimum value of the discharge pulse. This ratio is equivalent to further standardizing the normalized mean value and can reflect the relative size of the normalized mean value in the discharge pattern. The minimum values of the discharge pulses in the five groups of discharge data are usually relatively consistent and close to the amplitude of the discharge pulse triggered near the discharge inception voltage. Therefore, this numerical value is a relative representation of the size of the normalized pulse mean value and can be used to judge the severity of defects;

[0070] Set parameter C as the standard deviation σamp of ampn, and set parameter D as the standard deviation σave of aveΦ. The basis is the variation law of ultra-high frequency data during the deterioration of the insulation surface defect until breakdown in the experiment. The standard deviation mainly reflects the volatility of the characteristic values between multiple discharge data. During the defect deterioration process, the volatility of the characteristic values gradually increases. Therefore, the standard deviation of the characteristic quantity can be used for supplementary judgment of the defect severity. When calculating, it is necessary to subtract the mean value from the five pieces of data respectively, that is, to zero the data mean value before comparing the standard deviation.

[0071] S4: Obtain the evaluation result of the defect severity according to the value ranges of the four characteristic parameters.

[0072] As Figure 4 shown, step S4 specifically includes the following steps:

[0073] S41: As shown in Table 1-2, encode parameter A, parameter B, parameter C, and parameter D to obtain the value ranges corresponding to the four parameters under different defect severities. Then, according to the experimental data, a comparison table of different characteristic parameter combinations and defect severities, as well as an evaluation process, are formulated;

[0074] Table 1: Characteristic parameter coding and parameter value ranges corresponding to different severities of insulation surface defects

[0075]

[0076]

[0077] Table 2: Comparison Table of Different Characteristic Parameter Combinations and Defect Severity

[0078]

[0079] S42: Differentiate the data with low severity according to the codes of parameter A and parameter B, calculate the stages corresponding to the numerical values of characteristic parameters A and B in the five pieces of data, and set the stage with a higher proportion as the characteristic parameter code of this group of five pieces of data. When the number of times of the proportion is the same, take the higher severity level as the result;

[0080] S43: If the data does not belong to the low severity, differentiate the medium severity and high severity according to the codes of parameter B, parameter C, and parameter D. Among them, the coding results of parameter C and parameter D are determined by the numerical size;

[0081] S44: Obtain the corresponding severity of the code through the method of looking up the table.

[0082] In the test of the test set, it is found that the method proposed in this embodiment has a good classification effect on the data in the initial stage of deterioration. The accuracy rate on the test set reaches 97%. For the development stage and the dangerous stage, the accuracy rate of the evaluation method reaches about 85%. Therefore, generally speaking, it can be considered that the method proposed in this embodiment can accurately divide the deterioration stage of the surface discharge defect.

[0083] Therefore, the present invention adopts the above-mentioned method for evaluating the severity of the surface discharge defect of a GIS device. Starting from the microscopic discharge mechanism, the relationship between the number of UHF pulses, the pulse mean value and the nature of the defect itself is explored, and four dimensionless physical quantities are extracted as the basis for severity evaluation, enhancing the universality of the method.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the method scheme of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the method scheme of the present invention, and these modifications or equivalent replacements cannot make the modified method scheme deviate from the spirit and scope of the method scheme of the present invention.

Claims

1. A method for evaluating the severity of insulation surface defects of GIS equipment, characterized in that It includes the following steps: S1: Construct a simulation circuit model to obtain five consecutive UHF spectrum data; S2: Calculate the forward normalized pulse mean value; S3: Calculate four characteristic parameters according to the five consecutive UHF spectrum data and the forward normalized pulse mean value. The four characteristic parameters include parameter A, parameter B, parameter C, and parameter D; S4: Obtain the defect severity assessment result according to the value ranges of the four characteristic parameters.

2. The evaluation method for the severity of insulation surface defects of a GIS device according to claim 1, characterized in that, Step S1 specifically includes the following steps: S11: Construct a simulation circuit model to simulate the UHF signal generated by the insulation along the surface of the GIS device; S12: Confirm the UHF signal phase and the UHF signal amplitude according to the current pulse phase of the simulation circuit model; S13: Obtain the UHF spectrum according to the UHF signal phase and the UHF signal amplitude. The UHF spectrum includes five consecutive UHF spectrum data.

3. The evaluation method for the severity of the insulation surface defect of a GIS device according to claim 2, characterized in that, In step S13, each piece of consecutive UHF spectrum data includes the UHF signal of the GIS surface defect discharge within 20 power frequency cycles.

4. The evaluation method for the severity of insulation surface defects of a GIS device according to claim 1, characterized in that Step S2 specifically includes the following steps: S21: Calculate the work function value Φ according to the positive half-cycle UHF pulse mean value; S22: Fit the work function value Φ and the number of discharge pulses within 20 power frequency cycles through a cubic function to obtain the relationship between the work function value Φ and the number of discharge pulses; S23: Iteratively find the root of the function according to the measured actual number of discharge pulses to obtain the work function value Φ corresponding to the defect surface when the discharge occurs; S24: Exclude the influence of the work function value Φ on the positive half-cycle UHF pulse mean value according to the relationship between the positive half-cycle UHF pulse mean value and the work function value Φ; S25: Obtain the change situation of the UHF pulse mean value caused by the change of the defect severity.

5. The evaluation method for the severity of insulation surface defects of a GIS device according to claim 4, characterized in that, In step S22, the relationship between the work function value Φ and the number of discharge pulses is specifically set as: num = a3Φ 3 + a2Φ 2 + a1Φ + a0 Among them, num represents the number of discharge pulses, and a0, a1, a2, and a3 all represent fitting parameters. a0 is set to -2.915×10 5 , a1 is set to 6.886×10 5 , a2 is set to -5.409×10 5 , a3 is set to 1.413×10 5 .

6. The evaluation method for the severity of the insulation surface defect of a GIS device according to claim 4, wherein In step S25, the UHF pulse mean value increases with the deepening of the defect severity.

7. The evaluation method for the severity of insulation surface defects of a GIS device according to claim 1, characterized in that Step S4 specifically includes the following steps: S41: Encode parameters A, B, C, and D, and construct a comparison table of encoding and severity; S42: Distinguish the data of low severity according to the encodings of parameters A and B; S43: If the data does not belong to low severity, distinguish medium severity and high severity according to the encodings of parameters B, C, and D; S44: Obtain the severity corresponding to the encoding by looking up the table.

8. The evaluation method for the severity of the insulation surface defect of a GIS device according to claim 1, wherein In step S3, parameter A is set as the ratio of the pulse peak value to the sum of the pulse amplitudes ampn, parameter B is set as the ratio of the forward normalized pulse mean value to the minimum discharge pulse amplitude aveΦ, parameter C is set as the standard deviation σamp of ampn, and parameter D is set as the standard deviation σave of aveΦ.

Images