Method based on time-frequency analysis and for extracting typical traveling wave feature parameters

Abstract

Disclosed in the present invention is a method based on time-frequency analysis and for extracting typical traveling wave feature parameters, the method comprising the following steps: collecting a traveling wave signal, obtaining collection point position information and fault point position information of the traveling wave signal, and using the traveling wave signal as a first signal; performing interval sampling on the first signal on the basis of a time-domain sequence to generate a second signal, and selecting the second signal on the basis of a frequency-domain interval to generate a third signal; performing discrete sampling on the third signal and then performing recombination to form a fourth signal; and performing feature extraction on the fourth signal to obtain waveform features, and acquiring an association relationship between the waveform features and the collection point position information and the fault point position information. The present invention can extract waveform features of a traveling wave signal in the case of a small sample size, thereby improving the traveling wave signal identification capability; and by means of calculating a correlation degree in sampled information, the traveling wave signal is identified on the basis of a correlation degree interval where a correlation integral is located, so as to identify a fault type.

Description

A method for extracting characteristic parameters of typical traveling waves based on time-frequency analysis Technical Field

[0001] This invention belongs to the field of power equipment fault analysis technology, specifically relating to a method for extracting typical traveling wave characteristic parameters based on time-frequency analysis. Background Technology

[0002] Currently, traveling wave fault location methods theoretically possess high location accuracy, leading to their widespread research and application. The timing and location of abrupt changes in the traveling wave signal represent specific fault information; therefore, accurately detecting these abrupt changes is crucial for traveling wave fault location. Analysis methods such as wavelet analysis and Hilbert-Huang transform have been widely applied in traveling wave identification, achieving good fault location results. However, there are challenges: attenuation during traveling wave transmission weakens the singularity of the traveling wave; identification of lightning interference when the line is not faulty after a lightning strike; and accurate location of the flashover point when the lightning strike point and flashover point are inconsistent, requiring identification and detection of the second traveling wave front (the reflected wave from the fault point) on the lightning-struck side. In single-ended traveling wave fault location methods, accurate identification of the reflected wave from the fault point and the reflected wave from the opposite bus is also critical. Furthermore, noise interference during the transmission and sampling of the traveling wave signal prevents the effective extraction of useful signals, making it difficult to guarantee the reliability and accuracy of existing traveling wave identification methods that extract features in the time or time-frequency domains. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for extracting typical traveling wave feature parameters based on time-frequency analysis that can improve the recognition ability of traveling wave signals with a small sample size and ensure reliability and accuracy.

[0004] The technical solution of this invention is as follows:

[0005] A method for extracting characteristic parameters of typical traveling waves based on time-frequency analysis includes the following steps:

[0006] Step S1: Acquire traveling wave signals, obtain the acquisition point location information and fault point location information of the traveling wave signals, and use the traveling wave signals as the first signal;

[0007] Step S2: The first signal is sampled at intervals based on the time-domain sequence to generate a second signal, and the second signal is selected based on the frequency-domain interval to generate a third signal;

[0008] Step S3: Discretely sample the third signal and recombine them to form the fourth signal;

[0009] Step S4: Perform feature extraction on the fourth signal to obtain waveform features, and obtain the correlation between the waveform features and the acquisition point location information and the fault point location information.

[0010] Further, step S1 includes the following steps:

[0011] Step S11: Set up traveling wave signal acquisition points and fault simulation points on the transmission line, and set up multiple traveling wave signal acquisition points, and record the location information of each traveling wave signal acquisition point.

[0012] Step S12: Simulate various types of transmission line faults and record the simulated fault points as fault location information;

[0013] Step S13: After the fault traveling wave signal acquisition point detects the traveling wave signal, it takes the traveling wave signal as the first signal.

[0014] Furthermore, the formula for calculating the location of the fault point is:

[0015] L f =L1+d f or L f =L2-d f ,

[0016] Where, d f T represents the distance from the fault point to the data collection point. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f The time it takes for the fault traveling wave signal to reach acquisition point 2 is represented by v, the traveling wave propagation speed is represented by v, L1 represents the location of acquisition point 1, and L2 represents the location of acquisition point 2. f Indicates the location of the fault.

[0017] Further, step S2 includes the following steps:

[0018] Step S21: Perform interval sampling on the first signal based on the time-domain sequence to generate a second signal;

[0019] Step S22: Perform interval sampling on the second signal based on the frequency domain interval to obtain sampling information;

[0020] Step S23: The sampled information obtained by interval sampling is spliced ​​together in time order to obtain the third signal;

[0021] Step S24: By changing the distance between frequency intervals, different frequency intervals are sampled at intervals to obtain sampling information of different frequency intervals and obtain multiple third signals.

[0022] Furthermore, the expression for generating the second signal by performing interval sampling on the first signal based on the time-domain sequence is:

[0023] Where, r i (t) represents the signal received in the i-th time interval, i.e., the second signal, A represents the amplitude of the second signal, f0 represents the center frequency of the second signal, θ0 represents the initial phase of the second signal, j represents the imaginary unit, w(t) represents the noise part of the second signal, i represents the index of the time segment, i.e., the current signal is in the i-th time interval, t represents the time variable, and T0 represents the length of the time segment, i.e. the duration of each interval;

[0024] The expression for obtaining sampling information by interval sampling of the second signal based on frequency domain intervals is:

[0025] Where X(k) represents the sampling information, that is, the sampled values ​​of a finite number of spectra in the second signal, and x(n) represents the nth discrete sampled value of the original signal. The sampling interval is represented by N, the total number of sampling points of the signal is represented by k, the frequency index is represented by n, the time index is represented by i, and the imaginary unit is represented by i.

[0026] The expression for the third signal is:

[0027] x k,m (n)=x(t k，m +nT k,m ), n = 0, 1, 2, ..., N k,m -1,

[0028] x third,k (n), k = 1, 2, ..., K

[0029] Where x(t) represents the original signal, x k,m (n) represents the sampled signal obtained from the k-th frequency interval and the m-th time interval, t k,m T represents the starting time point of sampling. k,m N represents the sampling interval, k represents the frequency interval index, m represents the segment number when performing interval sampling on the k-th frequency interval, and N represents the frequency interval index. k,m Let x represent the length of the m-th sampled signal within the k-th frequency interval. third,k (n) represents the third signal, K represents the total number of frequency intervals, and N k,j This represents the number of sampling points in the j-th segment of the signal within the k-th frequency interval, where n represents the sampling point index of the discrete signal.

[0030] Further, step S3 specifically involves: discretizing and recombining each of the third signals, then concatenating them to form a fourth signal, and obtaining the fourth signal group; the expression of the fourth signal is:

[0031] Where, r i (n) represents the received signal at the nth sampling point within the i-th time interval, i.e., the fourth signal; A represents the amplitude of the fourth signal; f0 represents the center frequency of the fourth signal; and Δt represents the sampling interval. The initial phase of the fourth signal is represented by w(n), the noise part of the fourth signal is represented by w(n), the index of the time segment is represented by i, the number of sampling points in each time segment is represented by N0, and the number of the signal sampling points is represented by n.

[0032] Further, step S4 includes the following steps:

[0033] Step S41: Establish the correlation vector degree model;

[0034] Step S42: Calculate the distance between the location information of each sampling point and the location information of the fault point in the sampling information. Given an observation threshold, when the relative distance is less than the observation threshold, it is called an associated vector pair. Then, count the associated vector pairs to obtain the relative statistical distance.

[0035] Step S43: Based on the actual distance between the acquisition point location information and the fault point location information, the waveform features are matched one-to-one with the actual distance between the acquisition point location information and the fault point location information, as well as the correlation interval, to form a correlation relationship. The traveling wave signal is identified based on the correlation relationship, and the fault point location is located.

[0036] Furthermore, the expression for the correlation vector degree model is:

[0037] Where C(q, r) represents the correlation vector degree value, q represents the singularity measure, r represents the observation regime of the correlation integral or the time delay or observation threshold of the time series, N represents the total number of samples, x represents the global signal eigenvalue, a represents the offset, and n represents the amplification factor. i Let x represent the value of the i-th sample point in the sample set. j Let H(x) represent the value of the j-th sample point in the sample set, and let H(x) represent the step function.

[0038] Furthermore, the relative distance in step S42 is the distance between the data collection point and the fault point, and the expression for the relative distance is: D ij =|L i -L f |,

[0039] Among them, D ij L represents the distance between data collection point i and fault point j. i L represents the location information of collection point i. f This indicates the location information of the fault point f;

[0040] When D ij When r < r, the collection point i and the fault point f form an association vector pair, and the counting formula is:

[0041] Where, N pair The number of associated vector pairs is represented by r, the observation threshold is represented by δ, and the decision function is represented by δ.

[0042] The formula for calculating the observation threshold r is: C(q,r)=r D ,

[0043] Where D represents the multifractal dimension and C(q, r) represents the correlation vector degree value.

[0044] Furthermore, the expression for the association relationship in step S43 is: R(L) i ,L f )=f(D ij ,φ),

[0045] Among them, R(L i ,L f ) indicates the collection point L i and fault point L f The relationship between D ij The distance between the acquisition point and the fault point is represented by φ, the waveform characteristics are represented by α and β, the adjustment coefficients are represented by v, and the traveling wave propagation speed is represented by T. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f L represents the time it takes for the fault traveling wave signal to arrive at acquisition point 2. i L represents the location of a known data collection point. f Indicates the location of the fault.

[0046] The beneficial effects of this invention are:

[0047] 1. High-precision fault location: By combining time-frequency analysis and correlation vector degree model, this invention can more accurately determine the location of the fault point, especially in complex environments (such as multipath propagation and noise interference). By using the traveling wave propagation time difference formula and combining correlation relationship, measurement error can be effectively reduced.

[0048] 2. Full-process feature extraction and analysis: From time-domain signal to frequency-domain sampling, and then to the construction and analysis of correlation, the steps are clear and progressive, ensuring that the key features of the traveling wave signal are fully extracted. Multi-level signal processing (first signal, second signal, third signal, fourth signal) can gradually filter out noise and highlight key features, providing a reliable data foundation for subsequent positioning.

[0049] 3. Multi-scale correlation characteristic analysis: The correlation vector degree model, combined with the flexible adjustment of the observation threshold r and singular measure q, can analyze the multi-scale correlation characteristics between the fault point and the collection point. By adjusting the parameters r and q, both local correlation and global distribution characteristics can be taken into account, adapting to different application scenarios.

[0050] 4. Effective anti-interference capability: This invention uses time-frequency analysis and interval sampling technology, which can effectively isolate the influence of noise interference on fault location. By using a decision function and observation threshold, only highly correlated vector pairs are counted, further improving the anti-noise performance.

[0051] 5. Comprehensive utilization of multi-band information: Through interval sampling and splicing technology in the frequency domain, comprehensive information from different time periods and frequency ranges is obtained, which helps to fully capture fault characteristics. The splicing of multiple third signals can further enhance the ability to describe complex fault characteristics.

[0052] 6. Adaptive optimization of observation threshold: The observation threshold is automatically calculated by the correlation vector degree and multifractal dimension, and can be adaptively adjusted to adapt to different system sizes and complexities;

[0053] 7. Wide range of applications: This invention is applicable to various transmission lines (such as overhead lines and cable lines) and different types of faults (such as single-phase grounding, phase-to-phase short circuit, open circuit, etc.), and also performs well in complex power networks (such as multi-branch networks).

[0054] 8. Improved fault diagnosis capability: This invention can better distinguish different fault types and locations through the correlation analysis between waveform features and actual distance, which helps to improve the accuracy and robustness of fault diagnosis. Feature extraction can identify significant patterns related to faults and avoid interference from too much irrelevant data.

[0055] 9. Multi-frequency range characteristic capture: Frequency domain range sampling and multi-frequency range signal splicing can capture the characteristics of traveling wave signals in different frequency bands, making fault location more comprehensive and accurate, especially suitable for complex transmission lines with significant multi-frequency characteristics.

[0056] In summary, this invention integrates time-frequency analysis, interval sampling, multi-band processing, and correlation vector analysis. Through a complete process design and flexible parameter adjustment, it has significant advantages in terms of accuracy, robustness, real-time performance, and applicability. It is suitable for the rapid location and feature extraction of various transmission line faults in power systems, providing important technical support for improving the operational stability and security of power systems. Attached Figure Description

[0057] Figure 1 is a schematic diagram of the process steps of a typical traveling wave characteristic parameter extraction method based on time-frequency analysis according to the present invention. Detailed Implementation

[0058] As shown in Figure 1, a method for extracting characteristic parameters of a typical traveling wave based on time-frequency analysis includes the following steps:

[0059] Step S1: Acquire traveling wave signals, obtain the acquisition point location information and fault point location information of the traveling wave signals, and use the traveling wave signals as the first signal;

[0060] Step S2: The first signal is sampled at intervals based on the time-domain sequence to generate a second signal, and the second signal is selected based on the frequency-domain interval to generate a third signal;

[0061] Step S3: Discretely sample the third signal and recombine them to form the fourth signal;

[0062] Step S4: Perform feature extraction on the fourth signal to obtain waveform features, and obtain the correlation between the waveform features and the acquisition point location information and the fault point location information.

[0063] Preferably, step S1 includes the following steps:

[0064] Step S11: Set up traveling wave signal acquisition points and fault simulation points on the transmission line, and set up multiple traveling wave signal acquisition points, and record the location information of each traveling wave signal acquisition point.

[0065] Step S12: Simulate various types of transmission line faults and record the simulated fault points as fault location information;

[0066] Step S13: After the fault traveling wave signal acquisition point detects the traveling wave signal, it takes the traveling wave signal as the first signal.

[0067] More preferably, the formula for calculating the location of the fault point is:

[0068] L f =L1+d f or L f =L2-d f ,

[0069] Where, d f T represents the distance from the fault point to the data collection point. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f The time it takes for the fault traveling wave signal to reach acquisition point 2 is represented by v, the traveling wave propagation speed is represented by v, L1 represents the location of acquisition point 1, and L2 represents the location of acquisition point 2. f Indicates the location of the fault.

[0070] Preferably, step S2 includes the following steps:

[0071] Step S21: Perform interval sampling on the first signal based on the time-domain sequence to generate a second signal;

[0072] Step S22: Perform interval sampling on the second signal based on the frequency domain interval to obtain sampling information;

[0073] Step S23: The sampled information obtained by interval sampling is spliced ​​together in time order to obtain the third signal;

[0074] Step S24: By changing the distance between frequency intervals, different frequency intervals are sampled at intervals to obtain sampling information of different frequency intervals and obtain multiple third signals.

[0075] More preferably, the expression for generating the second signal by performing interval sampling on the first signal based on the time-domain sequence is:

[0076] Where, r i (t) represents the signal received in the i-th time interval, i.e., the second signal, A represents the amplitude of the second signal, f0 represents the center frequency of the second signal, θ0 represents the initial phase of the second signal, j represents the imaginary unit, w(t) represents the noise part of the second signal, i represents the index of the time segment, i.e., the current signal is in the i-th time interval, t represents the time variable, and T0 represents the length of the time segment, i.e. the duration of each interval;

[0077] The expression for obtaining sampling information by interval sampling of the second signal based on frequency domain intervals is:

[0078] Where X(k) represents the sampling information, that is, the sampled values ​​of a finite number of spectra in the second signal, and x(n) represents the nth discrete sampled value of the original signal. The sampling interval is represented by N, the total number of sampling points of the signal is represented by k, the frequency index is represented by n, the time index is represented by i, and the imaginary unit is represented by i.

[0079] The expression for the third signal is:

[0080] x k,m (n)=x(t k,m +nT k,m ), n = 0, 1, 2, ..., N k，m -1,

[0081] x third,k (n), k = 1, 2, ..., K

[0082] Where x(t) represents the original signal, xk,m (n) represents the sampled signal obtained from the k-th frequency interval and the m-th time interval, t k,m T represents the starting time point of sampling. k,m N represents the sampling interval, k represents the frequency interval index, m represents the segment number when performing interval sampling on the k-th frequency interval, and N represents the frequency interval index. k,m Let x represent the length of the m-th sampled signal within the k-th frequency interval. third,k (n) represents the third signal, K represents the total number of frequency intervals, and N k,j This represents the number of sampling points in the j-th segment of the signal within the k-th frequency interval, where n represents the sampling point index of the discrete signal.

[0083] Preferably, step S3 specifically involves: discretizing each of the third signals and then recombining them to form a fourth signal, and obtaining the fourth signal group; the expression of the fourth signal is:

[0084] Where, r i (n) represents the received signal at the nth sampling point within the i-th time interval, i.e., the fourth signal; A represents the amplitude of the fourth signal; f0 represents the center frequency of the fourth signal; and Δt represents the sampling interval. The initial phase of the fourth signal is represented by w(n), the noise part of the fourth signal is represented by w(n), the index of the time segment is represented by i, the number of sampling points in each time segment is represented by N0, and the number of the signal sampling points is represented by n.

[0085] Preferably, step S4 includes the following steps:

[0086] Step S41: Establish the correlation vector degree model;

[0087] Step S42: Calculate the distance between the location information of each sampling point and the location information of the fault point in the sampling information. Given an observation threshold, when the relative distance is less than the observation threshold, it is called an associated vector pair. Then, count the associated vector pairs to obtain the relative statistical distance.

[0088] Step S43: Based on the actual distance between the acquisition point location information and the fault point location information, the waveform features are matched one-to-one with the actual distance between the acquisition point location information and the fault point location information, as well as the correlation interval, to form a correlation relationship. The traveling wave signal is identified based on the correlation relationship, and the fault point location is located.

[0089] More preferably, the expression for the correlation vector degree model is:

[0090] Where C(q,r) represents the correlation vector degree value, q represents the singularity measure, r represents the observation regime of the correlation integral or the time delay or observation threshold of the time series, N represents the total number of samples, x represents the global signal eigenvalue, a represents the offset, and n represents the amplification factor. i Let x represent the value of the i-th sample point in the sample set. j Let H(x) represent the value of the j-th sample point in the sample set, and let H(x) represent the step function.

[0091] More preferably, the relative distance in step S42 is the distance between the data collection point and the fault point, and the expression for the relative distance is: D ij =|L i -L f |,

[0092] Among them, D ij L represents the distance between data collection point i and fault point j. i L represents the location information of collection point i. f This indicates the location information of the fault point f;

[0093] When D ij When r < r, the collection point i and the fault point f form an association vector pair, and the counting formula is:

[0094] Where, N pair The number of associated vector pairs is represented by r, the observation threshold is represented by δ, and the decision function is represented by δ.

[0095] The formula for calculating the observation threshold r is: C(q,r)=r D ,

[0096] Where D represents the multifractal dimension and C(q,r) represents the correlation vector degree value.

[0097] More preferably, the expression for the association relationship in step S43 is: R(L) i ,L f )=f(D ij ,φ),

[0098] Among them, R(L i ,L f ) indicates the collection point L i and fault point L f The relationship between D ij The distance between the acquisition point and the fault point is represented by φ, the waveform characteristics are represented by α and β, the adjustment coefficients are represented by v, and the traveling wave propagation speed is represented by T. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f L represents the time it takes for the fault traveling wave signal to arrive at acquisition point 2.i L represents the location of a known data collection point. f Indicates the location of the fault.

[0099] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for extracting characteristic parameters of typical traveling waves based on time-frequency analysis, characterized in that, Includes the following steps: Step S1: Acquire traveling wave signals, obtain the acquisition point location information and fault point location information of the traveling wave signals, and use the traveling wave signals as the first signal; Step S2: The first signal is sampled at intervals based on the time domain sequence to generate a second signal, and the second signal is selected based on the frequency domain interval to generate a third signal; Step S3: Discretely sample the third signal and recombine them to form the fourth signal; Step S4: Perform feature extraction on the fourth signal to obtain waveform features, and obtain the correlation between the waveform features and the acquisition point location information and the fault point location information.

2. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Set up traveling wave signal acquisition points and fault simulation points on the transmission line, and set up multiple traveling wave signal acquisition points, and record the location information of each traveling wave signal acquisition point. Step S12: Simulate various types of transmission line faults and record the simulated fault points as fault location information; Step S13: After the fault traveling wave signal acquisition point detects the traveling wave signal, it takes the traveling wave signal as the first signal.

3. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 2, characterized in that, The formula for calculating the location of the fault point is: L f = L1 + d f or L f = L2 - d f , Where, d f T represents the distance from the fault point to the data collection point. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f The time it takes for the fault traveling wave signal to reach acquisition point 2 is represented by v, the traveling wave propagation speed is represented by v, L1 represents the location of acquisition point 1, and L2 represents the location of acquisition point 2. f Indicates the location of the fault.

4. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Perform interval sampling on the first signal based on the time-domain sequence to generate a second signal; Step S22: Perform interval sampling on the second signal based on the frequency domain interval to obtain sampling information; Step S23: The sampled information obtained by interval sampling is spliced ​​together in time order to obtain the third signal; Step S24: By changing the distance between frequency intervals, different frequency intervals are sampled at intervals to obtain sampling information of different frequency intervals and obtain multiple third signals.

5. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 4, characterized in that, The expression for generating the second signal by performing interval sampling on the first signal based on the time-domain sequence is as follows: Where, r i (t) represents the signal received in the i-th time interval, i.e., the second signal, A represents the amplitude of the second signal, f0 represents the center frequency of the second signal, θ0 represents the initial phase of the second signal, j represents the imaginary unit, w(t) represents the noise part of the second signal, i represents the index of the time segment, i.e., the current signal is in the i-th time interval, t represents the time variable, and T0 represents the length of the time segment, i.e. the duration of each interval; The expression for obtaining sampling information by interval sampling of the second signal based on frequency domain intervals is: Where X(k) represents the sampling information, that is, the sampled values ​​of a finite number of spectra in the second signal, and x(n) represents the nth discrete sampled value of the original signal. The sampling interval is represented by N, the total number of sampling points of the signal is represented by k, the frequency index is represented by n, the time index is represented by i, and the imaginary unit is represented by i. The expression for the third signal is: x k,m (n)=x(t k,m +nT k,m ),n=0,1,2,...,N k,m -1, x third,k (n),k=1,2,...,K, Where x(t) represents the original signal, x k,m (n) represents the sampled signal obtained from the k-th frequency interval and the m-th time interval, t k,m T represents the starting time point of sampling. k,m N represents the sampling interval, k represents the frequency interval index, m represents the segment number when performing interval sampling on the k-th frequency interval, and N represents the frequency interval index. k,m Let x represent the length of the m-th sampled signal within the k-th frequency interval. third,k (n) represents the third signal, K represents the total number of frequency intervals, and N k,j This represents the number of sampling points in the j-th segment of the signal within the k-th frequency interval, where n represents the sampling point index of the discrete signal.

6. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 1, characterized in that, Step S3 specifically involves: discretizing and recombining each of the third signals, then concatenating them to form a fourth signal, and obtaining the fourth signal group; the expression of the fourth signal is: Where, r i (n) represents the received signal at the nth sampling point within the i-th time interval, i.e., the fourth signal; A represents the amplitude of the fourth signal; f0 represents the center frequency of the fourth signal; and Δt represents the sampling interval. The initial phase of the fourth signal is represented by w(n), the noise part of the fourth signal is represented by w(n), the index of the time segment is represented by i, the number of sampling points in each time segment is represented by N0, and the number of the signal sampling points is represented by n.

7. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Establish the correlation vector degree model; Step S42: Calculate the distance between the location information of each sampling point and the location information of the fault point in the sampling information. Given an observation threshold, when the relative distance is less than the observation threshold, it is called an associated vector pair. Then, count the associated vector pairs to obtain the relative statistical distance. Step S43: Based on the actual distance between the acquisition point location information and the fault point location information, the waveform features are matched one-to-one with the actual distance between the acquisition point location information and the fault point location information, as well as the correlation interval, to form a correlation relationship. The traveling wave signal is identified based on the correlation relationship, and the fault point location is located.

8. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 7, characterized in that, The expression for the correlation vector degree model is: Where C(q,r) represents the correlation vector degree value, q represents the singularity measure, r represents the observation regime of the correlation integral or the time delay or observation threshold of the time series, N represents the total number of samples, x represents the global signal eigenvalue, a represents the offset, and n represents the amplification factor. i Let x represent the value of the i-th sample point in the sample set. j Let H(x) represent the value of the j-th sample point in the sample set, and let H(x) represent the step function.

9. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 7, characterized in that, The relative distance in step S42 is the distance between the data collection point and the fault point, and the expression for the relative distance is: D ij =|L i -L f |, Among them, D ij L represents the distance between data collection point i and fault point j. i L represents the location information of collection point i. f This indicates the location information of the fault point f; When D ij When r < r, the collection point i and the fault point f form an association vector pair, and the counting formula is: Where, N pair The number of associated vector pairs is represented by r, the observation threshold is represented by δ, and the decision function is represented by δ. The formula for calculating the observation threshold r is: C(q,r)=r D , Where D represents the multifractal dimension and C(q,r) represents the correlation vector degree value.

10. The method for extracting typical traveling wave characteristic parameters based on time-frequency analysis according to claim 7, characterized in that, The expression for the association relationship in step S43 is: R(L i ,L f )=f(D ij ,φ), Among them, R(L i ,L f ) indicates the collection point L i and fault point L f The relationship between D ij The distance between the acquisition point and the fault point is represented by φ, the waveform characteristics are represented by α and β, the adjustment coefficients are represented by v, and the traveling wave propagation speed is represented by T. 1f T represents the time it takes for the fault traveling wave signal to arrive at acquisition point 1. 2f L represents the time it takes for the fault traveling wave signal to arrive at acquisition point 2. i L represents the location of a known data collection point. f Indicates the location of the fault.

Images