Fault identification method and device based on transmission conductor, terminal equipment and storage medium

Abstract

The invention discloses a fault identification method and device based on a power transmission conductor, terminal equipment and a storage medium, and the method comprises the steps: determining a target filtering signal with a maximum second spectral kurtosis value through the first spectral kurtosis of a vibration signal of the power transmission conductor according to the first spectral kurtosis and a self-adaptive filter; the method comprises the following steps: extracting a target filtering signal from a self-adaptive filter, generating a to-be-optimized filter of the target filtering signal, carrying out iterative optimization on the self-adaptive filter capable of extracting the target filtering signal, taking the self-adaptive filter as a subsequent target filter for extracting a signal periodic fault feature, and finally filtering the vibration signal by adopting the target filter to generate the periodic fault feature. And according to the fault identification result of the transmission conductor based on the periodic fault features, the defects of end effect, false components and the like caused by the fact that the periodic fault features of the vibration signals cannot be extracted in the current EMD method are overcome, and the accuracy of transmission conductor fault identification is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a fault identification method, device, terminal equipment and storage medium based on a power transmission line. Background Art

[0002] As an important infrastructure in the power grid, conductors are directly related to the stable operation of the power grid and user experience. In the process of designing and selecting high-voltage line materials, designers need to have a deep understanding of the electrical and mechanical indicators of the wires, and evaluate the endurance and heat resistance of the wires. Because the wires work in a complex working environment for a long time, especially ultra-high voltage transmission lines, the wires exposed to the air will be affected by factors such as wind, ice, thunder, and the system's own impact, causing vibrations on the line, which may cause structural damage to the line in severe cases. The broken strands of the wires caused by line vibration will affect the transmission efficiency of the line. In severe cases, the wire breakage will cause local power outages. Therefore, the fault identification of transmission lines is of great significance to ensuring the safety of national power transmission.

[0003] Since the sensitivity of kurtosis to a single shock in the signal is much greater than the periodic shock components in the signal, the currently widely used empirical mode decomposition (Empirical Mode Decomposition) method can only deconvolute a few shock components for the periodic shock components in the vibration signal, resulting in problems such as endpoint effects and false components, which affects the accuracy of wire fault identification. Summary of the invention

[0004] The embodiments of the present invention provide a fault identification method, apparatus, terminal equipment and storage medium based on a transmission line, which overcome the defects of the current EMD method such as endpoint effect and false components caused by the inability to extract periodic fault characteristics of vibration signals, thereby effectively improving the accuracy of transmission line fault identification.

[0005] An embodiment of the present invention provides a method for fault identification based on a transmission line, comprising:

[0006] Acquire a vibration signal of a transmission line and a first spectral kurtosis corresponding to the vibration signal;

[0007] According to the first spectral kurtosis and a preset adaptive filter, the vibration signal is divided into a plurality of vibration sub-signals, and a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal are generated;

[0008] Determining a filtered signal corresponding to the maximum second spectral kurtosis value as a target filtered signal, and determining a filter that generates the target filtered signal as a filter to be optimized;

[0009] Iteratively optimize the filter parameters of the filter to be optimized, and in each iterative optimization process, use the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculate the relevant spectrum kurtosis of the new target filter signal, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, stop the iterative optimization operation and generate the target filter;

[0010] Using the target filter to filter the vibration signal to generate a periodic fault feature;

[0011] According to the periodic fault characteristics, a fault identification result of the transmission line is output.

[0012] Further, the obtaining of the vibration signal of the transmission line and the first spectral kurtosis corresponding to the vibration signal includes:

[0013] Acquiring initial vibration signals of transmission lines;

[0014] Filtering the initial vibration signal to generate a vibration signal after noise is removed;

[0015] Convert the vibration signal into a signal time series in a time-frequency form, decompose the signal time series, determine a non-stationary subsequence of the signal time series, and a 2n-order spectral moment of the signal time series;

[0016] According to the 2n-order spectral moment, time averaging is performed on the non-stationary subsequence along the time axis to generate a 2n-order spectral moment average value;

[0017] The fourth-order spectrum accumulation of the signal time series is calculated according to the 2n-order spectrum moment average value, and the fourth-order spectrum accumulation is energy normalized to generate the first spectrum kurtosis of the vibration signal.

[0018] Furthermore, the step of performing time averaging processing on the non-stationary subsequence along the time axis according to the 2n-order spectral moment to generate a 2n-order spectral moment average value includes:

[0019] The 2n-order spectral moment average value is calculated according to the following formula:

[0020]

[0021] Among them, S 2nY (f) is the average value of the 2n-order spectral moment, T is time, S 2nY (t,f) is the 2nth order spectral moment of the signal time series at the time-frequency (t,f).

[0022] Further, the vibration signal is divided into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal are generated, including:

[0023] Get the preset window function;

[0024] Performing Fourier transform on the vibration signal to generate a signal frequency domain sequence;

[0025] According to the window function and the first spectral kurtosis, the signal frequency domain sequence is windowed, the signal frequency domain sequence is divided into a plurality of windowed signal subsequences with the same width, and a subsequence kurtosis value of each windowed signal subsequence is determined;

[0026] According to the subsequence kurtosis value and the adaptive filter, a window fusion operation is performed on the windowed signal subsequence to generate a plurality of vibration sub-signals and a filter corresponding to each vibration sub-signal, and then a filtered signal corresponding to each vibration sub-signal and a second spectral kurtosis value corresponding to each filtered signal are generated.

[0027] Further, the window fusion operation is performed on the windowed signal subsequence according to the subsequence kurtosis value and the adaptive filter to generate a plurality of vibration subsignals and a filter corresponding to each vibration subsignal, and then a filtered signal corresponding to each vibration subsignal and a second spectral kurtosis value corresponding to each filtered signal is generated, including:

[0028] Repeating the window fusion operation according to the subsequence kurtosis value and the windowed signal subsequence, and stopping the window fusion operation when it is determined that each of the windowed signal subsequences has undergone the window fusion operation, to generate a plurality of vibration sub-signals;

[0029] Inputting the vibrator signals into the adaptive filter in sequence, so that the adaptive filter adaptively adjusts its own parameters to generate a filter corresponding to each vibrator signal;

[0030] According to the vibrator signals and the corresponding filters, generating a filter signal corresponding to each vibrator signal and a second spectral kurtosis value corresponding to each filter signal;

[0031] The window fusion operation includes:

[0032] Acquire a sequence to be fused and a first kurtosis value of the sequence to be fused; wherein, initially, the sequence to be fused is the first windowed signal subsequence on the signal frequency domain sequence;

[0033] Determine whether there is a sequence to be optimized;

[0034] If not, the next windowed signal subsequence of the sequence to be fused is used as the sequence to be evaluated, and the second kurtosis value of the sequence to be evaluated is obtained.

[0035] Fusing the sequence to be fused with the sequence to be evaluated to generate a first fused sequence, and calculating a third kurtosis value of the first fused sequence;

[0036] When it is determined that the third kurtosis value is greater than the first kurtosis value, and the third kurtosis value is greater than the second kurtosis value, taking the next windowed signal subsequence of the sequence to be evaluated as the sequence to be fused for the next round of window fusion operation, and taking the first fused sequence as the sequence to be optimized for the next round of window fusion operation;

[0037] When it is determined that the third kurtosis value is not greater than the first kurtosis value, or the third kurtosis value is not greater than the second kurtosis value, performing an inverse Fourier transform on the sequence to be fused to generate a oscillator signal, and using the sequence to be evaluated as the sequence to be fused for the next round of window fusion operation;

[0038] If yes, then obtain the fourth kurtosis value of the sequence to be optimized, fuse the sequence to be fused and the sequence to be optimized to generate a second fused sequence, and calculate the fifth kurtosis value of the second fused sequence;

[0039] When it is determined that the fourth kurtosis value is greater than the first kurtosis value, and the fourth kurtosis value is greater than the fifth kurtosis value, taking the next windowed signal subsequence of the sequence to be fused as the sequence to be fused for the next round of window fusion operation, and taking the second fused sequence as the sequence to be optimized for the next round of window fusion operation;

[0040] When it is determined that the fourth kurtosis value is not greater than the first kurtosis value, or the fourth kurtosis value is not greater than the fifth kurtosis value, the sequence to be optimized is extracted and inverse Fourier transform is performed to generate a oscillator signal, and the sequence to be fused is used as the sequence to be fused for the next round of window fusion operation.

[0041] Furthermore, the filter parameters include: deconvolution period, filter length, and displacement number;

[0042] The iterative optimization operation is performed on the filter parameters of the filter to be optimized, and in each iterative optimization operation, the optimized filter to be optimized is used to filter the vibrator signal corresponding to the target filter signal, and the relevant spectrum kurtosis of the new target filter signal is calculated, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, the iterative optimization operation is stopped to generate the target filter, including:

[0043] Acquiring the sampling frequency of the vibration signal and the initial displacement value of the filter to be optimized;

[0044] Determining an initial deconvolution period of the filter to be optimized according to the sampling frequency of the vibration signal, and generating a plurality of deconvolution periods to be evaluated according to the initial deconvolution period;

[0045] According to the initial displacement value and a number of deconvolution cycles to be evaluated, repeatedly performing an iterative optimization operation on the filter to be optimized until a target filter is generated;

[0046] The iterative optimization operation includes:

[0047] Obtaining a displacement value to be evaluated, a deconvolution period to be evaluated, and a filter to be optimized; wherein, initially, the displacement value to be evaluated is an initial displacement value;

[0048] According to a number of deconvolution cycles to be evaluated and displacement values ​​to be evaluated, the filter to be optimized is updated to generate a filter to be evaluated;

[0049] The vibrator signal corresponding to the target filter signal is filtered by using the filter to be evaluated to obtain the filter signals to be evaluated generated by the filter to be evaluated with different deconvolution periods to be evaluated under the displacement value to be evaluated, and the relevant spectral kurtosis corresponding to each filter signal to be evaluated;

[0050] Sorting the relevant spectral kurtosis, and multiplying the second largest relevant spectral kurtosis by a preset multiple to generate a judgment threshold;

[0051] Determine whether the maximum correlation spectrum kurtosis is greater than the determination threshold;

[0052] If not, the displacement value to be evaluated is increased by 1 to generate the displacement value to be evaluated required for the next round of iterative optimization operation;

[0053] If yes, the deconvolution period to be evaluated corresponding to the maximum correlation spectrum kurtosis is taken as the target deconvolution period, and the target filter length is calculated according to the filter signal to be evaluated corresponding to the maximum correlation spectrum kurtosis, and then the target displacement number is determined according to the target deconvolution period and the target filter length;

[0054] According to the target deconvolution period, the target filter length, and the target displacement number, the filter parameters of the filter to be optimized are updated to generate a target filter.

[0055] Further, determining the initial deconvolution period of the filter to be optimized according to the sampling frequency of the vibration signal, and generating a plurality of deconvolution periods to be evaluated according to the initial deconvolution period, comprises:

[0056] Get the preset frequency range and the preset period value range;

[0057] According to the frequency range and the sampling frequency, the initial deconvolution period is calculated according to the following formula:

[0058]

[0059] Among them, T f is the initial deconvolution period, f s is the sampling frequency, f fault is the frequency range;

[0060] According to the cycle value range and the initial deconvolution cycle, a plurality of deconvolution cycles to be evaluated are generated.

[0061] Another embodiment of the present invention provides a fault identification device based on a power transmission line, comprising:

[0062] A data acquisition module, used to acquire a vibration signal of the transmission line and a first spectral kurtosis corresponding to the vibration signal;

[0063] an adaptive filtering module, configured to divide the vibration signal into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and generate a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal;

[0064] A filter determination module, used to determine the filter signal corresponding to the maximum second spectral kurtosis value as the target filter signal, and determine the filter that generates the target filter signal as the filter to be optimized;

[0065] A filter optimization module, used to iteratively optimize the filter parameters of the filter to be optimized, and in each iterative optimization process, use the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculate the relevant spectrum kurtosis of the new target filter signal, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, stop the iterative optimization operation and generate the target filter;

[0066] A feature generation module, used to filter the vibration signal using the target filter to generate periodic fault features;

[0067] A fault identification module is used to output a fault identification result of the transmission line according to the periodic fault characteristics.

[0068] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a fault identification method based on a transmission line as described in any one of the embodiments is implemented.

[0069] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a fault identification method based on a transmission line as described in any of the above embodiments.

[0070] The following beneficial effects are achieved by implementing the present invention:

[0071] The present invention discloses a fault identification method, device, terminal equipment and storage medium based on a transmission line. The first spectral kurtosis of the vibration signal of the transmission line is used, and then according to the first spectral kurtosis and an adaptive filter, a target filter signal with the largest second spectral kurtosis value is determined, and a filter to be optimized for generating the target filter signal is determined. It can be understood that the spectral kurtosis value is very sensitive to the transient impact component in the signal. The filter signal most likely to contain fault information is determined by comparing the second spectral kurtosis value of each filter signal, and then the adaptive filter that can extract the target filter signal is iteratively optimized as a target filter for subsequently extracting the periodic fault characteristics of the signal. It can be understood that the relevant spectral kurtosis is an important indicator for measuring the periodic impact characteristics of the signal. By determining in the iterative optimization process that the correlation spectrum kurtosis is greater than the judgment threshold, it can be determined that the filter to be optimized can completely extract the periodic impact characteristics of the signal, and the filter to be optimized at this time is used as the target filter, and finally the target filter is used to filter the vibration signal to generate periodic fault characteristics, and the fault identification result of the transmission line is based on the periodic fault characteristics. Therefore, the present invention overcomes the defects of the current EMD method such as endpoint effect and false components caused by the inability to extract the periodic fault characteristics of the vibration signal, thereby effectively improving the accuracy of transmission line fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a flow chart of a fault identification method based on a transmission line provided by an embodiment of the present invention.

[0073] Figure 2 It is a structural schematic diagram of a fault identification device based on a transmission line provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0075] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0076] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0077] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0078] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0079] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0080] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the internal connection of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0081] See also Figure 1 , is a flow chart of a fault identification method based on a transmission line provided by an embodiment of the present invention, comprising:

[0082] S1. Acquire a vibration signal of a transmission line and a first spectral kurtosis corresponding to the vibration signal;

[0083] Preferably, the step of obtaining a vibration signal of the transmission line and a first spectral kurtosis corresponding to the vibration signal comprises:

[0084] S11, obtaining an initial vibration signal of the transmission line;

[0085] S12, filtering the initial vibration signal to generate a vibration signal after noise is removed;

[0086] S13, converting the vibration signal into a signal time series in a time-frequency form, and decomposing the signal time series to determine a non-stationary subsequence of the signal time series and a 2n-order spectral moment of the signal time series;

[0087] S14, performing time averaging processing on the non-stationary subsequence along the time axis according to the 2n-order spectral moment to generate a 2n-order spectral moment average value;

[0088] Preferably, performing time averaging processing on the non-stationary subsequence along the time axis according to the 2n-order spectral moment to generate the 2n-order spectral moment average value includes:

[0089] S141. Calculate the 2n-order spectral moment average value according to the following formula:

[0090]

[0091] Among them, S 2nY (f) is the average value of the 2n-order spectral moment, T is time, S 2nY (t,f) is the 2nth order spectral moment of the signal time series at the time-frequency (t,f).

[0092] S15. Calculate the 4th order spectrum cumulant of the signal time series according to the 2nth order spectrum moment average value, and perform energy normalization on the 4th order spectrum cumulant to generate a first spectral kurtosis of the vibration signal.

[0093] In a preferred embodiment of the present invention, the kurtosis is defined as the normalized 4th-order center distance. In the fault diagnosis of power transmission lines, the kurtosis reflects the degree to which the vibration signal deviates from the Gaussian process. It is defined as:

[0094]

[0095] Where K is the kurtosis, σ is the standard deviation of the signal x(t), is the mean of the signal x(t), and N is the signal length.

[0096] Spectral kurtosis is defined as the energy-normalized fourth-order spectral cumulant, which is essentially a measure of the average time dispersion of the signal's time-frequency energy distribution over all frequencies. For a time-varying system h(t,s), s is a time imaginary variable, and the signal y(t) is the system response to the input x(t), then the Wold-Cramer decomposition of y(t) is:

[0097]

[0098] Where H(t,f) is the Fourier transform of h(t,s), which can be expressed as the complex envelope of signal y(t) at frequency f. X(f) is the frequency domain representation of signal x(t), and dX(f) is the spectral process of x(t).

[0099] From the Wold-Cramer decomposition of the signal, we can see that the complex envelope H(t,f) describes the characteristic information of the non-stationary process y(t), and the information contained in H(t,f) is measured by the statistic-spectral moment. The 2n-order spectral moment of the signal is expressed as:

[0100] s 2nY (t,f)=E<|H(t,f)| 2n >*S 2nx ;

[0101] where s 2nY (t,f) is the 2nth order spectral moment of the signal, E<> is the mean operation, S 2nx is the input signal. Since the complex envelope H(t,f) is time-stationary and independent of the spectral process dX(f), s 2nY (t,f) is a measure of the information contained in the complex envelope H(t,f), which measures the energy intensity of the signal at the time-frequency (t,f).

[0102] For non-stationary processes, the spectral moment is a statistical indicator that characterizes its characteristic information. In order to accurately obtain the spectral moment value of the signal in practice, it is necessary to perform time averaging processing on the instantaneous moment along the time axis, that is:

[0103]

[0104] Where S 2nY (f) is the average value of the 2n-order spectral moment, and T is the time.

[0105] Since the value of the fourth-order spectrum cumulant increases as the signal deviates from the Gaussian process, it is necessary to normalize the energy of the fourth-order spectrum cumulant, that is:

[0106]

[0107] Where K Y (f) is the spectral kurtosis, which describes the average time dispersion of the signal at frequency f, s4Y (f) is the spectral moment of the 4th order spectrum, c 4Y (f) is the cumulative amount of the fourth-order spectrum, which is more practical than the kurtosis index and can accurately determine the position of the non-stationary signal in the frequency domain.

[0108] It should be noted that the transmission line vibration signal is ASK filtered and the kurtosis is applied to the bearing vibration signal analysis. Since the kurtosis is independent of factors such as the conductor size and is only sensitive to the impact component in the signal, it is particularly suitable for conductor surface damage or pit faults. This characteristic of kurtosis is suitable as a time domain parameter for early transmission line vibration signal judgment to determine whether a fault has occurred.

[0109] S2. According to the first spectral kurtosis and the preset adaptive filter, the vibration signal is divided into a plurality of vibration sub-signals, and a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal is generated.

[0110] Preferably, the method of dividing the vibration signal into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and generating a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal, comprises:

[0111] S21, obtaining a preset window function;

[0112] S22, performing Fourier transform on the vibration signal to generate a signal frequency domain sequence;

[0113] S23, according to the window function and the first spectral kurtosis, windowing the signal frequency domain sequence, dividing the signal frequency domain sequence into a plurality of windowed signal subsequences with the same width, and determining a subsequence kurtosis value of each windowed signal subsequence;

[0114] S24. Perform a window fusion operation on the windowed signal subsequence according to the subsequence kurtosis value and the adaptive filter to generate a plurality of vibration sub-signals and a filter corresponding to each vibration sub-signal, and then generate a filtered signal corresponding to each vibration sub-signal and a second spectral kurtosis value corresponding to each filtered signal.

[0115] Preferably, the window fusion operation is performed on the windowed signal subsequence according to the subsequence kurtosis value and the adaptive filter to generate a plurality of vibrator subsignals and a filter corresponding to each vibrator subsignal, and then a filtered signal corresponding to each vibrator subsignal and a second spectral kurtosis value corresponding to each filtered signal, including:

[0116] S241, repeatedly performing a window fusion operation according to the subsequence kurtosis value and the windowed signal subsequence, and when it is determined that each of the windowed signal subsequences has undergone the window fusion operation, stopping the window fusion operation to generate a plurality of vibration sub-signals;

[0117] S242, sequentially inputting the vibrator signals into the adaptive filter, so that the adaptive filter adaptively adjusts its own parameters to generate a filter corresponding to each vibrator signal;

[0118] S243. Generate, according to the vibrator signals and the corresponding filters, a filtered signal corresponding to each vibrator signal and a second spectral kurtosis value corresponding to each filtered signal;

[0119] The window fusion operation includes:

[0120] S2411, obtaining a sequence to be fused and a first kurtosis value of the sequence to be fused; wherein, initially, the sequence to be fused is the first windowed signal subsequence on the signal frequency domain sequence;

[0121] S2412, determining whether there is a sequence to be optimized;

[0122] S2413: If not, taking the next windowed signal subsequence of the sequence to be fused as the sequence to be evaluated, and obtaining a second kurtosis value of the sequence to be evaluated.

[0123] S2414, fusing the sequence to be fused with the sequence to be evaluated to generate a first fused sequence, and calculating a third kurtosis value of the first fused sequence;

[0124] S2415. When it is determined that the third kurtosis value is greater than the first kurtosis value, and the third kurtosis value is greater than the second kurtosis value, taking the next windowed signal subsequence of the sequence to be evaluated as the sequence to be fused for the next round of window fusion operation, and taking the first fused sequence as the sequence to be optimized for the next round of window fusion operation;

[0125] S2416. When it is determined that the third kurtosis value is not greater than the first kurtosis value, or the third kurtosis value is not greater than the second kurtosis value, inverse Fourier transform is performed on the sequence to be fused to generate a oscillator signal, and the sequence to be evaluated is used as the sequence to be fused for the next round of window fusion operation;

[0126] S2417: If yes, obtain the fourth kurtosis value of the sequence to be optimized, fuse the sequence to be fused and the sequence to be optimized to generate a second fused sequence, and calculate the fifth kurtosis value of the second fused sequence;

[0127] S2418. When it is determined that the fourth kurtosis value is greater than the first kurtosis value, and the fourth kurtosis value is greater than the fifth kurtosis value, using the next windowed signal subsequence of the sequence to be fused as the sequence to be fused for the next round of window fusion operation, and using the second fused sequence as the sequence to be optimized for the next round of window fusion operation;

[0128] S2419. When it is determined that the fourth kurtosis value is not greater than the first kurtosis value, or the fourth kurtosis value is not greater than the fifth kurtosis value, extract the sequence to be optimized and perform inverse Fourier transform to generate a oscillator signal, and use the sequence to be fused as the sequence to be fused for the next round of window fusion operation.

[0129] In a preferred embodiment of the present invention, the adaptive spectral kurtosis method determines the center frequency and bandwidth through a greedy algorithm, and fuses adjacent window functions on the frequency axis according to the principle of maximum kurtosis, thereby traversing the window functions on the entire frequency axis, and finally determines the center frequency and bandwidth of the optimal filter based on the final spectral kurtosis diagram.

[0130] First, the adaptive spectral kurtosis method is to window the signal in the frequency domain, so the signal is first transformed into the frequency domain through Fourier transform;

[0131]

[0132] in, is the Fourier transform sequence of the signal, N is the signal length, k is the current time domain, x[k] is the current time domain signal, and n is the order of the spectrum.

[0133] The frequency domain sequence is windowed by the current window function and the next adjacent window function:

[0134]

[0135] in Add the window sequence to the current window. is the sequence after adding window to adjacent window functions, τ and l are the coefficients of the current window and the adjacent window respectively, n is the order of the spectrum, Add window function to the current window, T La is the windowing function for adjacent windows, and a is the interval between adjacent windows.

[0136] The signal frequency domain sequence is fused by the above two window functions to form a window function, which is expressed as follows after windowing:

[0137]

[0138] in is the sequence after window fusion, is the window fusion function, w[] is the basic window function, and i is the index value of the final window.

[0139] Assume that the initial width of the window function is M, and the range of the sequence on the frequency axis after windowing is [ria, (r + ri) a + M / 2]. Similarly, the range of the sequence on the frequency axis is [(r + ri) a, (r + ri) a + M], and the range of the sequence on the frequency axis is [ria, la + M / 2]. The filtered signal can be obtained by inverse Fourier transform:

[0140]

[0141] Among them, w ξ Represents w r , w Tl and w l G(λ,r) is the amplitude coefficient of the superimposed window function filter, and λ is the window overlap rate. The window overlap rate is related to the window length M and the interval a between adjacent windows.

[0142] Calculate the kurtosis value after bandpass filtering with a window function;

[0143] Definition of coincidence rate:

[0144]

[0145] Where λ is the window overlap rate, M is the window length, and a is the interval between adjacent windows.

[0146] When the window function is a hanning window:

[0147]

[0148] Where G(λ,r) is the amplitude coefficient.

[0149] Therefore, the kurtosis value after bandpass filtering by the window function is:

[0150]

[0151] where w ξ Represents w r , w Tl and w l <> indicates the mean operation, kri is the kurtosis value of the ri-th adaptive window function, and the constant "-2" is due to is a complex number. The kurtosis value in the above formula comes from the signal The signal It comes from the signal frequency domain window filtering, that is, the kurtosis value is related to the position of the window function in the frequency domain.

[0152] In the adaptive spectral kurtosis method, the fusion operation of the window function depends on the kurtosis value of the signal after windowing. Window fusion is performed only when the kurtosis value after fusion is greater than or equal to the maximum kurtosis value of the two window functions before fusion.

[0153]

[0154] Among them, w r , w Tl and w l They represent the current window, the next adjacent window, and the window after the two windows are fused. The above process is repeated until all windows have been tried to be fused.

[0155] S3, determining the filtered signal corresponding to the largest second spectral kurtosis value as the target filtered signal, and determining the adaptive filter that generates the target filtered signal as the filter to be optimized;

[0156] S4, performing an iterative optimization operation on the filter parameters of the filter to be optimized, and in each iterative optimization operation, using the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculating the relevant spectral kurtosis of the new target filter signal, and when it is determined that the relevant spectral kurtosis is greater than a preset judgment threshold, stopping the iterative optimization operation and generating a target filter;

[0157] Preferably, the filter parameters include: deconvolution period, filter length, and displacement number;

[0158] The iterative optimization operation is performed on the filter parameters of the filter to be optimized, and in each iterative optimization operation, the optimized filter to be optimized is used to filter the vibrator signal corresponding to the target filter signal, and the relevant spectrum kurtosis of the new target filter signal is calculated, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, the iterative optimization operation is stopped to generate the target filter, including:

[0159] S41, acquiring the sampling frequency of the vibration signal and the initial displacement value of the filter to be optimized;

[0160] S42, repeatedly performing an iterative optimization operation on the filter to be optimized according to the initial displacement value and a number of deconvolution cycles to be evaluated, until a target filter is generated;

[0161] The iterative optimization operation includes:

[0162] S421, obtaining a displacement value to be evaluated, a deconvolution period to be evaluated, and a filter to be optimized; wherein, initially, the displacement value to be evaluated is an initial displacement value;

[0163] S422, updating the filter to be optimized according to a number of deconvolution cycles to be evaluated and displacement values ​​to be evaluated, and generating a filter to be evaluated;

[0164] S423, filtering the vibrator signal corresponding to the target filter signal using the filter to be evaluated, to obtain the filter signals to be evaluated generated by the filters to be evaluated with different deconvolution periods to be evaluated under the displacement value to be evaluated, and the relevant spectral kurtosis corresponding to each filter signal to be evaluated;

[0165] S424, sorting the relevant spectral kurtosis, and multiplying the second largest relevant spectral kurtosis by a preset multiple to generate a judgment threshold;

[0166] S425, judging whether the maximum correlation spectrum kurtosis is greater than the judgment threshold;

[0167] S426, if not, adding 1 to the displacement value to be evaluated to generate the displacement value to be evaluated required for the next round of iterative optimization operation;

[0168] S427, if yes, then the deconvolution period to be evaluated corresponding to the maximum correlation spectrum kurtosis is used as the target deconvolution period, and the target filter length is calculated according to the filtered signal to be evaluated corresponding to the maximum correlation spectrum kurtosis, and then the target displacement number is determined according to the target deconvolution period and the target filter length;

[0169] S428. Update the filter parameters of the filter to be optimized according to the target deconvolution period, the target filter length, and the target displacement number to generate a target filter.

[0170] In a preferred embodiment of the present invention, MCKD parameter selection analysis and fault diagnosis are adopted; since the effect of maximum correlation kurtosis deconvolution algorithm in extracting fault features is related to its parameters, before using maximum correlation kurtosis to extract transmission line vibration signal features, the MCKD algorithm parameters must first be analyzed and determined.

[0171] First, calculate the range of the deconvolution period; the MCKD algorithm is affected by parameters such as the filter length L, deconvolution period T, displacement number M, and iteration number N. The deconvolution period T is determined by the characteristic frequency of the transmission line fault and the sampling frequency. Since the fault characteristic frequency fluctuates within a range according to theoretical calculations, the range of the deconvolution period can be calculated as follows:

[0172]

[0173] Where T f To calculate the value of the unwinding period, f s is the sampling frequency of the vibration signal, f fault is the theoretically calculated fault characteristic frequency range.

[0174] According to the above formula, T f ,Pick:

[0175] T fc =[[min(T f )]-5,[min(T f )]-5];

[0176] Where T fc T f range, take T fc The relevant kurtosis of the signal is calculated for each value in, and the unwrapping period T is determined according to the maximum relevant kurtosis value.

[0177] Secondly, determine the optimal deconvolution period T; in practical applications, due to the influence of various noises, the above calculation of the correlation kurtosis under different periods cannot obtain a unique and obvious maximum value, which makes it impossible to accurately determine the deconvolution period T. According to the calculation formula of the correlation kurtosis, the correlation kurtosis value is related to the value of M. The larger the value of M, the more accurate the calculation of the correlation kurtosis, and the more accurate the deconvolution period T parameter determined by the maximum correlation kurtosis, but the larger the value of M, the longer the calculation time. In order to accurately determine the optimal deconvolution period T in the shortest time, specifically:

[0178] 1. M is set to 1, and T is determined according to the characteristic frequency calculation formula fc ;

[0179] 2. Calculate T fc Each value in corresponds to the M displacement number related kurtosis value ckM of the signal, and the ckM sequence is arranged from large to small to obtain the ck_sort sequence;

[0180] 3. If the first value in the ck_sort sequence is greater than three times the second value, the optimal deconvolution period T is determined based on the location of the maximum correlation kurtosis value in ckM, and the algorithm ends. Otherwise, M = M + 1, and go to the previous step.

[0181] Regarding the filter length L parameter, according to relevant theories, the shorter the filter length, the worse the filtering effect; however, since the shorter the filter length, the faster its calculation speed, the selection of the filter length L should increase the filter length as much as possible while ensuring the calculation speed as much as possible to ensure the accuracy and speed of the algorithm.

[0182] For the displacement number M, the larger the value of M, the longer the algorithm calculation time; in the vibration signal analysis of the transmission line, according to the determined filter length L and deconvolution period T, different M values ​​are taken to extract the vibration signal MCKD algorithm respectively, and the displacement number related kurtosis value ck1 is calculated for the extracted signal, and the optimal parameter M is determined according to the maximum value of ck1.

[0183] S5. Filtering the vibration signal using the target filter to generate a periodic fault feature;

[0184] In a preferred embodiment of the present invention, since the sensitivity of kurtosis to a single shock in a signal is much greater than the periodic shock component in the signal, the MED method can only deconvolute a few shock components for the periodic shock component in the vibration signal. For the vibration signal of the transmission line, once the fault point exists, it will continue, resulting in periodic shock vibration. In order to solve this problem, the maximum correlation kurtosis deconvolution method is proposed to extract the periodic shock of the signal through the correlation kurtosis.

[0185] First, determine the deconvolution period T, the number of shifts M and the filter length L, and initialize the filter f; extract the periodic impact of the signal through the correlation kurtosis, which is defined as follows:

[0186]

[0187] Where M is the number of shifts, T is the deconvolution period, and N is the length of the input signal. When T = 0 and M = 1, the correlation kurtosis degenerates into kurtosis.

[0188] The maximum correlation kurtosis deconvolution takes the maximization of correlation kurtosis as the optimization principle. Similar to the MED algorithm, it also extracts the fault impact component from the vibration signal obtained by the sensor by iteratively determining the optimal filter f. The objective function of the MCKD algorithm is defined as:

[0189]

[0190] Where CK is the maximum correlation kurtosis, and f is the filter coefficient corresponding to CK.

[0191] Secondly, calculate the output y of the signal x after the filter f is filtered; in order to solve the filter coefficient f corresponding to the maximum correlation kurtosis CK, the following conditions must be met:

[0192]

[0193] Among them, x n-k+1 is the signal to be processed, and yn is the signal after x is filtered by filter f.

[0194] Then, the change of the relevant kurtosis before and after deconvolution ΔCKM(T) is calculated to determine whether it is less than a given threshold, or whether the number of iterations reaches a specified maximum value. If so, the iteration is stopped, otherwise, the execution is returned to the previous step to continue;

[0195] Calculate the first-order derivatives of the numerator and denominator of CK with respect to fk respectively:

[0196]

[0197] Where M is the number of displacements, N is the signal length, xn-k+1 is the signal before processing, and yn is the signal after xn-k+1 is filtered by filter f. is the matrix corresponding to yn.

[0198] Convert to matrix form:

[0199]

[0200] in:

[0201]

[0202] Where xr is the matrix corresponding to the signal before processing, x0 is the initial signal parameter, and xmT is each element in xr. is the matrix corresponding to yn, r represents different deconvolution cycles, and are two matrices related to yn respectively.

[0203] Finally, the signal y is extracted using the correlation spectral kurtosis convolution algorithm;

[0204]

[0205] Also because Therefore, the above formula can be transformed into:

[0206]

[0207] S6. Outputting the fault identification result of the transmission line according to the periodic fault characteristics.

[0208] The present embodiment provides a fault identification method based on a transmission line, which determines a target filter signal with the largest second spectral kurtosis value and a filter to be optimized for generating the target filter signal through the first spectral kurtosis of the vibration signal of the transmission line, and then according to the first spectral kurtosis and an adaptive filter. It can be understood that the spectral kurtosis value is very sensitive to transient impact components in the signal. The filter signal that is most likely to contain fault information is determined by comparing the second spectral kurtosis value of each filter signal, and then the adaptive filter that can extract the target filter signal is iteratively optimized as the target filter for extracting the periodic fault characteristics of the signal. It can be understood that the relevant spectral kurtosis is an important indicator for measuring the periodic impact characteristics of the signal. By determining in the iterative optimization process that the correlation spectrum kurtosis is greater than the judgment threshold, it can be determined that the filter to be optimized can completely extract the periodic impact characteristics of the signal, and the filter to be optimized at this time is used as the target filter, and finally the target filter is used to filter the vibration signal to generate periodic fault characteristics, and the fault identification result of the transmission line is based on the periodic fault characteristics. Therefore, this embodiment overcomes the defects of the current EMD method such as endpoint effect and false components caused by the inability to extract the periodic fault characteristics of the vibration signal, thereby effectively improving the accuracy of transmission line fault identification.

[0209] See also Figure 2 , is a schematic diagram of a structure of a fault identification device based on a power transmission line provided by an embodiment of the present invention, comprising:

[0210] A data acquisition module, used to acquire a vibration signal of the transmission line and a first spectral kurtosis corresponding to the vibration signal;

[0211] an adaptive filtering module, configured to divide the vibration signal into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and generate a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal;

[0212] A filter determination module, used to determine the filter signal corresponding to the maximum second spectral kurtosis value as the target filter signal, and determine the filter that generates the target filter signal as the filter to be optimized;

[0213] A filter optimization module, used to iteratively optimize the filter parameters of the filter to be optimized, and in each iterative optimization process, use the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculate the relevant spectrum kurtosis of the new target filter signal, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, stop the iterative optimization operation and generate the target filter;

[0214] A feature generation module, used to filter the vibration signal using the target filter to generate periodic fault features;

[0215] A fault identification module is used to output a fault identification result of the transmission line according to the periodic fault characteristics.

[0216] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0217] Those skilled in the art can clearly understand that for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0218] Another preferred embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a fault identification method based on a transmission line as described in any one of the above embodiments is implemented.

[0219] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0220] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0221] The memory can be used to store the computer program, and the processor realizes various functions of the terminal device by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart EMDia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0222] Another preferred embodiment of the present invention provides a storage medium, the storage medium is a computer-readable storage medium, the computer program is stored in the computer-readable storage medium, and the computer program, when executed by the processor, can implement the steps of the above-mentioned various method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0223] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A fault identification method based on a transmission line, characterized in that: include: Acquire a vibration signal of a transmission line and a first spectral kurtosis corresponding to the vibration signal; According to the first spectral kurtosis and a preset adaptive filter, the vibration signal is divided into a plurality of vibration sub-signals, and a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal are generated; Determining a filtered signal corresponding to the maximum second spectral kurtosis value as a target filtered signal, and determining a filter that generates the target filtered signal as a filter to be optimized; Iteratively optimize the filter parameters of the filter to be optimized, and in each iterative optimization process, use the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculate the relevant spectrum kurtosis of the new target filter signal, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, stop the iterative optimization operation and generate the target filter; Using the target filter to filter the vibration signal to generate a periodic fault feature; According to the periodic fault characteristics, a fault identification result of the transmission line is output.

2. A method for identifying a fault based on a transmission line according to claim 1, characterized in that: The step of obtaining a vibration signal of the transmission line and a first spectral kurtosis corresponding to the vibration signal includes: Acquiring initial vibration signals of transmission lines; Filtering the initial vibration signal to generate a vibration signal after noise is removed; Convert the vibration signal into a signal time series in a time-frequency form, decompose the signal time series, determine a non-stationary subsequence of the signal time series, and a 2n-order spectral moment of the signal time series; According to the 2n-order spectral moment, time averaging is performed on the non-stationary subsequence along the time axis to generate a 2n-order spectral moment average value; The fourth-order spectrum accumulation of the signal time series is calculated according to the 2n-order spectrum moment average value, and the fourth-order spectrum accumulation is energy normalized to generate the first spectrum kurtosis of the vibration signal.

3. A method for identifying a fault based on a transmission line according to claim 2, characterized in that: The step of performing time averaging processing on the non-stationary subsequence along the time axis according to the 2n-order spectral moment to generate a 2n-order spectral moment average value includes: The 2n-order spectral moment average value is calculated according to the following formula: Among them, S 2nY (f) is the average value of the 2n-order spectral moment, T is time, S 2nY (t,f) is the 2nth order spectral moment of the signal time series at the time-frequency (t,f).

4. A method for identifying a fault based on a transmission line according to claim 3, characterized in that: The method of dividing the vibration signal into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and generating a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal, comprises: Get the preset window function; Performing Fourier transform on the vibration signal to generate a signal frequency domain sequence; According to the window function and the first spectral kurtosis, the signal frequency domain sequence is windowed, the signal frequency domain sequence is divided into a plurality of windowed signal subsequences with the same width, and a subsequence kurtosis value of each windowed signal subsequence is determined; According to the subsequence kurtosis value and the adaptive filter, a window fusion operation is performed on the windowed signal subsequence to generate a plurality of vibration sub-signals and a filter corresponding to each vibration sub-signal, and then a filtered signal corresponding to each vibration sub-signal and a second spectral kurtosis value corresponding to each filtered signal are generated.

5. A method for identifying a fault based on a transmission line according to claim 4, characterized in that: The method of performing a window fusion operation on the windowed signal subsequence according to the subsequence kurtosis value and the adaptive filter to generate a plurality of vibrating subsignals and a filter corresponding to each vibrating subsignal, and then generating a filtered signal corresponding to each vibrating subsignal and a second spectral kurtosis value corresponding to each filtered signal, comprises: Repeating the window fusion operation according to the subsequence kurtosis value and the windowed signal subsequence, and stopping the window fusion operation when it is determined that each of the windowed signal subsequences has undergone the window fusion operation, to generate a plurality of vibration sub-signals; Inputting the vibrator signals into the adaptive filter in sequence, so that the adaptive filter adaptively adjusts its own parameters to generate a filter corresponding to each vibrator signal; According to the vibrator signals and the corresponding filters, generating a filter signal corresponding to each vibrator signal and a second spectral kurtosis value corresponding to each filter signal; The window fusion operation includes: Acquire a sequence to be fused and a first kurtosis value of the sequence to be fused; wherein, initially, the sequence to be fused is the first windowed signal subsequence on the signal frequency domain sequence; Determine whether there is a sequence to be optimized; If not, the next windowed signal subsequence of the sequence to be fused is used as the sequence to be evaluated, and the second kurtosis value of the sequence to be evaluated is obtained; Fusing the sequence to be fused with the sequence to be evaluated to generate a first fused sequence, and calculating a third kurtosis value of the first fused sequence; When it is determined that the third kurtosis value is greater than the first kurtosis value, and the third kurtosis value is greater than the second kurtosis value, taking the next windowed signal subsequence of the sequence to be evaluated as the sequence to be fused for the next round of window fusion operation, and taking the first fused sequence as the sequence to be optimized for the next round of window fusion operation; When it is determined that the third kurtosis value is not greater than the first kurtosis value, or the third kurtosis value is not greater than the second kurtosis value, performing an inverse Fourier transform on the sequence to be fused to generate a oscillator signal, and using the sequence to be evaluated as the sequence to be fused for the next round of window fusion operation; If yes, then obtain the fourth kurtosis value of the sequence to be optimized, fuse the sequence to be fused and the sequence to be optimized to generate a second fused sequence, and calculate the fifth kurtosis value of the second fused sequence; When it is determined that the fourth kurtosis value is greater than the first kurtosis value, and the fourth kurtosis value is greater than the fifth kurtosis value, taking the next windowed signal subsequence of the sequence to be fused as the sequence to be fused for the next round of window fusion operation, and taking the second fused sequence as the sequence to be optimized for the next round of window fusion operation; When it is determined that the fourth kurtosis value is not greater than the first kurtosis value, or the fourth kurtosis value is not greater than the fifth kurtosis value, the sequence to be optimized is extracted and inverse Fourier transform is performed to generate a oscillator signal, and the sequence to be fused is used as the sequence to be fused for the next round of window fusion operation.

6. A method for identifying a fault based on a transmission line according to claim 5, characterized in that: The filter parameters include: deconvolution period, filter length, and displacement number; The iterative optimization operation is performed on the filter parameters of the filter to be optimized, and in each iterative optimization operation, the optimized filter to be optimized is used to filter the vibrator signal corresponding to the target filter signal, and the relevant spectrum kurtosis of the new target filter signal is calculated, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, the iterative optimization operation is stopped to generate the target filter, including: Acquiring the sampling frequency of the vibration signal and the initial displacement value of the filter to be optimized; Determining an initial deconvolution period of the filter to be optimized according to the sampling frequency of the vibration signal, and generating a plurality of deconvolution periods to be evaluated according to the initial deconvolution period; According to the initial displacement value and a number of deconvolution cycles to be evaluated, repeatedly performing an iterative optimization operation on the filter to be optimized until a target filter is generated; The iterative optimization operation includes: Obtaining a displacement value to be evaluated, a deconvolution period to be evaluated, and a filter to be optimized; wherein, initially, the displacement value to be evaluated is an initial displacement value; According to a number of deconvolution cycles to be evaluated and displacement values ​​to be evaluated, the filter to be optimized is updated to generate a filter to be evaluated; The vibrator signal corresponding to the target filter signal is filtered by using the filter to be evaluated to obtain the filter signals to be evaluated generated by the filter to be evaluated with different deconvolution periods to be evaluated under the displacement value to be evaluated, and the relevant spectral kurtosis corresponding to each filter signal to be evaluated; Sorting the relevant spectral kurtosis, and multiplying the second largest relevant spectral kurtosis by a preset multiple to generate a judgment threshold; Determine whether the maximum correlation spectrum kurtosis is greater than the determination threshold; If not, the displacement value to be evaluated is increased by 1 to generate the displacement value to be evaluated required for the next round of iterative optimization operation; If yes, the deconvolution period to be evaluated corresponding to the maximum correlation spectrum kurtosis is taken as the target deconvolution period, and the target filter length is calculated according to the filter signal to be evaluated corresponding to the maximum correlation spectrum kurtosis, and then the target displacement number is determined according to the target deconvolution period and the target filter length; According to the target deconvolution period, the target filter length, and the target displacement number, the filter parameters of the filter to be optimized are updated to generate a target filter.

7. A method for identifying a fault based on a power transmission line according to claim 6, characterized in that: The step of determining an initial deconvolution period of the filter to be optimized according to the sampling frequency of the vibration signal, and generating a plurality of deconvolution periods to be evaluated according to the initial deconvolution period, comprises: Get the preset frequency range and the preset period value range; According to the frequency range and the sampling frequency, the initial deconvolution period is calculated according to the following formula: Among them, T f is the initial deconvolution period, f s is the sampling frequency, f fault is the frequency range; According to the cycle value range and the initial deconvolution cycle, a plurality of deconvolution cycles to be evaluated are generated.

8. A fault identification device based on a transmission line, characterized in that: include: A data acquisition module, used to acquire a vibration signal of the transmission line and a first spectral kurtosis corresponding to the vibration signal; an adaptive filtering module, configured to divide the vibration signal into a plurality of vibration sub-signals according to the first spectral kurtosis and a preset adaptive filter, and generate a filter corresponding to each vibration sub-signal, a filtered signal, and a second spectral kurtosis value corresponding to each filtered signal; A filter determination module, used to determine the filter signal corresponding to the maximum second spectral kurtosis value as the target filter signal, and determine the filter that generates the target filter signal as the filter to be optimized; A filter optimization module, used to iteratively optimize the filter parameters of the filter to be optimized, and in each iterative optimization process, use the optimized filter to be optimized to filter the vibrator signal corresponding to the target filter signal, and calculate the relevant spectrum kurtosis of the new target filter signal, and when it is determined that the relevant spectrum kurtosis is greater than a preset judgment threshold, stop the iterative optimization operation and generate the target filter; A feature generation module, used to filter the vibration signal using the target filter to generate periodic fault features; A fault identification module is used to output a fault identification result of the transmission line according to the periodic fault characteristics.

9. A terminal device, characterized in that: The invention comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a fault identification method based on a transmission line as claimed in any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is executed, the device where the storage medium is located is controlled to execute a fault identification method based on a transmission line as claimed in any one of claims 1 to 7.

Images