Automatic Secondary Pulse Waveform Selection Method and System for Cable Fault Location

Abstract

The present invention relates to the technical field of cable fault detection, and proposes a method and system for automatically selecting secondary pulse waveforms for cable fault ranging, including: using the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data; for each group of secondary pulse waveform data, calculating the reverse index of the correlation coefficient, the root mean square error of each group of waveforms, and the wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators; dynamically adjusting the distribution weights of each characteristic indicator according to the proportion of the coefficient of variation of each characteristic indicator; based on the obtained distribution weights, performing weighted calculation on the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy to obtain a comprehensive score, and selecting the secondary pulse waveform data based on the comprehensive score. The multi-dimensional feature weighted scoring that integrates the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy solves the problems of single judgment dimension and inaccurate selection in the existing methods.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault detection, and specifically, to a method and system for automatically selecting secondary pulse waveforms for cable fault ranging. Background Art

[0002] The statements in this part merely provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] Power cables are important infrastructure for transmitting and distributing electric energy in urban power distribution systems and are widely used in power systems for the electrical connection between various electrical equipment. Compared with traditional overhead lines, the cable laying form has the advantages of small floor area, high safety, strong adaptability, and low maintenance requirements, and is especially suitable for power transmission in densely built urban areas or complex terrain environments. In practical applications, power cables are mainly laid underground through pipelines or direct burial. Although this laying method is convenient for urban planning and protecting the safety of the lines, it also makes the cables extremely vulnerable to external factors such as moisture, corrosion, and mechanical extrusion, resulting in various fault types such as insulation aging failure, joint failure, and mechanical damage. In order to quickly and accurately locate the cable fault location, reduce the power outage time and maintenance costs, a widely used technical means at present is the secondary pulse method; this method estimates the cable fault distance by comparing the discharge reflection waveforms before and after breakdown and analyzing the position of their bifurcation points. In actual tests, multiple pairs of data of breakdown waveforms and non-breakdown waveforms are obtained, and the accurate fault location is judged by comparing the differences in the bifurcation points of these waveforms.

[0004] Currently, in fault ranging, the selection of secondary pulse waveforms requires manually selecting the optimal data pair from multiple sets of waveform data pairs for marking the bifurcation points. This process highly depends on the subjective judgment and experience level of the operator. Manually selecting data pairs not only consumes time, but also, due to the small differences between waveforms and the fuzzy judgment basis, it is easy to lead to inaccurate selection, thus affecting the final ranging accuracy.

[0005] In the existing secondary pulse method for power cable fault ranging, although there is already a certain waveform selection mechanism, some current automated waveform selection methods only judge the quality of waveforms based on a single characteristic index such as correlation. This one-dimensional evaluation model has problems of weak expression ability and poor adaptability, and cannot comprehensively describe the differences between breakdown waveforms and non-breakdown waveforms, easily leading to misselection, misjudgment, or failure for complex waveforms, ultimately affecting the accuracy and stability of the cable fault ranging results. Summary of the Invention

[0006] In order to solve the above problems, the present invention proposes a method and system for automatically selecting secondary pulse waveforms for cable fault ranging, which integrates a multi-dimensional feature weighted scoring mechanism of the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy, and automatically selects the optimal waveform data pair in the secondary pulse, solving the problems of single judgment dimension and inaccurate selection in the existing methods.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] The first aspect of the present invention provides a method for automatically selecting secondary pulse waveforms for cable fault ranging, including the following steps:

[0009] Use the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data;

[0010] For each group of secondary pulse waveform data, calculate the reverse index of the correlation coefficient, the root mean square error of each group of waveforms, and the wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators;

[0011] According to the proportion of the coefficient of variation of each characteristic indicator, dynamically adjust the distribution weights of each characteristic indicator;

[0012] Based on the obtained distribution weights, weight the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy to obtain a comprehensive score, and select the secondary pulse waveform data based on the comprehensive score.

[0013] The second aspect of the present invention is a system for automatically selecting secondary pulse waveforms for cable fault ranging, including:

[0014] A data acquisition module configured to use the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data;

[0015] A characteristic indicator calculation module configured to calculate the reverse index of the correlation coefficient, the root mean square error of each group of waveforms, and the wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators for each group of secondary pulse waveform data;

[0016] A weight dynamic update module configured to dynamically adjust the distribution weights of each characteristic indicator according to the proportion of the coefficient of variation of each characteristic indicator;

[0017] A scoring and selection module configured to weight the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy based on the obtained distribution weights to obtain a comprehensive score, and select the secondary pulse waveform data based on the comprehensive score.

[0018] The third aspect of the present invention provides a system for automatically selecting secondary pulse waveforms for cable fault ranging, including:

[0019] The secondary pulse method testing device is used to collect multiple groups of secondary pulse reflection waveform data;

[0020] A processor is configured to execute the steps in the above-mentioned automatic secondary pulse waveform selection method for cable fault location.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0022] The present invention locates faults by comparing and analyzing the breakdown waveforms and non-breakdown waveforms in the secondary pulse method measurement. It combines the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy of the breakdown discharge waveforms in each group of waveforms as characteristic indicators. By calculating the coefficient of variation of each indicator in the sample population and dynamically allocating weights according to the proportion of the coefficient of variation, the final comprehensive score can adaptively reflect the importance differences of each indicator, realizing waveform optimization with more representativeness and discriminative power.

[0023] The waveform selection method avoids the subjectivity and low efficiency problems in manual waveform selection, significantly improving the accuracy and reliability of waveform selection. Compared with the method that only makes judgments based on a single indicator such as correlation, this multi-dimensional comprehensive evaluation method has stronger discriminative ability in complex waveform scenarios, can effectively reduce the probability of misselection and misjudgment, thereby improving the accuracy and stability of cable fault location. At the same time, using the coefficient of variation as a measurement standard, it automatically determines the weights of each feature in the total score, realizing the dynamic allocation and quantitative adjustment of feature weights. The weight adjustment mechanism driven by the coefficient of variation realizes the ability to automatically optimize the importance of features for different test samples, enhancing the adaptability and intelligence level of the system, and having good prospects for engineering promotion and application.

[0024] The advantages of the present invention and the advantages of additional aspects will be described in detail in the following specific embodiments. Description of the Drawings

[0025] The specification drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute a limitation to the present invention.

[0026] Figure 1 It is a flowchart of the waveform automatic selection method in Embodiment 1 of the present invention;

[0027] Figure 2(a) is the first group of secondary pulse current waveform diagram in the simulation test example of Embodiment 1 of the present invention;

[0028] Figure 2(b) is the second group of secondary pulse current waveform diagram in the simulation test example of Embodiment 1 of the present invention;

[0029] Figure 2(c) is the third group of secondary pulse current waveform diagram in the simulation test example of Embodiment 1 of the present invention;

[0030] Figure 2(d) is the waveform diagram of the fourth group of secondary pulse currents in the simulation test example of Embodiment 1 of the present invention;

[0031] Figure 2(e) is the waveform diagram of the fifth group of secondary pulse currents in the simulation test example of Embodiment 1 of the present invention;

[0032] Figure 2(f) is the waveform diagram of the sixth group of secondary pulse currents in the simulation test example of Embodiment 1 of the present invention;

[0033] Figure 2(g) is the waveform diagram of the seventh group of secondary pulse currents in the simulation test example of Embodiment 1 of the present invention;

[0034] Figure 2(h) is the waveform diagram of the eighth group of secondary pulse currents in the simulation test example of Embodiment 1 of the present invention; Detailed implementation manners

[0035] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0036] It should be noted that the following detailed descriptions are all exemplary and are intended to provide further descriptions of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0037] It should be noted that the terms used herein are only for describing specific implementation manners and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof. It should be noted that, without conflict, the various embodiments and features in the present invention can be combined with each other. The embodiments will be described in detail below in conjunction with the accompanying drawings.

[0038] Embodiment 1

[0039] In the technical solutions disclosed in one or more embodiments, as Figure 1 shown, an automatic selection method for secondary pulse waveforms for cable fault location includes the following steps:

[0040] Step 1: Use the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data;

[0041] Step 2: For each group of secondary pulse waveform data, calculate the correlation coefficient reverse index, root mean square error, and wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators;

[0042] Step 3: Dynamically adjust the distribution weights of each characteristic index according to the proportion of the coefficient of variation of each characteristic index.

[0043] Step 4: Based on the obtained distribution weights, perform weighting on the reverse correlation coefficient, root mean square error, and wavelet energy entropy to obtain a comprehensive score, and select the secondary pulse waveform data based on the comprehensive score.

[0044] In this embodiment, the fault is located by comparing and analyzing the breakdown waveform and non-breakdown waveform in the secondary pulse method measurement. The reverse correlation coefficient, root mean square error, and wavelet energy entropy of the breakdown discharge waveform in each group of waveforms are fused as characteristic indexes. By calculating the coefficient of variation of each index in the sample population and dynamically allocating weights according to the proportion of the coefficient of variation, the final comprehensive score can adaptively reflect the importance differences of each index, realizing waveform optimization with more representativeness and discrimination ability. Fault location based on the secondary pulse waveform data selected based on the comprehensive score can improve the accuracy of fault location.

[0045] The waveform selection method of this embodiment avoids the subjectivity and low efficiency problems in manual waveform selection, and significantly improves the accuracy and reliability of waveform selection. Compared with the method that only judges based on a single index such as correlation, this multi-dimensional comprehensive evaluation method has stronger discrimination ability in complex waveform scenarios, can effectively reduce the probability of misselection and misjudgment, and thus improve the accuracy and stability of cable fault ranging. At the same time, using the coefficient of variation as a measurement standard, the weights of each characteristic in the total score are automatically determined, realizing the dynamic allocation and quantitative adjustment of characteristic weights. The weight adjustment mechanism driven by the coefficient of variation realizes the ability to automatically optimize the importance of characteristics for different test samples, enhances the adaptability and intelligence level of the system, and has good prospects for engineering promotion and application.

[0046] First, explain the basic principle of the secondary pulse method for measuring cable faults: Apply a high-voltage pulse to the cable with a high-resistance or flashover fault through a high-voltage generator to cause arcing discharge at the fault point. Since the arc resistance is very small, the original high-resistance or flashover fault becomes a low-resistance short-circuit fault during the arcing period. At this time, inject a low-voltage pulse signal into the faulty cable through a pulse generating device, and record the low-voltage pulse reflection waveform at this time (referred to as the live arc waveform), then the low-resistance reflection pulse of the fault point can be clearly observed; after the fault arc extinguishes, inject a low-voltage pulse (secondary pulse) into the faulty cable again, and record the low-voltage pulse reflection waveform at this time (referred to as the non-arc waveform). At this time, because the fault resistance returns to high resistance, the low-voltage pulse signal has no reflection or very little reflection at the fault point. Compare the live arc waveform and the non-arc waveform to locate the fault.

[0047] In Step 1, multiple groups of secondary pulse reflection waveform data in power cable fault location are collected. Optionally, for each secondary pulse discharge test, 8 groups of waveform data can be collected. Each group of secondary pulse reflection waveform data includes a breakdown discharge reflection waveform and a non-breakdown reflection waveform. As shown in Figures 2(a) to 2(h), they are the first group to the eighth group of secondary pulse reflection waveform data collected. The secondary pulse reflection waveform can be generally divided into two stages: the front stage and the rear stage. Among them, the coincidence degree of the front-stage waveform is relatively high and the difference is relatively small, corresponding to the reflection waveform before the low-voltage pulse signal encounters the fault point; the coincidence degree of the rear-stage waveform is relatively low and the difference is relatively large, corresponding to the reflection waveform after the low-voltage pulse signal encounters the fault point; the bifurcation point that divides the waveform into the front stage and the rear stage is the fault point to be calibrated.

[0048] Theoretically, the earliest waveform has the highest arc discharge energy and clear waveform characteristics, and the fault point should be easier to identify. However, in actual tests, due to factors such as signal interference, coupling characteristic changes, and unstable dielectric breakdown paths, the early waveforms (especially the first group) may not necessarily provide the optimal judgment basis. Therefore, selecting waveforms for fault location based on experience or single-factor judgment is not accurate, resulting in inaccurate fault location.

[0049] In some embodiments, in Step 2, the calculation of the characteristic indexes of the breakdown discharge reflection waveform and the non-breakdown reflection waveform is as follows:

[0050] Optionally, the Pearson correlation coefficient formula is used to calculate the correlation coefficient of each group of secondary pulse reflection waveforms, which is used to quantify the linear correlation between the breakdown discharge reflection waveform and the non-breakdown reflection waveform within the same group , and the calculation formula is:

[0051] (1);

[0052] Where represents the amplitude of the i-th sampling point of the breakdown discharge reflection waveform and represents the amplitude of the i-th sampling point of the non-breakdown reflection waveform , and are the average values of the amplitudes of the two waveforms respectively, and n is the total number of sampling points.

[0053] In this embodiment, the Pearson correlation coefficient is introduced as a waveform similarity evaluation tool, providing a means of simple calculation and intuitive results, which helps to quickly identify waveform pairs with significant differences from non-breakdown waveforms in multiple groups of breakdown waveform data. Compared with empirical judgment or morphological comparison, this method has the advantages of strong objectivity and good adaptability, and is particularly suitable for the automatic processing scenario of large-scale waveform samples. In addition, as a linear matching tool, the correlation measure complements non-linear indicators such as mean square error and wavelet energy entropy, jointly constructing a complete multi-dimensional waveform quality evaluation system.

[0054] In the above embodiment, the Pearson correlation coefficient can also be replaced by other correlation measurement methods, such as the Spearman rank correlation coefficient or the Kendall rank correlation coefficient.

[0055] In Formula 2, for the breakdown discharge reflection waveform and the non-breakdown reflection waveform, the root mean square error is calculated according to the amplitudes at the same sampling points of the two waveforms. Specifically, after aligning the two waveforms point by point on the time axis, the square of the amplitude difference at each sampling point is calculated in turn, and the sum of all square errors is averaged and then square-rooted to obtain the root mean square error value of the waveform pair. , the formula is as follows:

[0056] (2);

[0057] Where, represents the amplitude sum of the i-th sampling point of the breakdown discharge reflection waveform and represents the amplitude of the i-th sampling point of the non-breakdown reflection waveform, and n is the total number of sampling points.

[0058] In a further technical solution, when calculating the root mean square error, before calculating the amplitude difference, preprocessing is performed using time window alignment and interpolation methods, which can ensure the accuracy of the amplitude difference calculation when there is a small offset on the time axis during the signal acquisition process.

[0059] Compared with the correlation coefficient that only considers the waveform trend, RMSE focuses more on the absolute deviation of the signal intensity, and can effectively make up for the lack of judgment of waveform pairs with similar trends but significantly different amplitudes. By introducing this index, it helps to identify waveforms with more prominent discharge characteristics from the amplitude perspective, improve the integrity of the waveform evaluation dimension, and enhance the accuracy and robustness of the fault location model.

[0060] In an alternative technical solution, the root mean square error in the characteristic index can be replaced by the mean absolute error (MAE) to weaken the influence of extreme deviations on the results and is suitable for high-noise waveform environments.

[0061] Optionally, in each group of waveform data, the breakdown discharge reflection waveform Determination method of wavelet energy entropy, comprising the following steps:

[0062] Step 21: Perform wavelet decomposition on the breakdown discharge reflection waveform to obtain the approximation coefficients and detail coefficients of each layer;

[0063] Specifically, perform wavelet decomposition on the breakdown discharge reflection waveform for N layers to obtain the approximation coefficients of each layer and detail coefficients . Then the number of subbands is (N + 1), including 1 approximation coefficient subband and N detail coefficient subbands. Among them, the approximation coefficients correspond to the low-frequency components of the signal, and the detail coefficients correspond to the high-frequency components at different scales (i.e., resolutions).

[0064] Among them, a subband refers to dividing the original signal into multiple frequency bands (sub-frequency bands) according to the frequency content, and each frequency band corresponds to a signal component;

[0065] Step 22: Calculate the energy values of the approximation coefficient subband and the detail coefficient subband for the approximation coefficients and detail coefficients respectively;

[0066] Optionally, the energy of the approximation coefficients The calculation formula is:

[0067] (3);

[0068] The energy of the detail coefficients of the k-th layer The calculation formula is:

[0069] (4);

[0070] Among them, , are the th approximation coefficient and detail coefficient respectively; and are the lengths of the subband coefficients respectively.

[0071] Step 23: Add up the obtained energy values of each subband to obtain the total energy of the approximation coefficients and detail coefficients , and the calculation formula is:

[0072] (5);

[0073] Step 24: Calculate the proportion of the energy of each subband in the total energy to obtain the energy distribution of each subband. The calculation formula is as follows:

[0074] (6);

[0075] Step S25: Calculate the wavelet energy entropy based on the subband energy distribution , and its calculation formula is:

[0076] (7);

[0077] Further, after step 2, it also includes normalizing the calculated characteristic indexes, normalizing the correlation coefficient, root mean square error, and wavelet energy entropy, and using the maximum-minimum normalization method to map each eigenvalue to the interval [0, 1]. The normalized correlation coefficient is , then is the reverse index of the correlation coefficient; the normalized root mean square error is ; the normalized wavelet energy entropy is .

[0078] The method in step 2 uses wavelet decomposition to perform multi-scale analysis on non-stationary signals, and can accurately extract the detailed structure and energy distribution characteristics in the breakdown discharge waveform. Through the information theory index of energy entropy, it can effectively measure the local complexity of the waveform, help identify abnormal waveforms with strong local mutations, and improve the sensitivity and accuracy of waveform screening. Compared with the global spectrum information provided by the traditional Fourier transform, the wavelet energy entropy has the advantages of time and frequency localization, and is more suitable for signal analysis of breakdown events with strong transient characteristics.

[0079] In step 3, the method of dynamically adjusting the allocation weights of each characteristic index according to the proportion of the coefficient of variation of each characteristic index includes the following steps:

[0080] Step 31: Calculate the mean and standard deviation of each characteristic index;

[0081] Characteristic index The mean The calculation formula is:

[0082] (8);

[0083] Characteristic index The standard deviation The calculation formula is:

[0084] (9);

[0085] Among them, represents groups of secondary pulse reflection waveform data. In this embodiment, M = 8; represents the cth characteristic index corresponding to the mth group of waveform data. In this embodiment, there are 3 characteristic indexes, namely the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy.

[0086] Step 32: Calculate the coefficient of variation for each characteristic index based on the calculated mean and standard deviation, including the coefficient of variation of the reverse index of the correlation coefficient, the coefficient of variation of the root mean square error, and the coefficient of variation of the wavelet energy entropy. The calculation formulas are as follows:

[0087] (10);

[0088] (11);

[0089] (12);

[0090] Wherein, and represent the standard deviation and mean of the reverse index of the correlation coefficient ; and represent the standard deviation and mean of the root mean square error ; and represent the standard deviation and mean of the wavelet energy entropy ; represents the coefficient of variation of the reverse index of the correlation coefficient; represents the coefficient of variation of the root mean square error, represents the coefficient of variation of the wavelet energy entropy;

[0091] For an alternative technical solution, in calculating the coefficient of variation, in addition to using the ratio of the standard deviation to the mean, when the sample data does not follow a normal distribution or there are extreme values, the median absolute deviation (MAD) can also be considered to replace the standard deviation to enhance robustness, and the ratio of the mean to the median absolute deviation is used as the coefficient of variation of the characteristic index;

[0092] Step 33: Calculate the proportion of the coefficient of variation for each characteristic index respectively, and use the proportion corresponding to the characteristic index as the allocation weight of the characteristic index;

[0093] Calculating the proportion of the coefficient of variation for each characteristic index includes the proportion of the coefficient of variation of the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy in the total coefficient of variation, and using the proportion corresponding to the characteristic index as the allocation weight of the characteristic index. The calculation formulas are as follows:

[0094] (13);

[0095] (14);

[0096] (15);

[0097] Wherein, Indicates the proportion of the coefficient of variation representing the reverse index of the correlation coefficient in the total coefficient of variation; Indicates the proportion of the coefficient of variation of the root mean square error in the total coefficient of variation; Indicates the proportion of the coefficient of variation of the wavelet energy entropy in the total coefficient of variation. These three proportion values are the weights to be assigned to the corresponding characteristic indicators.

[0098] In the above embodiments, by dynamically calculating the proportion of the coefficient of variation to weight each characteristic indicator, the scoring mechanism has the self-adjusting ability driven by data. In this way, the importance of each characteristic can be adaptively adjusted according to the actual test data, ensuring that the scoring result can more comprehensively and objectively reflect the waveform difference, improving the discrimination of waveform screening and the stability of the overall ranging algorithm.

[0099] In step 4, the calculation formula for the comprehensive score Score is:

[0100] (16);

[0101] For the method of selecting secondary pulse waveform data based on the comprehensive score, optionally, based on the obtained comprehensive score Score, the comprehensive scores calculated for multiple groups of secondary pulse discharge waveform data are arranged, and the group with a comprehensive score greater than the set score threshold or the group with the highest score of the secondary pulse reflection waveform data is selected as the selection result for power cable fault ranging.

[0102] Among them, arranging the comprehensive scores calculated for multiple groups of secondary pulse discharge waveform data can be in ascending order or descending order;

[0103] To illustrate the above selection process and the improvement of the fault judgment accuracy by the selected waveform data, a comparative experiment was conducted;

[0104] (1) Explanation of the selection process;

[0105] First, obtain 8 groups of secondary pulse waveform data obtained during the power cable fault test, as shown in Figures 2(a) to 2(h); calculate the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy of the breakdown discharge waveform in each group of waveforms respectively, and normalize the obtained characteristic indicators to obtain the results shown in Table 1;

[0106] Table 1 Calculation results of the characteristic indicators of the secondary pulse waveform;

[0107]

[0108] The calculation results of the coefficient of variation and weight of the normalized reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy of the secondary pulse waveforms shown in Figures 2(a) to 2(h) are shown in Table 2:

[0109] Table 2 shows the calculation results of the coefficient of variation and weight of three characteristic indexes;

[0110]

[0111] Combined with Table 3, the calculation results of the comprehensive scores of the 8 groups of secondary pulse waveforms shown in Figs. 2(a) to 2(h) are given. The third group is the selected optimal secondary pulse waveform.

[0112] Table 3 is the comprehensive score of 8 groups of secondary pulse waveforms;

[0113]

[0114] (2) Conduct a comparative test;

[0115] Combined with the data in Table 1, if only the reverse index of the correlation coefficient is used as the basis for waveform selection, the second group of waveforms corresponding to Fig. 2(b) will be screened out as the optimal secondary pulse waveform; if only the root mean square error index is used as the basis for waveform selection, the fifth group of waveforms corresponding to Fig. 2(e) will be screened out as the optimal secondary pulse waveform; if only the wavelet energy entropy is used as the basis for waveform selection, the sixth group of waveforms corresponding to Fig. 2(f) will be screened out as the optimal secondary pulse waveform. Comparing the screening results of these three groups of waveforms with the screening results of the method proposed in this embodiment, it is found that the screening results based on the comprehensive scoring method proposed in this embodiment are significantly better than those based on a single characteristic index. For the waveform signal screening results of single index and comprehensive index for Figs. 2(a) to 2(h), the fault location is calculated respectively, and the accuracy of the location results is shown in Table 4. The signal fault point selected by the comprehensive index of this embodiment has the most accurate location;

[0116] Table 4 is the comparison result of fault location of waveforms screened by different indexes;

[0117]

[0118] To further illustrate the beneficial effects of this embodiment, the comprehensive scoring method and the single feature index method of this embodiment are respectively used to screen the optimal secondary pulse waveform, which are respectively used for bifurcation point marking and fault location. According to experience, the relative ranging error within ±2% is used as the criterion for accurate ranging. If the relative ranging error is within ±2%, it is judged that the fault location is accurate. In this embodiment, the effects are compared by the number of accurate groups. In 20 comparison tests, the number of accurate ranging groups of the optimal secondary pulse waveform screened by the comprehensive scoring of this embodiment reaches 18 groups; while the number of accurate ranging groups of the optimal secondary pulse waveform screened by the single correlation coefficient reverse index is 14 groups, the number of accurate ranging groups of the optimal secondary pulse waveform screened by the single root mean square error index is 12 groups, and the number of accurate ranging groups of the optimal secondary pulse waveform screened by the single wavelet energy entropy is also 12 groups. Obviously, the optimal secondary pulse waveform screened by the comprehensive scoring of this embodiment has higher ranging accuracy.

[0119] Embodiment 2

[0120] Based on Embodiment 1, this embodiment provides an automatic selection system for secondary pulse waveforms for cable fault location, including:

[0121] A data acquisition module, configured to perform tests using the secondary pulse method and acquire multiple groups of secondary pulse reflection waveform data;

[0122] A feature index calculation module, configured to calculate the correlation coefficient reverse index, the root mean square error of each group of waveforms, and the wavelet energy entropy of the breakdown discharge waveform in each group of waveforms for each group of secondary pulse waveform data as feature indexes;

[0123] A weight dynamic update module, configured to dynamically adjust the distribution weights of each feature index according to the proportion of the coefficient of variation of each feature index;

[0124] A scoring selection module, configured to obtain a comprehensive score by weighting the correlation coefficient reverse index, the root mean square error, and the wavelet energy entropy based on the obtained distribution weights, and select secondary pulse waveform data based on the comprehensive score.

[0125] Further, the method for dynamically adjusting the distribution weights of each feature index according to the proportion of the coefficient of variation of each feature index includes the following steps:

[0126] Calculate the mean and standard deviation of each feature index;

[0127] Calculate the coefficient of variation of each feature index according to the calculated mean and standard deviation;

[0128] Calculate the proportion of the coefficient of variation of each feature index respectively, and use the proportion corresponding to the feature index as the distribution weight of the feature index.

[0129] It should be noted here that each module in this embodiment corresponds one by one to each step in Embodiment 1, and the specific implementation process is the same, so it will not be repeated here.

[0130] Embodiment 3

[0131] Based on Embodiment 1, this embodiment provides a secondary pulse waveform automatic selection system for cable fault ranging, including:

[0132] A secondary pulse method testing device for collecting multiple groups of secondary pulse reflection waveform data;

[0133] A processor configured to execute the steps in the secondary pulse waveform automatic selection method for cable fault ranging described in Embodiment 1.

[0134] Among them, the secondary pulse method testing device includes a high-voltage generator, a pulse generating device, and a receiving device;

[0135] The high-voltage generator is used to apply a high-voltage pulse to a cable with a high-resistance or flashover fault, causing arc discharge at the fault point;

[0136] The pulse generating device is used to inject a low-voltage pulse signal into the faulty cable;

[0137] The receiving device is used to collect and record the low-voltage pulse reflection waveform, that is, the secondary pulse reflection waveform data.

[0138] The process of measuring cable faults by the secondary pulse method is as follows: A high-voltage pulse is applied to a cable with a high-resistance or flashover fault through a high-voltage generator, causing arc discharge at the fault point. Since the arc resistance is very small, the original high-resistance or flashover fault becomes a low-resistance short-circuit fault during the arcing period. At this time, a low-voltage pulse signal is injected into the faulty cable through the pulse generating device, and the low-voltage pulse reflection waveform at this time (referred to as the live arc waveform) is recorded, and the low-resistance reflection pulse of the fault point can be clearly observed; after the fault arc goes out, another low-voltage pulse (secondary pulse) is injected into the faulty cable, and the low-voltage pulse reflection waveform at this time (referred to as the non-arc waveform) is recorded. At this time, since the fault resistance returns to high resistance, the low-voltage pulse signal has no reflection or very little reflection at the fault point. By comparing the live arc waveform and the non-arc waveform, the fault is located.

[0139] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0140] Although the specific implementation manners of the present invention have been described above in conjunction with the accompanying drawings, it is not a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications or deformations that can be made without creative efforts on the basis of the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. An automatic selection method for secondary pulse waveforms for cable fault location, characterized in that It includes the following steps: Use the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data; For each group of secondary pulse waveform data, calculate the correlation coefficient reverse index, root mean square error of each group of waveforms, and wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators; Dynamically adjust the distribution weights of each characteristic indicator according to the proportion of the coefficient of variation of each characteristic indicator; Based on the obtained distribution weights, weight the correlation coefficient reverse index, root mean square error, and wavelet energy entropy to obtain a comprehensive score, and select the secondary pulse waveform data based on the comprehensive score; The method for determining the wavelet energy entropy of the breakdown discharge reflection waveform includes the following steps: Perform wavelet decomposition on the breakdown discharge reflection waveform to obtain the approximate coefficients and detail coefficients of each layer after wavelet decomposition; For the approximate coefficients and detail coefficients, calculate the energy values of the approximate coefficient subband and the detail coefficient subband respectively; Add the obtained energy values of each subband to obtain the total energy of the approximate coefficients and detail coefficients; Calculate the proportion of the energy of each subband in the total energy to obtain the energy distribution of each subband; Calculate the wavelet energy entropy based on the subband energy distribution; Each group of secondary pulse reflection waveform data includes a breakdown discharge reflection waveform and an unbroken reflection waveform. Calculate the root mean square error based on the amplitudes at the same sampling points of the two waveforms. Specifically, after aligning the two waveforms point-to-point on the time axis, calculate the square of the amplitude difference at each sampling point in turn, sum all the square errors, take the average, and then take the square root to obtain the root mean square error value of the waveform pair; The method for dynamically adjusting the distribution weights of each characteristic indicator according to the proportion of the coefficient of variation of each characteristic indicator includes the following steps: Calculate the mean and standard deviation of each characteristic indicator; Based on the calculated mean and standard deviation, obtain the coefficient of variation of each characteristic indicator, including the coefficient of variation of the correlation coefficient reverse index, the coefficient of variation of the root mean square error, and the coefficient of variation of the wavelet energy entropy. The calculation formulas are respectively: （10）； （11）； （12）； Among them, and represent the standard deviation and mean of the reverse index of the correlation coefficient; of and represent the root mean square error of the standard deviation and mean; and represent the wavelet energy entropy of the standard deviation and mean; represents the coefficient of variation of the reverse index of the correlation coefficient; represents the coefficient of variation of the root mean square error, represents the coefficient of variation of the wavelet energy entropy; Calculate the proportion of the coefficient of variation of each characteristic indicator respectively, and use the proportion corresponding to the characteristic indicator as the distribution weight of the characteristic indicator.

2. The automatic secondary pulse waveform selection method for cable fault ranging according to claim 1, wherein: Use the Pearson correlation coefficient formula to calculate the correlation coefficient of each group of secondary pulse reflection waveforms.

3. The automatic secondary pulse waveform selection method for cable fault location according to claim 1, characterized in that: When calculating the root mean square error, perform preprocessing based on time window alignment and interpolation method before calculating the amplitude difference.

4. The automatic secondary pulse waveform selection method for cable fault ranging according to claim 1, characterized in that: Based on the obtained comprehensive score, rank the comprehensive scores calculated from multiple groups of secondary pulse discharge waveform data, and select the group with a comprehensive score greater than the set score threshold or the group with the highest score of the secondary pulse reflection waveform data as the selection result.

5. An automatic secondary pulse waveform selection system for cable fault location, which is based on the automatic secondary pulse waveform selection method for cable fault location described in claim 1, is characterized in that It includes: A data acquisition module configured to use the secondary pulse method for testing to obtain multiple groups of secondary pulse reflection waveform data; A characteristic indicator calculation module configured to calculate the correlation coefficient reverse index, root mean square error of each group of waveforms, and wavelet energy entropy of the breakdown discharge waveform in each group of waveforms as characteristic indicators for each group of secondary pulse waveform data; A weight dynamic update module configured to dynamically adjust the distribution weights of each characteristic indicator according to the proportion of the coefficient of variation of each characteristic indicator; A scoring selection module, configured to obtain a comprehensive score by weighting the reverse index of the correlation coefficient, the root mean square error, and the wavelet energy entropy based on the obtained allocation weights, and select the secondary pulse waveform data based on the comprehensive score.

6. A secondary pulse waveform automatic selection system for cable fault location, characterized in that It includes: A secondary pulse method testing device for collecting multiple groups of secondary pulse reflection waveform data; A processor, configured to execute the steps in the method for automatically selecting secondary pulse waveforms for cable fault location according to any one of claims 1-4.

Images