Double-end traveling wave positioning optimization method, device and equipment and readable storage medium

Abstract

The invention discloses a double-end traveling wave positioning optimization method, device and equipment and a readable storage medium. The positioning accuracy is improved from the three aspects of signal processing, fault detection and positioning calculation. Multi-level wavelet packet decomposition is carried out on an original traveling wave signal, a sub-band adaptive soft threshold is generated according to noise statistical characteristics, a traveling wave head signal with a high signal-to-noise ratio is reconstructed through filtering and inverse transformation, and the problem that signal integrity and noise suppression cannot be considered at the same time in a traditional method is solved. And time-frequency transformation and instantaneous energy envelope analysis are carried out on the reconstructed signal, after a correlation matrix and a sequence are generated, a lightweight time sequence model is input, and a fault triggering moment is output in combination with a constraint condition, so that the detection sensitivity of the weak high-resistance fault signal in a low signal-to-noise ratio environment is effectively improved, and the omission ratio is reduced. The coefficient compensation time difference is calculated according to the original time difference in combination with the wave amplitude value, and the positioning parameters are continuously corrected by dynamically updating the parameters and the coefficients, so that the limitation of the existing algorithm is overcome, and the positioning accuracy under a complex line is improved.

Description

Technical Field

[0001] The present application relates to the technical field of distribution network fault detection, and more specifically, to a dual-end traveling wave positioning optimization method, apparatus, device, and readable storage medium. Background Art

[0002] In modern power distribution systems, fault location technology is crucial for ensuring the stability and reliability of power supply. Traveling wave location, with its fast response and high positioning accuracy, has gained widespread application in fault detection and isolation on single overhead lines or pure cable lines. This method captures the voltage and current surge signals that rapidly propagate along the line at the moment of the fault and calculates the arrival time difference of the traveling waves at two measuring points, thereby accurately determining the fault location.

[0003] However, the traditional two-end traveling wave positioning algorithm has significant defects: first, it is extremely sensitive to signal noise. The commonly used bandpass filtering or simple threshold triggering method cannot effectively suppress noise while ensuring signal integrity. In a low signal-to-noise ratio environment, the traveling wave characteristics are easily masked by noise, seriously affecting the positioning performance; second, in the face of high-resistance grounding faults, the traveling wave generated by the fault has a small amplitude and low energy, which is often lower than the system noise level. The detection method based on amplitude threshold or single frequency domain characteristics is very easy to miss the detection; third, in the overhead and cable mixed line scenario, the traveling wave propagation speed and attenuation characteristics of the overhead section and the cable section are very different. The traveling wave is reflected and transmitted multiple times at the junction of the medium, resulting in complex time domain and frequency domain characteristics. The time difference positioning algorithm based on a single propagation speed assumption or a fixed model is difficult to be compatible with various working conditions and cannot effectively control the positioning error.

[0004] Therefore, an optimization method is urgently needed to solve the positioning error problem caused by multiple reflections and weak superposition of high-impedance signals in hybrid lines. Summary of the Invention

[0005] The present application provides a two-end traveling wave positioning optimization method, apparatus, equipment and readable storage medium, which reconstructs high signal-to-noise ratio signals through multi-level wavelet packet decomposition and adaptive threshold filtering, improves fault detection sensitivity by fusing time-frequency and energy features, dynamically compensates for time differences and corrects positioning parameters, and effectively solves the limitations of traditional methods in balancing noise suppression and signal integrity, weak fault detection and complex line positioning accuracy.

[0006] A dual-terminal traveling wave positioning optimization method, comprising:

[0007] The original traveling wave signal collected by the two terminals is decomposed by multi-level wavelet packets, and a sub-band adaptive soft threshold is generated according to the statistical characteristics of the noise. The two-terminal traveling wave head signal with high signal-to-noise ratio is reconstructed through soft threshold filtering and inverse wavelet packet transform.

[0008] Performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence respectively;

[0009] Input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering moment of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point;

[0010] Determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation;

[0011] According to the time difference compensation results and the actual fault location feedback, the recursive least squares method is used to dynamically update the wave velocity parameters and interface correction coefficients.

[0012] Optionally, performing multi-level wavelet packet decomposition on the original traveling wave signal collected from both ends, generating a sub-band adaptive soft threshold according to noise statistical characteristics, and reconstructing a two-end traveling wave head signal with a high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform, includes:

[0013] Perform multi-level wavelet packet decomposition on the original traveling wave signal collected by both ends to obtain high-frequency sub-bands and low-frequency sub-bands;

[0014] The median absolute deviation method is used to estimate the noise standard deviation on the first-level high-frequency sub-band;

[0015] Calculating the sub-band adaptive soft threshold of each sub-band level based on the noise standard deviation and combining the number of sub-band samples and the empirical adjustment coefficient;

[0016] Soft threshold filtering is applied to the sub-bands at each level according to the sub-band adaptive soft threshold, and a double-ended traveling wave head signal with a high signal-to-noise ratio is reconstructed by inverse wavelet packet transform.

[0017] Optionally, the calculation formula of the sub-band adaptive soft threshold is:

[0018]

[0019] The calculation formula for the soft threshold filtering process is:

[0020]

[0021] in, For the Sub-band adaptive soft threshold, For the The empirical adjustment coefficient of the sub-band, is the noise standard deviation, For the The number of sub-band samples, is the first The kth sample signal of the sub-band, is the first The kth sample signal of the sub-band.

[0022] Optionally, the step of synchronously performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal to generate corresponding time-frequency feature matrices and energy mutation sequences respectively includes:

[0023] Synchronously performing short-time Fourier transform and continuous wavelet transform on the reconstructed double-ended traveling wave head signal to generate a corresponding time-frequency feature matrix;

[0024] The instantaneous energy envelope is extracted through Hilbert transform, and the mean and standard deviation of the envelope in each window are calculated based on the sliding window;

[0025] The energy mutation points are identified according to the envelope mutation threshold and judgment conditions to form an energy mutation sequence.

[0026] Optionally, the determination condition of the energy mutation point is:

[0027] and

[0028]

[0029] The calculation formulas for the envelope mean and standard deviation within the window are:

[0030]

[0031]

[0032] in, is the instantaneous energy envelope value at time t, is the envelope mutation threshold at time t, 、 are the envelope mean and standard deviation in the window corresponding to time t, is the experience factor, and L is the window length.

[0033] Optionally, the formula for determining the triggering moment of the double-terminal high-resistance fault is:

[0034]

[0035] in, is the triggering moment of the double-ended high-resistance fault, is the confidence of the lightweight timing model for the moment, is the confidence threshold, is the energy mutation point moment, To allow time tolerance.

[0036] Optionally, the calculation formulas for the reflection coefficient and the transmission coefficient are:

[0037]

[0038] Where R is the reflection coefficient, T is the transmission coefficient, is the reflected wave amplitude, is the transmitted wave amplitude, is the incident wave amplitude.

[0039] A dual-terminal traveling wave positioning optimization device, comprising:

[0040] The decomposition and reconstruction unit is used to perform multi-level wavelet packet decomposition on the original traveling wave signal collected by the two terminals, generate sub-band adaptive soft thresholds according to the statistical characteristics of the noise, and reconstruct the two-terminal traveling wave head signal with high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform;

[0041] An envelope mutation unit is used to synchronously perform time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal, and generate corresponding time-frequency feature matrix and energy mutation sequence respectively;

[0042] A fault time unit is used to input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output a double-terminal high-resistance fault triggering time in combination with the confidence threshold and the spatiotemporal constraint of the mutation point;

[0043] a time difference compensation unit, configured to determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation;

[0044] The correction optimization unit is used to dynamically update the wave velocity parameters and interface correction coefficients using the recursive least squares method based on the time difference compensation results and the actual fault location feedback.

[0045] A dual-terminal traveling wave positioning optimization device, comprising a memory and a processor;

[0046] The memory is used to store programs;

[0047] The processor is used to execute the program to implement each step of the double-terminal traveling wave positioning optimization method as described in any one of the above items.

[0048] A readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the computer program implements the various steps of the double-terminal traveling wave positioning optimization method as described in any one of the above items.

[0049] A computer program product comprises a computer program, wherein when the computer program is run by a processor, the computer program executes each step of the two-terminal traveling wave positioning optimization method as described in any one of the above items.

[0050] As can be seen from the above technical solutions, the embodiments of the present application provide a dual-end traveling wave positioning optimization method, apparatus, device, and readable storage medium. These methods perform multi-level wavelet packet decomposition on the original traveling wave signal, generate subband adaptive soft thresholds based on noise statistics, and reconstruct a high signal-to-noise ratio traveling wave head signal through soft threshold filtering and inverse wavelet packet transform. This strategy, which combines multi-resolution wavelet packet decomposition with adaptive threshold filtering, can suppress power frequency interference and high-frequency impulse noise in different frequency bands, preserving the traveling wave head details to the greatest extent possible. This effectively addresses the difficulty of traditional methods in balancing signal integrity and noise suppression, significantly improving the signal-to-noise ratio and extraction accuracy of the traveling wave head.

[0051] Next, the reconstructed signal is subjected to simultaneous time-frequency transformation and instantaneous energy envelope analysis to generate a time-frequency feature matrix and energy mutation sequence. This is then fed into a lightweight timing model and, combined with constraints, outputs the fault trigger moment. This process integrates the spectral distribution and energy peak variation trends, significantly improving the detection sensitivity of weak high-resistance fault signals in low signal-to-noise ratio environments and effectively reducing the missed detection rate.

[0052] Finally, the original time difference is determined based on the difference in the triggering time of the two-terminal fault. The reflection and transmission coefficients are calculated using the measured wave amplitude. A propagation velocity model is selected based on the line type to compensate for the time difference. The wave velocity parameters and interface correction coefficients are dynamically updated using a recursive least squares method. This process enables dynamic estimation of interface coefficients, precise compensation of time difference, and continuous correction of positioning parameters, enabling the algorithm to quickly adapt to different media conditions. This overcomes the limitations of existing algorithms based on a single propagation velocity assumption or fixed models, effectively controls positioning errors, and significantly improves positioning accuracy on complex lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0054] Figure 1 This is a flow chart of a dual-terminal traveling wave positioning optimization method disclosed in an embodiment of the present application;

[0055] Figure 2 A schematic diagram of a dual-terminal traveling wave positioning optimization device disclosed in an embodiment of the present application;

[0056] Figure 3 This is a hardware structure block diagram of a dual-terminal traveling wave positioning optimization device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0058] The present application can be used in a variety of general or special computing device environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multi-processor devices, and distributed computing environments including any of the above devices or devices.

[0059] Next, we will introduce the application scheme. This application proposes the following technical scheme, please see below for details.

[0060] Figure 1 This is a flowchart of a dual-terminal traveling wave positioning optimization method disclosed in an embodiment of the present application.

[0061] like Figure 1 As shown, the method may include:

[0062] Step S1: perform multi-level wavelet packet decomposition on the original traveling wave signal collected by the two terminals, generate a sub-band adaptive soft threshold according to the noise statistical characteristics, and reconstruct a two-terminal traveling wave head signal with high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform.

[0063] Specifically, the original traveling wave signal collected from both ends is decomposed by multi-level wavelet packets, and a sub-band adaptive soft threshold is generated according to the statistical characteristics of the noise. The dual-end traveling wave head signal with high signal-to-noise ratio is reconstructed through soft threshold filtering and inverse wavelet packet transform. Specifically, first, multi-level wavelet packet decomposition is performed on the original traveling wave signal collected by the dual-end measurement device, and the signal is decomposed layer by layer into sub-band signals of different frequency bands (such as high-frequency sub-band and low-frequency sub-band), so as to realize multi-resolution frequency domain feature separation; the noise standard deviation is estimated on the first-level high-frequency sub-band by the median absolute deviation (MAD) algorithm, and the noise level is quantified; for each sub-band, the adaptive soft threshold is calculated based on the number of sub-band samples and the empirical adjustment coefficient, so that the threshold is dynamically adjusted with the sub-band noise characteristics; the soft threshold function is applied to each sub-band signal for filtering, so as to suppress the power frequency interference and high-frequency pulse noise while retaining the traveling wave head details; finally, the processed sub-band signal is reconstructed by inverse wavelet packet transform to obtain a dual-end traveling wave head signal with a high signal-to-noise ratio, which solves the technical problem that traditional fixed threshold filtering is difficult to take into account both signal integrity and noise suppression, and improves the accuracy of traveling wave feature extraction.

[0064] Step S2: synchronously perform time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal to generate corresponding time-frequency feature matrix and energy mutation sequence respectively.

[0065] Specifically, the reconstructed double-ended traveling wave head signal is subjected to time-frequency transformation and instantaneous energy envelope analysis simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence respectively. Specifically, the reconstructed double-ended traveling wave head signal is simultaneously subjected to short-time Fourier transform (STFT) and continuous wavelet transform (CWT). The localized characteristics of the signal in the time-frequency domain are obtained by short-time Fourier transform (STFT), and the multi-scale time-frequency characteristics of the signal are captured by continuous wavelet transform (CWT). The two are combined to generate a time-frequency feature matrix to comprehensively characterize the spectral distribution of the signal. The signal is processed using Hilbert transform to obtain the instantaneous energy envelope. The mean and standard deviation of the envelope in the window are calculated using a sliding window of a set length. The window threshold is set based on an empirical factor. When the envelope value at a certain moment meets the conditions of exceeding the sum of the mean and the threshold and the envelope change rate is greater than zero, it is determined to be an energy mutation point, thereby generating an energy mutation sequence. This process integrates the spectral distribution and the energy peak change trend, providing a multi-dimensional feature basis for the subsequent detection of weak high-resistance fault signals, and effectively improving the capture accuracy of the fault triggering moment in a low signal-to-noise ratio environment.

[0066] Step S3: input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering moment of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point.

[0067] Specifically, the time-frequency feature matrix and the energy mutation sequence are input into the lightweight timing model. The lightweight timing model analyzes each moment and provides a confidence level for each moment. A confidence threshold is pre-set; only when the confidence level of a moment exceeds this threshold is it considered further. Furthermore, considering the spatiotemporal characteristics of the energy mutation point, spatiotemporal constraints are imposed on the mutation point, requiring that the time difference between the selected moment and the energy mutation point be within the allowable time tolerance. Among the moments that meet these conditions, the earliest moment with a confidence level greater than the confidence threshold and a difference from the energy mutation point within the allowable time tolerance is selected as the triggering moment of the double-terminal high-resistance fault.

[0068] The formula for determining the triggering time of the double-terminal high-resistance fault is:

[0069]

[0070] in, is the triggering moment of the double-ended high-resistance fault, is the confidence of the lightweight timing model for the moment, is the confidence threshold, is the energy mutation point moment, To allow time tolerance.

[0071] Step S4: Determine the difference between the triggering moments of the double-ended high-resistance fault as the double-ended original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select the propagation velocity model based on the line section type to perform time difference compensation.

[0072] Specifically, after determining the triggering time of the double-ended high-resistance fault, the difference between the fault triggering times at both ends is calculated to obtain the original double-ended time difference. At the junction of the overhead section and the cable section, the incident wave amplitude, reflected wave amplitude, and transmitted wave amplitude are actually measured. The reflection coefficient is calculated by dividing the reflected wave amplitude by the incident wave amplitude, and the transmission coefficient is calculated by dividing the transmitted wave amplitude by the incident wave amplitude. Depending on the specific type of line, if it is an overhead line, the overhead line propagation velocity model is selected, and if it is a cable line, the cable line propagation velocity model is selected. If it is an overhead section, the time difference remains unchanged; if it is a cable section, the original time difference is multiplied by the ratio of the overhead line propagation velocity to the cable line propagation velocity to complete the time difference compensation.

[0073] ① Based on the measured wave amplitude values (including reflected wave amplitude, transmitted wave amplitude, and incident wave amplitude) at the junction of the overhead section and the cable section, the reflection coefficient and transmission coefficient are calculated.

[0074] The calculation formulas for the reflection coefficient and the transmission coefficient are:

[0075]

[0076] Where R is the reflection coefficient, T is the transmission coefficient, is the reflected wave amplitude, is the transmitted wave amplitude, is the incident wave amplitude.

[0077] ② Select the overhead line propagation velocity model (propagation velocity) according to the line type ) or cable line propagation velocity model (propagation velocity ), original time difference compensation for:

[0078]

[0079] is the original time difference between the two ends, that is, the difference between the triggering moments of the high-resistance fault at the two ends.

[0080] Step S5: Based on the time difference compensation result and the actual fault position feedback, the wave velocity parameters and the interface correction coefficient are dynamically updated using the recursive least squares method.

[0081] Specifically, after obtaining the time difference compensation result, it is combined with the actual fault location feedback information. The recursive least squares method is an algorithm that can continuously update the estimated parameters based on new data. Here, we regard the wave velocity parameters (including the wave velocity of the overhead line and the wave velocity of the cable line) and the interface correction coefficient (the coefficient reflecting the propagation characteristics of the traveling wave at different line interfaces) as parameters to be updated. Every time there is new fault location feedback information, the algorithm will adjust the above parameters based on this new data. Through continuous iterative calculations, the wave velocity parameters and interface correction coefficients can continuously adapt to changes that may occur during line operation, such as the impact of changes in ambient temperature on cable wave velocity, or changes in propagation characteristics caused by line aging. In this way, the positioning algorithm can maintain good convergence and accuracy under different working conditions, achieving more accurate positioning of the fault location.

[0082] ① Construct parameter vector by feedback of each fault location result and measurement models .

[0083] in, For the The observed time difference of the secondary fault (including the overhead section and the cable section); is the regression vector, are the equivalent path lengths of the overhead and cable segments, respectively; To measure noise.

[0084] ②Use recursive least squares method to dynamically update the wave velocity parameters and interface correction coefficients, that is, update Parameters are used to ensure the convergence and accuracy of the algorithm under different working conditions.

[0085] As can be seen from the above technical solutions, the embodiments of the present application provide a dual-end traveling wave positioning optimization method, apparatus, device, and readable storage medium. These methods perform multi-level wavelet packet decomposition on the original traveling wave signal, generate subband adaptive soft thresholds based on noise statistics, and reconstruct a high signal-to-noise ratio traveling wave head signal through soft threshold filtering and inverse wavelet packet transform. This strategy, which combines multi-resolution wavelet packet decomposition with adaptive threshold filtering, can suppress power frequency interference and high-frequency impulse noise in different frequency bands, preserving the traveling wave head details to the greatest extent possible. This effectively addresses the difficulty of traditional methods in balancing signal integrity and noise suppression, significantly improving the signal-to-noise ratio and extraction accuracy of the traveling wave head.

[0086] Next, the reconstructed signal is subjected to simultaneous time-frequency transformation and instantaneous energy envelope analysis to generate a time-frequency feature matrix and energy mutation sequence. This is then fed into a lightweight timing model and, combined with constraints, outputs the fault trigger moment. This process integrates the spectral distribution and energy peak variation trends, significantly improving the detection sensitivity of weak high-resistance fault signals in low signal-to-noise ratio environments and effectively reducing the missed detection rate.

[0087] Finally, the original time difference is determined based on the difference in the triggering time of the two-terminal fault. The reflection and transmission coefficients are calculated using the measured wave amplitude. A propagation velocity model is selected based on the line type to compensate for the time difference. The wave velocity parameters and interface correction coefficients are dynamically updated using a recursive least squares method. This process enables dynamic estimation of interface coefficients, precise compensation of time difference, and continuous correction of positioning parameters, enabling the algorithm to quickly adapt to different media conditions. This overcomes the limitations of existing algorithms based on a single propagation velocity assumption or fixed models, effectively controls positioning errors, and significantly improves positioning accuracy on complex lines.

[0088] In some embodiments of the present application, the process of step S1, performing multi-level wavelet packet decomposition on the original traveling wave signal collected from both ends, generating a sub-band adaptive soft threshold according to the statistical characteristics of the noise, and reconstructing a two-end traveling wave head signal with a high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform is introduced. Specifically, it may include:

[0089] Step S11: performing multi-level wavelet packet decomposition on the original traveling wave signal collected by both ends to obtain high-frequency sub-bands and low-frequency sub-bands;

[0090] Step S12: estimating the noise standard deviation using the median absolute deviation method on the first-level high-frequency sub-band;

[0091] Step S13: Calculating the sub-band adaptive soft threshold of each sub-band based on the noise standard deviation, the number of sub-band samples and the empirical adjustment coefficient;

[0092] Step S14: applying soft threshold filtering to the sub-bands at each level according to the sub-band adaptive soft threshold, and reconstructing a double-ended traveling wave head signal with a high signal-to-noise ratio through inverse wavelet packet transform.

[0093] Specifically, a multi-level wavelet packet decomposition operation is first performed on the original traveling wave signal collected from both ends. Wavelet packet decomposition is a multi-resolution analysis method that can gradually subdivide the original signal according to its different frequency components. Through this decomposition, the signal is divided into high-frequency subbands and low-frequency subbands. The high-frequency subband contains the most volatile parts of the signal and is often mixed with more noise interference; the low-frequency subband, on the other hand, reflects the general trend and outline of the signal, providing a basis for subsequent processing.

[0094] After completing the multi-level wavelet packet decomposition, we focus on the first-level high-frequency subband. The median absolute deviation method is used to estimate the standard deviation of the noise in this subband. This robust statistical estimation method calculates the median of the absolute deviations of each data point from the median in the dataset and then performs a specific transformation. This method accurately reflects the degree of noise dispersion in this subband, thereby obtaining the noise standard deviation. This standard deviation is crucial for subsequent threshold determination, quantifying the intensity of the noise and providing a key parameter for further processing.

[0095] Based on the noise standard deviation obtained in step S12, combined with the number of samples in each subband and the empirical adjustment coefficient, the subband-adaptive soft threshold for each subband is calculated. Because different subbands have different sample numbers and contain different noise characteristics, it is necessary to calculate an appropriate threshold for each subband. The empirical adjustment coefficient is determined based on extensive practical experience and experimental data. It can fine-tune the threshold, making the calculated threshold more consistent with the actual noise conditions in each subband, thereby more effectively suppressing noise.

[0096] After obtaining the subband-adaptive soft threshold for each subband, soft threshold filtering is applied to each subband. The principle of soft threshold filtering is that when the signal value in a subband is less than the threshold, it is approximated to 0, effectively removing small fluctuations that may be noise. When the signal value is greater than the threshold, it is shrunk to a certain extent, further suppressing noise while preserving the main signal characteristics. After soft threshold filtering, the processed subband signals are recombined through an inverse wavelet packet transform to recover a two-ended traveling wave head signal with a high signal-to-noise ratio. This reconstructed signal provides a more reliable foundation for subsequent fault detection and analysis.

[0097] ① The original traveling wave signal collected by both ends Perform wavelet packet decomposition to obtain high-frequency subbands and low-frequency subband ;

[0098] ② Using the median absolute deviation (MAD) method, in the first-level high-frequency sub-band The noise standard deviation is estimated above. The noise standard deviation calculation formula is:

[0099]

[0100] ③For the Level subband, calculate subband adaptive soft threshold , the calculation formula of the sub-band adaptive soft threshold is:

[0101]

[0102] ④ Apply soft threshold filtering to the sub-bands at each level. The calculation formula of the soft threshold filtering is:

[0103]

[0104] in, For the Sub-band adaptive soft threshold, For the The empirical adjustment coefficient of the sub-band, is the noise standard deviation, For the The number of sub-band samples, is the first The kth sample signal of the sub-band, is the first The kth sample signal of the sub-band.

[0105] Reconstructing the double-ended traveling wave head signal with high signal-to-noise ratio through inverse wavelet packet transform .

[0106] In some embodiments of the present application, the process of step S2, which is to simultaneously perform time-frequency transformation and instantaneous energy envelope analysis on the reconstructed dual-end traveling wave head signal to generate corresponding time-frequency feature matrices and energy mutation sequences, is introduced. Specifically, it may include:

[0107] Step S21, synchronously performing short-time Fourier transform and continuous wavelet transform on the reconstructed double-ended traveling wave head signal to generate a corresponding time-frequency feature matrix;

[0108] Step S22: extracting the instantaneous energy envelope through Hilbert transform, and calculating the envelope mean and standard deviation in each window based on the sliding window;

[0109] Step S23: Identify energy mutation points according to the envelope mutation threshold and the judgment condition to form an energy mutation sequence.

[0110] Specifically, the reconstructed double-ended traveling wave head signal is subjected to a short-time Fourier transform (STFT) and a continuous wavelet transform (CWT) simultaneously to generate a time-frequency feature matrix. The short-time Fourier transform (STFT) divides the signal into multiple short-time segments using a sliding window function. Each segment is Fourier transformed to form an amplitude distribution matrix on the time-frequency two-dimensional plane, capturing the signal's local time-domain characteristics at specific frequency bands. The continuous wavelet transform (CWT) uses wavelet basis functions of different scales to perform multi-resolution analysis on the signal, generating a wavelet coefficient matrix containing scale-time information to characterize the signal's time-frequency characteristics at different frequency scales. The combined time-frequency feature matrix can comprehensively and multidimensionally characterize the spectral distribution and time-frequency localization characteristics of the traveling wave signal, providing rich frequency-domain information for subsequent fault feature identification.

[0111] The instantaneous energy envelope of the signal is extracted through the Hilbert transform. Specifically, the reconstructed signal is analyzed and processed to calculate the instantaneous envelope value of its amplitude. This envelope value reflects the temporal trend of the signal energy. Subsequently, a sliding window of length L is set, and the mean and standard deviation of the envelope within the window are calculated for each time instant t. The envelope mean represents the average energy level within the window, while the standard deviation reflects the severity of energy fluctuations. Combined, the two quantify the local statistical characteristics of the signal energy and provide benchmark parameters for subsequent mutation point detection.

[0112] Based on the envelope mutation threshold and judgment criteria, energy mutation points are identified and an energy mutation sequence is formed. This sequence effectively captures weak fault signals by integrating energy amplitude mutations with temporal variation trends, providing key time-domain characteristics for precise positioning of subsequent fault triggering moments.

[0113] ① Reconstructed double-ended traveling wave head signal with high signal-to-noise ratio Simultaneously perform short-time Fourier transform (STFT) and continuous wavelet transform (CWT):

[0114]

[0115] ②Use Hilbert transform to obtain instantaneous envelope , choose the length Window, for each moment Calculate the envelope mean and standard deviation within the window:

[0116] The calculation formulas for the envelope mean and standard deviation within the window are:

[0117]

[0118]

[0119] ③ According to the envelope mutation threshold and judgment conditions, the energy mutation point is identified to form an energy mutation sequence:

[0120] If at the time If the determination conditions of the energy mutation point are met, it is considered that an envelope mutation occurs here, and this moment is recorded as the energy mutation point. .

[0121] The determination conditions of the energy mutation point are:

[0122] and

[0123]

[0124] in, is the instantaneous energy envelope value at time t, is the envelope mutation threshold at time t, 、 are the envelope mean and standard deviation in the window corresponding to time t, is the experience factor, and L is the window length.

[0125] A dual-terminal traveling wave positioning optimization device provided in an embodiment of the present application is described below. The dual-terminal traveling wave positioning optimization device described below and the dual-terminal traveling wave positioning optimization method described above can be referenced to each other.

[0126] See also Figure 2 , Figure 2 This is a schematic diagram of a dual-terminal traveling wave positioning optimization device disclosed in an embodiment of the present application.

[0127] like Figure 2 As shown, the dual-end traveling wave positioning optimization device may include:

[0128] The decomposition and reconstruction unit 110 is used to perform multi-level wavelet packet decomposition on the original traveling wave signal collected by the two terminals, generate a sub-band adaptive soft threshold according to the noise statistical characteristics, and reconstruct the two-terminal traveling wave head signal with high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform;

[0129] The envelope mutation unit 120 is used to synchronously perform time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal, and generate corresponding time-frequency feature matrix and energy mutation sequence respectively;

[0130] The fault time unit 130 is used to input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering time of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point;

[0131] a time difference compensation unit 140 for determining the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, combining the reflection coefficient and the transmission coefficient calculated based on the measured wave amplitude at the junction of the overhead section and the cable section, and selecting a propagation velocity model based on the line section type to perform time difference compensation;

[0132] The correction optimization unit 150 is used to dynamically update the wave velocity parameters and the interface correction coefficient using a recursive least square method according to the time difference compensation result and the actual fault position feedback.

[0133] As can be seen from the above technical solutions, the embodiments of the present application provide a dual-end traveling wave positioning optimization method, apparatus, device, and readable storage medium. These methods perform multi-level wavelet packet decomposition on the original traveling wave signal, generate subband adaptive soft thresholds based on noise statistics, and reconstruct a high signal-to-noise ratio traveling wave head signal through soft threshold filtering and inverse wavelet packet transform. This strategy, which combines multi-resolution wavelet packet decomposition with adaptive threshold filtering, can suppress power frequency interference and high-frequency impulse noise in different frequency bands, preserving the traveling wave head details to the greatest extent possible. This effectively addresses the difficulty of traditional methods in balancing signal integrity and noise suppression, significantly improving the signal-to-noise ratio and extraction accuracy of the traveling wave head.

[0134] Next, the reconstructed signal is subjected to simultaneous time-frequency transformation and instantaneous energy envelope analysis to generate a time-frequency feature matrix and energy mutation sequence. This is then fed into a lightweight timing model and, combined with constraints, outputs the fault trigger moment. This process integrates the spectral distribution and energy peak variation trends, significantly improving the detection sensitivity of weak high-resistance fault signals in low signal-to-noise ratio environments and effectively reducing the missed detection rate.

[0135] Finally, the original time difference is determined based on the difference in the triggering time of the two-terminal fault. The reflection and transmission coefficients are calculated using the measured wave amplitude. A propagation velocity model is selected based on the line type to compensate for the time difference. The wave velocity parameters and interface correction coefficients are dynamically updated using a recursive least squares method. This process enables dynamic estimation of interface coefficients, precise compensation of time difference, and continuous correction of positioning parameters, enabling the algorithm to quickly adapt to different media conditions. This overcomes the limitations of existing algorithms based on a single propagation velocity assumption or fixed models, effectively controls positioning errors, and significantly improves positioning accuracy on complex lines.

[0136] Optionally, performing multi-level wavelet packet decomposition on the original traveling wave signal collected from both ends, generating a sub-band adaptive soft threshold according to noise statistical characteristics, and reconstructing a two-end traveling wave head signal with a high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform, includes:

[0137] Perform multi-level wavelet packet decomposition on the original traveling wave signal collected by both ends to obtain high-frequency sub-bands and low-frequency sub-bands;

[0138] The median absolute deviation method is used to estimate the noise standard deviation on the first-level high-frequency sub-band;

[0139] Calculating the sub-band adaptive soft threshold of each sub-band based on the noise standard deviation and combining the number of sub-band samples and the empirical adjustment coefficient;

[0140] Soft threshold filtering is applied to the sub-bands at each level according to the sub-band adaptive soft threshold, and a double-ended traveling wave head signal with a high signal-to-noise ratio is reconstructed by inverse wavelet packet transform.

[0141] Optionally, the calculation formula of the sub-band adaptive soft threshold is:

[0142]

[0143] The calculation formula for the soft threshold filtering process is:

[0144]

[0145] in, For the Sub-band adaptive soft threshold, For the The empirical adjustment coefficient of the sub-band, is the noise standard deviation, For the The number of sub-band samples, is the first The kth sample signal of the sub-band, is the first The kth sample signal of the sub-band.

[0146] Optionally, the step of synchronously performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal to generate corresponding time-frequency feature matrices and energy mutation sequences respectively includes:

[0147] Synchronously performing short-time Fourier transform and continuous wavelet transform on the reconstructed double-ended traveling wave head signal to generate a corresponding time-frequency feature matrix;

[0148] The instantaneous energy envelope is extracted through Hilbert transform, and the mean and standard deviation of the envelope in each window are calculated based on the sliding window;

[0149] The energy mutation points are identified according to the envelope mutation threshold and judgment conditions to form an energy mutation sequence.

[0150] Optionally, the determination condition of the energy mutation point is:

[0151] and

[0152]

[0153] The calculation formulas for the envelope mean and standard deviation within the window are:

[0154]

[0155]

[0156] in, is the instantaneous energy envelope value at time t, is the envelope mutation threshold at time t, 、 are the envelope mean and standard deviation in the window corresponding to time t, is the experience factor, and L is the window length.

[0157] Optionally, the formula for determining the triggering moment of the double-terminal high-resistance fault is:

[0158]

[0159] in, is the triggering moment of the double-ended high-resistance fault, is the confidence of the lightweight timing model for the moment, is the confidence threshold, is the energy mutation point moment, To allow time tolerance.

[0160] Optionally, the calculation formulas for the reflection coefficient and the transmission coefficient are:

[0161]

[0162] Where R is the reflection coefficient, T is the transmission coefficient, is the reflected wave amplitude, is the transmitted wave amplitude, is the incident wave amplitude.

[0163] The dual-terminal traveling wave positioning optimization device provided in the embodiment of the present application can be applied to dual-terminal traveling wave positioning optimization equipment. Figure 3 The hardware structure diagram of the double-ended traveling wave positioning optimization device is shown. Figure 3 ,The hardware structure of the double-terminal traveling wave positioning optimization device may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4;

[0164] In the embodiment of the present application, the number of the processor 1, the communication interface 2, the memory 3, and the communication bus 4 is at least one, and the processor 1, the communication interface 2, and the memory 3 communicate with each other through the communication bus 4;

[0165] The processor 1 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention;

[0166] The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory;

[0167] The memory stores a program, and the processor can call the program stored in the memory, wherein the program is used to:

[0168] The original traveling wave signal collected by the two terminals is decomposed by multi-level wavelet packets, and a sub-band adaptive soft threshold is generated according to the statistical characteristics of the noise. The two-terminal traveling wave head signal with high signal-to-noise ratio is reconstructed through soft threshold filtering and inverse wavelet packet transform.

[0169] Performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence respectively;

[0170] Input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering moment of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point;

[0171] Determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation;

[0172] According to the time difference compensation results and the actual fault location feedback, the recursive least squares method is used to dynamically update the wave velocity parameters and interface correction coefficients.

[0173] Optionally, the refined functions and extended functions of the program may refer to the above description.

[0174] The present application also provides a readable storage medium, which may store a program suitable for execution by a processor, wherein the program is used to:

[0175] The original traveling wave signal collected by the two terminals is decomposed by multi-level wavelet packets, and a sub-band adaptive soft threshold is generated according to the statistical characteristics of the noise. The two-terminal traveling wave head signal with high signal-to-noise ratio is reconstructed through soft threshold filtering and inverse wavelet packet transform.

[0176] Performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence respectively;

[0177] Input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering moment of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point;

[0178] Determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation;

[0179] According to the time difference compensation results and the actual fault location feedback, the recursive least squares method is used to dynamically update the wave velocity parameters and interface correction coefficients.

[0180] Optionally, the refined functions and extended functions of the program may refer to the above description.

[0181] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0182] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0183] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A two-terminal traveling wave positioning optimization method, characterized in that: include: The original traveling wave signal collected by the two terminals is decomposed by multi-level wavelet packets, and a sub-band adaptive soft threshold is generated according to the statistical characteristics of the noise. The two-terminal traveling wave head signal with high signal-to-noise ratio is reconstructed through soft threshold filtering and inverse wavelet packet transform. Performing time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence respectively; Input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output the triggering moment of the double-terminal high-resistance fault by combining the confidence threshold and the spatiotemporal constraint of the mutation point; Determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation; According to the time difference compensation results and the actual fault location feedback, the recursive least squares method is used to dynamically update the wave velocity parameters and interface correction coefficients.

2. The method according to claim 1, characterized in that The method performs multi-level wavelet packet decomposition on the original traveling wave signal collected at both ends, generates a sub-band adaptive soft threshold according to the noise statistical characteristics, and reconstructs the dual-end traveling wave head signal with a high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform, including: Perform multi-level wavelet packet decomposition on the original traveling wave signal collected by both ends to obtain high-frequency sub-bands and low-frequency sub-bands; The median absolute deviation method is used to estimate the noise standard deviation on the first-level high-frequency sub-band; Calculating the sub-band adaptive soft threshold of each sub-band based on the noise standard deviation and combining the number of sub-band samples and the empirical adjustment coefficient; Soft threshold filtering is applied to the sub-bands at each level according to the sub-band adaptive soft threshold, and a double-ended traveling wave head signal with a high signal-to-noise ratio is reconstructed by inverse wavelet packet transform.

3. The method according to claim 2, characterized in that The calculation formula of the sub-band adaptive soft threshold is: The calculation formula for the soft threshold filtering process is: in, For the Sub-band adaptive soft threshold, For the The empirical adjustment coefficient of the sub-band, is the noise standard deviation, For the The number of sub-band samples, is the first The kth sample signal of the sub-band, is the first The kth sample signal of the sub-band.

4. The method according to claim 1, wherein The reconstructed double-ended traveling wave head signal is subjected to time-frequency transformation and instantaneous energy envelope analysis simultaneously to generate corresponding time-frequency feature matrix and energy mutation sequence, respectively, including: Synchronously performing short-time Fourier transform and continuous wavelet transform on the reconstructed double-ended traveling wave head signal to generate a corresponding time-frequency feature matrix; The instantaneous energy envelope is extracted through Hilbert transform, and the mean and standard deviation of the envelope in each window are calculated based on the sliding window; The energy mutation points are identified according to the envelope mutation threshold and judgment conditions to form an energy mutation sequence.

5. The method according to claim 4, characterized in that The determination conditions of the energy mutation point are: and The calculation formulas for the envelope mean and standard deviation within the window are: in, is the instantaneous energy envelope value at time t, is the envelope mutation threshold at time t, 、 are the envelope mean and standard deviation in the window corresponding to time t, is the experience factor, and L is the window length.

6. The method according to claim 1, characterized in that The formula for determining the triggering time of the double-terminal high-resistance fault is: in, is the triggering moment of the double-ended high-resistance fault, is the confidence of the lightweight timing model for the moment, is the confidence threshold, is the energy mutation point moment, To allow time tolerance.

7. The method according to claim 1, characterized in that The calculation formulas for the reflection coefficient and the transmission coefficient are: Where R is the reflection coefficient, T is the transmission coefficient, is the reflected wave amplitude, is the transmitted wave amplitude, is the incident wave amplitude.

8. A double-ended traveling wave positioning optimization device, characterized in that: include: The decomposition and reconstruction unit is used to perform multi-level wavelet packet decomposition on the original traveling wave signal collected by the two terminals, generate sub-band adaptive soft thresholds according to the statistical characteristics of the noise, and reconstruct the two-terminal traveling wave head signal with high signal-to-noise ratio through soft threshold filtering and inverse wavelet packet transform; An envelope mutation unit is used to synchronously perform time-frequency transformation and instantaneous energy envelope analysis on the reconstructed double-ended traveling wave head signal, and generate corresponding time-frequency feature matrix and energy mutation sequence respectively; A fault time unit is used to input the time-frequency feature matrix and the energy mutation sequence into a lightweight timing model, and output a double-terminal high-resistance fault triggering time in combination with the confidence threshold and the spatiotemporal constraint of the mutation point; a time difference compensation unit, configured to determine the difference between the triggering moments of the double-end high-resistance fault as the double-end original time difference, calculate the reflection coefficient and transmission coefficient based on the measured wave amplitude at the junction of the overhead section and the cable section, and select a propagation velocity model based on the line section type to perform time difference compensation; The correction optimization unit is used to dynamically update the wave velocity parameters and interface correction coefficients using the recursive least squares method based on the time difference compensation results and the actual fault location feedback.

9. A double-ended traveling wave positioning optimization device, characterized in that: including memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the dual-terminal traveling wave positioning optimization method according to any one of claims 1 to 7.

10. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, each step of the dual-terminal traveling wave positioning optimization method according to any one of claims 1 to 7 is implemented.

Images