Cable defect location method, apparatus, device, and medium

Abstract

The present application relates to a kind of cable defect positioning method, device, equipment and medium, it relates to defect detection technical field.The method includes: based on generalized S transformation estimation, the reflection coefficient spectrum of cable is transformed, the time-frequency spectrum of cable is obtained;Spectrum peak screening and pseudo-peak elimination are carried out to time-frequency spectrum, and the target signal of cable is obtained;Frequency domain summation is carried out to target signal, and the positioning curve of cable is obtained;Based on positioning curve and signal propagation velocity, the defect position of cable is determined.The precision positioning of cable defect can be realized by using the present method.

Description

Technical Field

[0001] This invention relates to the field of defect detection technology, and in particular to a method, apparatus, equipment and medium for locating cable defects. Background Technology

[0002] In recent years, frequency domain reflectometry (FDR) has been widely used in cable defect detection.

[0003] However, due to the dispersion effect, the frequency of the reflected wave changes with the sampling time, which leads to the generation of simple harmonic waves. The traditional FDR method uses the Fast Fourier Transform (FFT) as the kernel function. It transmits a signal of a specific frequency, receives the reflected wave generated by the defect in the cable, and uses the FFT to process the signal to finally generate a defect location map, thereby determining the location of the cable defect.

[0004] However, the final defect location map contains many interfering factors, resulting in low accuracy in defect location. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and medium for locating cable defects, which can improve the accuracy of locating cable defects.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for locating cable defects, comprising:

[0008] Based on the generalized S-transform estimation, the reflection coefficient spectrum of the cable is transformed to obtain the time spectrum of the cable;

[0009] The target signal of the cable is obtained by performing peak filtering and spurious peak elimination on the time spectrum.

[0010] The cable positioning curve is obtained by summing the target signal in the frequency domain.

[0011] Based on the positioning curve and signal propagation speed, the location of the cable defect is determined.

[0012] In one embodiment, peak filtering and spurious peak elimination are performed on the time spectrum to obtain the target signal of the cable, including:

[0013] Local peak points are extracted from the time spectrum to determine the first set of peak points corresponding to the time spectrum; wherein, the first set of peak points includes at least one peak point;

[0014] Based on the first set of peak points, pseudo-peak elimination is performed on the time spectrum to obtain the target signal of the cable.

[0015] In one embodiment, based on a first set of peak points, pseudo-peak elimination is performed on the time spectrum to obtain the target signal of the cable, including:

[0016] Based on the cable's structural parameters, the predicted defect time interval for the cable is determined; among which, the structural parameters include cable length and signal propagation speed.

[0017] Based on the predicted defect time interval and the peak points in the first peak point set, the time spectrum is subjected to the first pseudo-peak elimination to obtain the second time spectrum and the second peak point set corresponding to the second time spectrum.

[0018] Based on the frequency range of historical cable defects and the peak points in the second peak point set, a second pseudo-peak elimination is performed on the second time spectrum to obtain the third time spectrum and the third peak point set corresponding to the third time spectrum.

[0019] A third pseudo-peak elimination process is performed on the isolated pseudo-peaks in the third time spectrum and the third peak point set to obtain the target signal of the cable.

[0020] In one embodiment, frequency domain summation of the target signal is performed to obtain the cable positioning curve, including:

[0021] For the target signal, energy accumulation is performed in the frequency domain to obtain the cable location curve; the location curve is used to characterize the defect signal intensity at different locations of the cable.

[0022] In one embodiment, determining the location of a cable defect based on a positioning curve and signal propagation speed includes:

[0023] Based on the positioning curve, determine the location of the main peak in the positioning curve;

[0024] The location of cable defects is determined based on the location of the main peak and the signal propagation speed.

[0025] In one embodiment, determining the location of a cable defect based on the peak position and signal propagation speed includes:

[0026] The product of the main peak position and the signal propagation speed is used to determine the first defect position;

[0027] The defect location of the cable is determined by half of the first defect location.

[0028] In one embodiment, the reflection coefficient spectrum of the cable is transformed based on the generalized S-transform estimation to obtain the time spectrum of the cable, including:

[0029] Based on the generalized S-transform estimation, the real part of the reflection coefficient spectrum of the cable is transformed to obtain the time spectrum of the cable.

[0030] In a second aspect, the present invention provides a cable defect location device, comprising:

[0031] The transformation module is used to transform the reflection coefficient spectrum of the cable based on the generalized S-transform estimation to obtain the time spectrum of the cable.

[0032] The processing module is used to perform peak filtering and spurious peak elimination on the time spectrum to obtain the target signal of the cable;

[0033] The summation module is used to perform frequency domain summation on the target signal to obtain the cable positioning curve;

[0034] The positioning module is used to determine the location of cable defects based on the positioning curve and signal propagation speed.

[0035] Thirdly, the present invention provides a computing device, including a memory and a processor;

[0036] The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of the first aspects.

[0037] Fourthly, the present invention provides a computer-readable storage medium for storing a computer program for performing the method as described in any one of the first aspects.

[0038] Fifthly, the present invention provides a computer program product comprising one or more computer instructions, wherein when the computer instructions are executed by a computer, the computer performs the method as described in any one of the first aspects.

[0039] As can be seen from the above technical solution, the present invention has at least the following beneficial effects:

[0040] In this invention, the reflection coefficient spectrum of the cable is transformed based on generalized S-transform estimation to obtain the cable's time-frequency spectrum, providing a data foundation for determining the defect location. Furthermore, peak filtering and spurious peak elimination are performed on the time-frequency spectrum to obtain the target signal of the cable, improving signal accuracy by eliminating interference. Next, frequency domain summation of the target signal yields the cable's location curve, again clearly distinguishing the defect signal from the normal signal. Finally, the defect location of the cable can be determined based on the location curve and signal propagation speed. This scheme lays the foundation for improving signal resolution by introducing generalized S-transform estimation; furthermore, the peak filtering and spurious peak elimination processes effectively suppress interference from noise and other signals, ultimately achieving accurate cable defect location.

[0041] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this invention do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0042] Figure 1 This is an application environment diagram of a cable defect location method provided in an embodiment of the present invention;

[0043] Figure 2 This is a flowchart illustrating a cable defect location method provided in an embodiment of the present invention;

[0044] Figure 3 This is a structural diagram of a multi-defect cable provided in an embodiment of the present invention;

[0045] Figure 4 This is a schematic diagram of a process for obtaining a target signal of a cable according to an embodiment of the present invention;

[0046] Figure 5 This is a structural diagram of cable distribution parameters provided in an embodiment of the present invention;

[0047] Figure 6 This is a structural diagram of a defect-free cable provided in an embodiment of the present invention;

[0048] Figure 7 This is a structural block diagram of a cable defect location device provided in an embodiment of the present invention;

[0049] Figure 8 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0050] The terms "first," "second," and "third," etc., used in this specification and accompanying drawings are used to distinguish different objects, not to limit a specific order.

[0051] In embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in embodiments of the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0052] To ensure clarity and conciseness in the description of the following embodiments, a brief introduction to the related technologies is given first:

[0053] In recent years, frequency domain reflectometry (FDR) has gained widespread application due to its unique technical advantages in cable defect detection. Compared with traditional time domain reflectometry (TDR), FDR injects a swept-frequency signal into the cable and analyzes the spectral characteristics of the reflected signal, enabling it to more accurately capture impedance changes at defects. It exhibits higher sensitivity and frequency resolution, especially in identifying minute defects (such as localized insulation aging and poor joint contact). By utilizing the differences in the propagation characteristics of different frequency signals in cables, it locates defects and assesses their severity through the peak position and amplitude of the reflection coefficient spectrum. Therefore, it has become one of the mainstream technologies in condition monitoring of various types of cables, including power cables and communication cables.

[0054] However, FDR still faces two major technical bottlenecks in practical applications, severely limiting the accuracy of defect detection. One prominent issue stems from the dispersion effect in cable transmission. As a typical lossy transmission line, the distributed parameters (resistance, inductance, capacitance, and conductance) of cables are not constant but exhibit nonlinear characteristics that vary with signal frequency. This characteristic directly causes signals of different frequencies to propagate at different speeds within the cable, i.e., the dispersion effect. When a signal encounters a defect during propagation, the frequency components of the reflected wave separate due to the dispersion effect, causing the frequency of the reflected wave to dynamically change with propagation time. This dynamic change disrupts the single-frequency characteristic of simple harmonic waves, resulting in a large number of non-ideal simple harmonic wave components in the reflected signal.

[0055] In traditional FDR methods, the Fast Fourier Transform (FFT) is used as the kernel function for spectral analysis. However, the FFT assumes the signal is a stationary simple harmonic wave. Therefore, these non-ideal harmonic waves can create false peaks in the defect location map. The amplitude of these false peaks can sometimes be comparable to, or even higher than, the peak value of the reflected signal from the real defect, making them easily misidentified as real defects. Especially in multi-defect scenarios, the superposition of false peaks and real peak values ​​can severely interfere with the inspector's judgment of the number and location of defects, increasing the risk of missed detections and misjudgments.

[0056] In real-world testing environments, various noises are inevitably introduced during measurement, including thermal noise from electronic components, external electromagnetic interference (such as strong electromagnetic fields from substations and high-frequency interference from motor operation), and attenuation noise during signal transmission. For defects far from the test end, their reflected signals are significantly attenuated during propagation due to cable losses (such as dielectric and conductor losses), resulting in very weak signal amplitudes. When these weak reflected signals are superimposed on noise, the noise may completely obscure the reflection characteristics of the defect, making it impossible to effectively identify the reflected signals of distant defects in the spectrum. For example, in the testing of long-distance high-voltage cables, the reflected signals of minor defects several kilometers away from the test end may drop below the noise level after long-distance transmission. Traditional FDR methods, limited by their noise suppression capabilities, cannot capture such signals, leading to missed detections of distant defects. If these distant defects are not detected in time, they may gradually worsen over time, eventually causing serious accidents such as cable breakdown and power outages, posing a significant threat to the safe and stable operation of the power system.

[0057] Therefore, it is necessary to address the problem of poor positioning accuracy in cable defect detection using FDR technology in traditional solutions. Solving this problem can not only reduce false peaks and improve the accuracy of defect identification, but also enhance the ability to capture weak reflected signals, enabling effective detection of remote defects. This will provide reliable technical support for the full life-cycle health management of cables and reduce operational risks caused by missed or misjudged defects.

[0058] To make the technical solution of the present invention clearer and easier to understand, the application scenarios of the technical solution of the present invention will be described below with reference to the accompanying drawings. For example... Figure 1 As shown in the figure, this figure is a schematic diagram of an application scenario provided by an embodiment of the present invention.

[0059] In this application scenario, server 104, as the core processing unit, plays a crucial role in the in-depth analysis and precise processing of the cable reflection coefficient spectrum. For example, by performing a generalized S-transform estimation, and dynamically adjusting the time-frequency window, it can accurately capture the frequency characteristics of the reflected signal over time while effectively suppressing spectral aliasing caused by dispersion. For instance, when the defect reflection wave exhibits frequency separation due to dispersion, the generalized S-transform can clearly distinguish the frequency trajectory of the real defect from the chaotic distribution of spurious peaks in the time-frequency two-dimensional plane, providing a foundation for subsequent peak selection. Furthermore, based on the processing results of the generalized S-transform, server 104 can further execute peak selection and spurious peak elimination algorithms. For example, for spurious peaks superimposed in multi-defect scenarios, the algorithm can accurately remove spurious peak components by analyzing the phase consistency and energy distribution of the spectral peaks, ensuring that the remaining spectral peaks correspond to the real defect reflection signals. Furthermore, the server 104 combines the physical parameters of the cable (such as characteristic impedance and propagation speed) and the frequency information of the real spectral peaks in the reflection coefficient spectrum, and calculates the defect location of the cable through the established mathematical model (such as the mapping formula between the reflection coefficient and the defect distance). Furthermore, the server 104 formats the calculated defect location information (including specific coordinates, defect type probability, severity assessment, etc.) and transmits it to the terminal 102 through a stable communication network.

[0060] Terminal 102 serves as a human-machine interface, presenting information to relevant technicians in an intuitive and easy-to-understand format (such as 3D cable model annotations, defect location heatmaps, and inspection report documents). For example, technicians can clearly see the precise location of defects along the cable laying path on the terminal screen. Combined with the accompanying defect characteristic descriptions (such as "a loose joint exists 123.5 meters from the test end, with moderate reflected signal amplitude"), they can quickly formulate targeted troubleshooting plans—eliminating the need for a full cable inspection; they only need to carry specialized tools to the target location to carry out repairs, significantly shortening fault location time. Simultaneously, Terminal 102 also supports historical data review and comparative analysis. Technicians can access past inspection records for the cable to observe defect development trends, providing data support for determining the root cause of the fault (such as whether long-term vibration caused the joint to loosen), thus realizing a shift from passive maintenance to proactive prevention.

[0061] To make the technical solution of the present invention clearer and easier to understand, the cable defect location method provided by the embodiments of the present invention will be described below in conjunction with the above application scenarios. Figure 2 As shown in the figure, this is a flowchart of a cable defect location method provided by an embodiment of the present invention.

[0062] S201. Based on the generalized S-transform estimation, the reflection coefficient spectrum of the cable is transformed to obtain the time spectrum of the cable.

[0063] The generalized S-transform is a time-frequency analysis method developed from the short-time Fourier transform and wavelet transform. It preserves the phase information of the Fourier transform and, by introducing a frequency-varying window function, automatically adjusts the time resolution at different frequencies. It possesses high time resolution at high frequencies and high frequency resolution at low frequencies. This characteristic overcomes the limitation of the fixed window function in the short-time Fourier transform, thus allowing for more flexible time-frequency characteristic analysis of non-stationary signals. The time spectrum describes the energy distribution of a signal in both time and frequency dimensions, showing the frequency components and their intensity at different times. In cable inspection, it can help locate the time of fault occurrence (corresponding to cable location) and the fault-related frequency characteristics.

[0064] When a signal is transmitted in a cable and encounters impedance mismatch (such as cable joints, fault points, etc.), it will be reflected. The ratio of the incident signal to the reflected signal is the reflection coefficient. The reflection coefficient spectrum is the characteristic curve of the reflection coefficient changing with frequency. It contains frequency domain information of impedance changes in the cable and can be used to analyze the fault location, insulation condition, etc. of the cable.

[0065] like Figure 3 As shown, for a cable of length L0, when there are n (n>1) defect locations in the cable, It is the distance from the starting point of the first defect to the beginning. It is the distance from the end point of the first defect to the beginning point; The propagation coefficient of the first defect; It is the characteristic impedance of the first defect; It is the distance from the starting point of the second defect to the beginning. It is the distance from the end point of the second defect to the beginning point; The propagation coefficient of the second defect; It is the characteristic impedance of the second defect; It is the distance from the starting point of the nth defect to the beginning of the defect; It is the distance from the termination point of the nth defect to the beginning point; Let be the propagation coefficient of the nth defect; It is the characteristic impedance of the nth defect, and further, based on the reflected wave generated at the cable end. The superposition of all reflected waves at the cable head end can be obtained. Furthermore, the reflection coefficient spectrum at the cable head end can be derived based on the relationship between the reflection coefficient and the traveling wave voltage. The expression:

[0066] Formula (1);

[0067] in, The reflection coefficient spectrum at the cable head end. The sum of all reflected waves received at the cable head end; U0 is the signal voltage; A k Let α be the amplitude of the k-th reflected wave; k x is the attenuation constant of the k-th reflected wave; k v represents the location of the k-th reflected wave (i.e., the defect location); k is the propagation speed of the k-th reflected wave; n is the number of defects in the cable; k is the k-th defect; f is the frequency of the signal in the cable; e is the natural constant; exp() is an exponential function with base e.

[0068] One feasible approach is to transform the real part of the reflection coefficient spectrum of the cable based on the generalized S-transform estimation to obtain the time spectrum of the cable.

[0069] For example, the reflection coefficient spectrum can be analyzed. Extract the real part of the and denote it as s:

[0070] Formula (2);

[0071] Where Re[] represents taking the real part of the complex number.

[0072] Furthermore, considering that f is the frequency of the signal, which is a variable and not the inherent frequency of the reflection coefficient spectrum, for ease of writing, f in formula (2) is replaced with the variable m, resulting in:

[0073] Formula (3);

[0074] Performing a generalized S-transform on the reflection coefficient spectrum signal s(m) yields the time spectrum of the cable. :

[0075] Formula (4);

[0076] in, This can represent the amplitude / intensity of the reflection coefficient spectrum signal when the time delay is η, where η is the time delay; t is the time, i.e., the actual propagation time of the signal in the cable; j is the imaginary unit; f represents the frequency of the signal in the cable, which is a key parameter describing the rate of periodic change of the signal; W(η-t, f) is the window function, expressed as:

[0077] Formula (5);

[0078] in, It is an adaptive window width function, where λ and β are adjustment parameters that can be dynamically optimized based on the signal frequency components and noise level.

[0079] Optionally, λ and β can also be dynamically optimized. The specific optimization process can be expressed as follows:

[0080] Formula (6);

[0081] Formula (7);

[0082] Where k1 and k2 are empirical coefficients that can be determined based on the cable type (such as high-voltage cable, communication cable) and the test environment. k1 and k2 can achieve adaptive adjustment of the window function in the low-frequency band with wide-window smoothing and in the high-frequency band with narrow-window focusing; σ is the noise standard deviation, which can be estimated using the median absolute deviation; P(f) is the signal power spectral density. It is the median of the signal power spectral density; f low The initial frequency of P(f), f high It is the cutoff frequency of P(f).

[0083] The generalized S-transform estimation can overcome the limitations of the traditional Fourier transform in non-stationary signal processing, accurately capturing the time-varying frequency characteristics of the reflection coefficient spectrum, and providing a clear time-frequency domain basis for distinguishing between real faults and interference signals. Cable reflection signals are affected by dispersion effects, and their frequency components dynamically change with propagation time, exhibiting obvious non-stationary characteristics.

[0084] The generalized S-transform, by introducing an adjustable Gaussian window function, can automatically adjust the width of the time window according to the signal frequency—using a narrow window for high-frequency components to improve time resolution and a wide window for low-frequency components to ensure frequency resolution, thus clearly outlining the time distribution trajectory of different frequency components in the time spectrum. For example, the reflected signal of a real fault usually shows a continuous frequency trajectory in the time spectrum, and the trajectory shape corresponds to the type of defect (such as loose joints or insulation damage); while the spurious peaks generated by the dispersion effect show chaotic frequency jumps, and the difference between the two in the time spectrum is obvious.

[0085] This invention, through precise feature differentiation, reduces the possibility of misidentifying spurious peaks as real faults at the source, providing a data foundation for subsequent peak screening and spurious peak elimination. Secondly, this transformation process effectively suppresses noise interference and enhances the identifiability of weak fault signals, especially improving accuracy in detecting remote defects. In the cable reflection coefficient spectrum, the reflected signals from remote defects are often weak in amplitude due to propagation loss and are easily masked by measurement noise.

[0086] The generalized S-transform, through its energy concentration characteristics in the time-frequency domain, concentrates signal energy at its true time-frequency location, while noise energy is dispersed throughout the time-frequency plane, thus achieving effective separation of signal and noise. For example, for a tiny defect several kilometers away from the test site, its reflected signal may be submerged by noise in the time domain. However, in the time-frequency spectrum obtained by the generalized S-transform, its corresponding frequency components will form a relatively concentrated local region of energy. By setting an energy threshold, this region can be accurately extracted, avoiding the annihilation of weak signals by noise. This noise suppression capability significantly improves the detection sensitivity of distant defects, ensures the comprehensiveness of fault location, and reduces the risk of missed detections.

[0087] In this invention, the obtained time-frequency spectrum provides rich parameter dimensions for the quantitative analysis of fault characteristics, facilitating centimeter-level precise location. The time-frequency spectrum not only includes the frequency information of the fault reflection signal but also synchronously records the signal arrival time. By establishing a time-frequency-distance mapping relationship, feature points on the time-frequency spectrum can be directly associated with their physical locations on the cable. For example, the start time of a continuous frequency trajectory in the time-frequency spectrum corresponds to the reflection time of the fault signal. Combined with the signal propagation speed at that frequency, the distance between the fault and the test end can be accurately calculated. Simultaneously, changes in the trajectory amplitude can reflect the severity of the defect (e.g., a larger amplitude may correspond to a more severe defect). This multi-dimensional feature quantification means that fault location no longer relies on a single frequency or time parameter but combines comprehensive features in the time and frequency domains, further reducing location errors and ensuring high accuracy of the results.

[0088] In this invention, the intuitiveness of the time spectrum facilitates subsequent automated processing and manual judgment, improving the efficiency and reliability of fault location. The time spectrum presents the time-frequency distribution of the signal as a two-dimensional image. Whether it's the server's subsequent peak screening algorithm or the manual review by technicians, key features can be quickly identified through the image. For example, the algorithm can automatically mark potential fault locations by identifying continuous energy concentration bands in the time spectrum, while technicians can verify and correct the algorithm results by observing the morphological characteristics of the time spectrum. This forms a dual guarantee mechanism of machine processing and manual verification, further improving the reliability of fault location results.

[0089] In summary, the time spectrum of the cable obtained by estimating the generalized S-transform, by accurately capturing non-stationary signal characteristics, effectively suppressing noise, providing multi-dimensional quantization parameters, and enhancing the intuitiveness of the results, fundamentally improves the accuracy, comprehensiveness, and precision of fault location, providing an indispensable data foundation for the precise location of cable defects.

[0090] S202. Perform peak filtering and spurious peak elimination on the time spectrum to obtain the target signal of the cable.

[0091] Among them, a spectral peak refers to a local peak point where the energy is relatively concentrated and significantly higher than the surrounding area. It corresponds to the obvious energy distribution of the signal at a specific time and frequency, and is usually related to the reflected signal generated by impedance discontinuities in the cable (such as normal joints, defects, etc.). A spurious peak is a false peak in the time spectrum that is not generated by the actual impedance discontinuities of the cable. It may be caused by noise interference during signal acquisition (such as electromagnetic interference, equipment noise, etc.) or errors in signal preprocessing or transformation (such as spectral leakage in Fourier transform, improper selection of window function in generalized S-transform, etc.). The target signal is the signal component that can accurately reflect the true condition of the cable (such as defect location, nature, etc.), that is, the signal corresponding to the effective spectral peak.

[0092] As shown in the example above, you can first analyze the time spectrum. Preprocessing is performed, such as smoothing using moving averages to locally average the energy values ​​of the time spectrum, reducing minor fluctuations caused by noise and making the peak outlines clearer. Simultaneously, a reasonable energy threshold can be set to initially eliminate regions with excessively low energy, narrowing the scope of subsequent processing and improving efficiency. Further, peak detection algorithms (such as gradient-based methods) can be used to identify potential peaks in the time spectrum. For each point on the time spectrum, the energy difference between it and its surrounding neighboring points is calculated; when the energy of a point is greater than the energy of its adjacent points (up, down, left, and right), it is marked as a potential peak. Further, feature analysis can be performed on the detected potential peaks. Screening can be conducted based on the known characteristics of the cable target signal (e.g., peaks generated by defects typically have a relatively stable frequency range, certain energy intensity, and temporal location patterns). For example, an energy intensity threshold can be set to eliminate peaks with energy below that threshold, retaining those with stronger energy. Combined with normal cable structural information (e.g., known joint locations), peaks located within a reasonable time (location) range can be selected. Finally, the frequency distribution of the peaks is analyzed, retaining peaks within the typical frequency range related to cable defects to complete the peak selection.

[0093] To identify spurious peaks caused by noise, their continuity and stability can be analyzed. Noise spurious peaks typically appear as isolated points with large energy fluctuations and no continuous energy distribution trend around them. Connectivity analysis can be used to calculate the number of connected high-energy points around each peak. When the number falls below a set threshold, it is identified as a noise spurious peak and eliminated, thus obtaining the target signal for the cable. Peak filtering can accurately pinpoint effective peaks related to the fault from the time spectrum by establishing multi-dimensional feature criteria, avoiding interference from irrelevant signals. The time spectrum often contains a large amount of complex peak information, including characteristic peaks generated by real fault reflections, as well as stray peaks formed by inherent equipment noise and external electromagnetic interference. The peak filtering process can efficiently filter out redundant peaks that do not meet the characteristics of fault signals by setting strict filtering conditions—such as peak energy thresholds (eliminating low-energy noise peaks), frequency continuity (real fault peaks usually show a continuous frequency trajectory), and time stability (matching the theoretical reflection time window). For example, fixed-frequency clutter peaks generated by the oscillation of the test equipment itself may have high energy in the time spectrum but a disordered temporal distribution, and can be excluded by the continuity criterion; while the peaks of real faults, due to the propagation characteristics of reflected signals, show a continuous trajectory from low frequency to high frequency (or vice versa), and can be retained. This targeted filtering ensures that subsequent analysis focuses only on signals directly related to the fault, reducing the interference of invalid information on the localization algorithm from the source.

[0094] Secondly, spurious peak elimination, through in-depth analysis of the causes of spurious peaks, accurately removes false features generated by factors such as dispersion effects, significantly reducing the risk of misjudgment of faults. The generation of spurious peaks is closely related to the dispersion characteristics of cables—differences in the propagation speed of signals at different frequencies cause the reflected wave frequency to separate over time, forming spectral peaks that appear real but have no corresponding fault. If these spurious peaks are not eliminated, they are easily misjudged as real defects. The spurious peak elimination algorithm can learn the typical characteristics of spurious peaks (such as drastic frequency jumps, broken trajectories, and mismatches with the cable impedance model) and compare them with the physical model of real faults (such as the theoretical reflection frequency range of impedance abrupt changes at the defect). For example, when the frequency trajectory of a spectral peak deviates from the theoretical reflection trajectory calculated based on cable distribution parameters by more than a threshold, the algorithm can determine it as a spurious peak and eliminate it. In multi-defect scenarios where spurious peaks overlap with real peaks, the algorithm can also decompose the energy components of the spectral peaks to separate the energy contributed by the spurious peaks, ensuring the purity of the remaining target signal. This precise spurious peak elimination capability effectively avoids misjudging dispersion effects as faults, significantly improving the accuracy of fault location.

[0095] Furthermore, the target signal obtained after screening and elimination provides high-quality input data for the fault location algorithm, directly improving the quantitative level of location accuracy. The true spectral peaks contained in the target signal have clear time-frequency characteristics—their time information corresponds to the propagation duration of the reflected signal, and their frequency information is related to the propagation speed of the signal in the cable. The combination of these two can be used to accurately calculate the fault location using the formula "fault distance = (propagation speed × time corresponding to the main peak) / 2". Since the target signal has been freed from spurious peaks and noise interference, the measurement errors of its time and frequency parameters are controlled within a very small range (e.g., time error less than 10 nanoseconds, frequency error less than 1 kHz). Taking a 110kV cable as an example, the propagation speed is approximately... With a speed of meters per second, a time error of 10 nanoseconds corresponds to a distance error of only 2 millimeters. Combined with precise correction of frequency parameters, the final positioning error can be stabilized at the centimeter level, far superior to the positioning results of unprocessed signals. This high-precision parameter extraction enables the target signal to directly support the accurate calculation of the fault location, providing a data foundation for achieving "centimeter-level positioning."

[0096] Furthermore, the clean target signal provides a reliable basis for subsequent defect type identification and severity assessment, expanding the application value of fault location. The spectral peak amplitude and frequency range of the target signal are closely related to the nature of the defect—for example, the reflected signal from a loose joint typically has a strong amplitude in the low-to-mid frequency range, while insulation breakdown may produce a significant response in the high-frequency range. By analyzing these characteristics of the target signal and combining them with feature templates for different defect types in the sample database, the specific type of fault (such as mechanical damage, insulation aging, etc.) can be further determined. Simultaneously, the amplitude of the spectral peak is positively correlated with the severity of the defect (e.g., the reflected amplitude of severely damaged cables is much higher than that of minor scratches), and the energy integral value can also quantitatively assess the degree of impact of the defect on the cable insulation performance. In this invention, through location, qualitative, and quantitative analysis, not only is precise fault location locking achieved, but comprehensive information is also provided for the formulation of maintenance strategies, improving the targeting and efficiency of fault handling.

[0097] In summary, by performing peak filtering and spurious peak elimination on the time spectrum to obtain the target signal, and by accurately preserving effective features, removing false interference, and improving data purity, high-quality input is provided for the fault location algorithm. This not only significantly improves the location accuracy (centimeter level) but also reduces the risk of misjudgment and missed detection, while expanding the ability to deeply analyze the nature of the fault.

[0098] S203. Perform frequency domain summation on the target signal to obtain the cable positioning curve.

[0099] Frequency domain summation refers to the process of accumulating the energy or amplitude of a signal in the frequency dimension. For the time spectrum of the target signal, frequency domain summation is to add up the energy values ​​of all frequency components at each time point (corresponding to the cable position), thereby transforming the two-dimensional time spectrum information into a one-dimensional signal that changes with time (position), highlighting the energy accumulation effect at different positions. The positioning curve is a curve with the cable position (which can be converted from time) as the horizontal axis and the defect signal intensity at the corresponding position (i.e., the total energy value after frequency domain summation) as the vertical axis. It can intuitively reflect the signal energy distribution at different positions along the cable. Positions with higher energy values ​​often correspond to defect points and can be used as a basis for locating cable defects.

[0100] Optionally, the target signal after spectral peak filtering and spurious peak elimination can be... The data is standardized to ensure that the time and frequency axes accurately correspond to the actual physical quantities (e.g., time units are microseconds and frequency units are megahertz), and the energy values ​​are converted into numerical forms that can be directly accumulated (e.g., linear amplitude or energy density). Furthermore, based on the typical frequency characteristics of cable defect signals, an effective frequency summation range can be defined. For example, if it is known that a certain type of cable defect mainly generates signals in the range of 1 to 10 MHz, then only the energy in this frequency range in the time spectrum is accumulated to eliminate interference from irrelevant frequency components and improve the specificity of the summation results.

[0101] One feasible method is to perform energy accumulation calculations in the frequency domain for the target signal to obtain the cable's positioning curve.

[0102] Among them, the energy accumulation operation refers to the summation of the energy of all frequency components corresponding to each time point in the time spectrum of the target signal to obtain the total energy value at that time point. It can integrate the signal energy at different frequencies, enhance the strength of the effective signal, and weaken the influence of noise dispersed in each frequency. The positioning curve is used to characterize the defect signal strength at different locations of the cable.

[0103] For example, it can be targeted at the signal. At each point in time (i.e., at each corresponding location on the cable), the energy values ​​of all frequency points within the defined frequency range are accumulated, such as... And finally obtain the cable positioning curve. .in, The cutoff frequency for the signal power spectral density; The starting frequency of the signal power spectral density.

[0104] It should be noted that frequency domain summation can aggregate the energy of the target signal dispersed across different frequency components, significantly enhancing the identifiability of fault features on the location curve. Even after peak filtering and spurious peak elimination, the target signal, though free from interference, still retains fault reflection energy at different frequencies across multiple points in the time spectrum. Frequency domain summation, by accumulating the energy of each frequency component, concentrates the dispersed fault feature energy onto the location curve, making the corresponding peak more prominent. For example, a defect may generate reflection signals at multiple frequencies. Analyzing a single frequency might result in a relatively flat peak, but through frequency domain summation, these dispersed energies superimpose to form a steep peak, clearly visible on the location curve. This avoids missed fault detection due to weak energy at a single frequency, especially for defects with weak reflection signals at a specific frequency but responses at multiple frequencies; this energy aggregation significantly improves their detectability.

[0105] Secondly, this process effectively suppresses the influence of residual noise, reduces background interference on the positioning curve, and improves the accuracy of peak positions. Although the target signal has undergone spurious peak elimination, a small amount of random noise may still remain. This noise manifests as dispersed low-energy fluctuations in the frequency domain. Frequency domain summation averages the energy of a large number of frequency components, utilizing the randomness of the noise to cause them to cancel each other out during the summation process. Meanwhile, the energy of the fault signal is preserved and enhanced due to its consistency (the reflected signals of the same defect at different frequencies are correlated), thereby reducing the background noise level of the positioning curve. For example, in low signal-to-noise ratio scenarios, a single-frequency signal may be submerged by noise, but through frequency domain summation, the noise energy is weakened, and the peak of the fault signal becomes prominent, significantly improving the signal-to-noise ratio of the positioning curve and providing a clear waveform basis for accurately identifying the peak position (i.e., the fault position).

[0106] Furthermore, in the positioning curve generated by frequency domain summation, the fault location typically corresponds to a significant peak in the curve. This intuitive mapping avoids the complex process of cross-validation of multi-frequency data, allowing the positioning algorithm to directly determine the fault point by identifying the peak location. In addition, because frequency domain summation integrates multi-frequency information, the peak position of the positioning curve is less affected by fluctuations in a single frequency. Even if there is a slight error in the position calculation at a certain frequency, information from other frequencies can correct for it, thus ensuring the stability of the positioning results. For example, when the signal propagation speed changes slightly due to environmental factors (such as temperature), the position calculation at a single frequency may deviate. However, frequency domain summation, through the integration of multi-frequency information, can effectively offset this deviation, keeping the peak position stable and ensuring that the positioning error is controlled within the centimeter range.

[0107] Furthermore, the morphological characteristics of the location curve can provide a basis for judging the fault distribution in multi-defect scenarios, improving the location capability under complex working conditions. When there are multiple defects in a cable, the reflected signals of each defect will form their own energy distribution in the frequency domain. After summing the frequency domain signals, the location curve will show multiple independent peaks, each peak corresponding to the location of a defect. By analyzing the amplitude, width, and spacing of the peaks, the severity of the defects (the larger the amplitude, the more severe the defect) and their relative positions (the peak spacing corresponds to the physical distance between defects) can be distinguished. For example, when there are two defects in the middle section of the cable: loose joint and insulation aging, the location curve will show two obvious peaks, corresponding to the positions of the two defects respectively. The peak of insulation aging may have a lower amplitude due to weaker reflected energy. This clear multi-peak feature provides an intuitive basis for the simultaneous location of multiple defects, solving the problem of easily confusing locations in multi-defect scenarios using traditional methods.

[0108] In summary, the location curve is obtained by summing the target signal in the frequency domain. By aggregating fault energy, suppressing residual noise, establishing an intuitive mapping, and supporting the identification of multiple defects, the identifiability of fault characteristics is further enhanced, and the accuracy, stability, and adaptability of the location results are improved. This provides the final waveform basis for the accurate location of cable defects and is the core link connecting signal processing and actual location judgment.

[0109] S204. Based on the positioning curve and signal propagation speed, determine the location of cable defects.

[0110] Signal propagation speed refers to the speed at which a signal (such as an electromagnetic wave) propagates in the cable medium. Its magnitude is determined by factors such as the cable's material (such as the insulation material) and structure (such as the distance between the core wire and the shielding layer). It can generally be found in the cable's technical parameter manual or obtained through experimental measurement. The unit is usually meters per microsecond.

[0111] One possible approach is to determine the location of the main peak in the positioning curve based on the positioning curve; and to determine the location of cable defects based on the location of the main peak and the signal propagation speed.

[0112] The main peak position refers to the horizontal axis (such as time) corresponding to the highest and most significant peak value in the positioning curve.

[0113] The main peak is the key point on the positioning curve that best reflects the characteristics of the defect signal. Since the reflected signal energy generated by the defect is usually the strongest, the main peak often corresponds to the main defect of the cable.

[0114] Optionally, the product of the peak position and the signal propagation speed can be used to determine the first defect position; half of the first defect position can be used to determine the cable defect position.

[0115] The first defect location is a position parameter obtained by directly multiplying the main peak location (expressed in time) by the signal propagation speed.

[0116] It should be noted that since the signal is reflected back to the detection end after propagating to the defect point in the cable, the product of the main peak position and the signal propagation speed is actually the total distance of the signal's round-trip propagation, not the actual distance from the defect to the detection end. This can be referred to as the first defect position.

[0117] For example, the position f of the main peak can be... peak The product of f and the signal propagation speed v is used to determine the first defect location x0; half of the first defect location, i.e., 0.5 × x0, is used to determine the cable defect location; that is, defect location = 0.5 × f peak ×v.

[0118] Determining the location of the main peak based on the positioning curve allows for precise identification of the energy focal point corresponding to the fault, providing a highly reliable feature anchor point for subsequent calculations. After frequency domain summation processing, the energy of the fault signal is highly aggregated in the positioning curve, and the main peak, as the most prominent peak in the curve, directly corresponds to the location of the strongest energy in the defect reflection signal. In practice, by setting reasonable peak detection thresholds (e.g., more than 3 times higher than the background noise) and morphological criteria (e.g., the steepness and symmetry of the peak), small stray peaks (such as residual noise or weak interference) that may exist in the positioning curve can be effectively eliminated, ensuring that the extracted main peak is unique and accurate. For example, for a joint fault in the middle section of a cable, the main peak on the positioning curve will show a significant energy bulge, and the peak width is related to the physical size of the defect (the larger the defect, the wider the peak may be). Through the algorithm's accurate identification of this feature, it is possible to avoid misjudging nearby secondary fluctuations as the main peak, providing a stable reference point for subsequent location calculations.

[0119] Secondly, based on the quantitative calculation of the main peak position and signal propagation speed, a precise mapping between time-frequency characteristics and physical space was established, achieving centimeter-level location of the defect. Signal propagation speed is an inherent parameter of the cable (determined by the dielectric constant, permeability, etc. of the cable medium), and its value can be obtained through cable model lookup or prior calibration experiments (such as signal transmission testing on a defect-free cable section). The time parameter corresponding to the main peak position on the location curve (i.e., the round-trip time of the reflected signal from the test end to the defect and back), combined with the signal propagation speed, allows for direct calculation of the straight-line distance between the defect and the test end using the formula "Defect distance = (Propagation speed × Time corresponding to the main peak) / 2". Because the time parameter of the main peak position undergoes multiple optimizations through time-frequency analysis and frequency domain summation, its measurement error can be controlled at the nanosecond level. Combined with the precise propagation speed (error typically less than 0.5%), the final calculated defect location error can be stabilized at the centimeter level. For example, for a propagation speed of... For a cable traveling at meters per second, if the measurement error for the time corresponding to the main peak is 10 nanoseconds, the converted distance error is only 1 meter × (10 nanoseconds / 1 second) × m / s ÷ 2 = 0.001 m (i.e. 1 mm). This high-precision quantitative calculation is far superior to the meter-level error of traditional methods, meeting the needs of refined inspection of cable defects.

[0120] Furthermore, this process, through the combination of characteristic anchor points and quantitative formulas, effectively counteracts the interference of environmental factors on the positioning results, improving positioning stability under different working conditions. While the signal propagation speed of a cable is an inherent parameter, it is subject to slight fluctuations due to environmental factors such as temperature and humidity (e.g., for every 10°C increase in temperature, the propagation speed of some cables may change by 0.1% to 0.3%), and the time parameter of the main peak position may also deviate due to changes in signal attenuation characteristics. However, since the main peak position is the concentrated manifestation of fault energy, its time parameter changes are correlated with the fluctuations in propagation speed (e.g., when a slight decrease in propagation speed due to increased temperature, the reflection time will correspondingly lengthen), and the two form a complementary correction in the calculation. For example, in a high-temperature environment, if the measured propagation speed is slightly lower than the actual value, and the corresponding time of the main peak is slightly longer due to signal delay, the deviations are partially offset in the formula calculation, and the final distance error can still be controlled within the allowable range. This self-correcting capability ensures that the positioning results remain stable in complex environments, avoiding the significant impact of single parameter fluctuations on the final result.

[0121] Furthermore, this method also demonstrates high accuracy in locating multiple defects, enabling simultaneous differentiation and localization of multiple fault points. When a cable has multiple defects, the location curve will exhibit multiple independent peaks, each corresponding to the energy characteristics of a defect. By sequentially extracting the position parameters of each peak and calculating them in conjunction with the signal propagation speed, the specific location of each defect can be obtained. Moreover, the amplitude differences between the peaks can reflect the relative severity of the defects (e.g., a peak with a higher amplitude corresponds to a more severe defect). For example, if there is insulation damage 100 meters from the test end and a loose joint 200 meters away, two peaks will appear on the location curve, corresponding to the 100-meter and 200-meter locations respectively. Technicians can then formulate priority repair strategies based on the distance and severity of these peaks. This multi-peak identification and calculation capability solves the problem of chaotic location in multi-defect scenarios using traditional methods, significantly improving the efficiency of handling complex faults.

[0122] In summary, by determining the main peak position based on the positioning curve and calculating the defect position in combination with the signal propagation speed, and by accurately extracting feature anchor points, establishing quantitative mapping relationships, offsetting environmental interference, and supporting multi-defect positioning, a high-precision conversion of cable defects from "time-frequency characteristics" to "physical location" is achieved. This provides a decisive basis for accurate fault diagnosis and repair and is the final closed-loop link to ensure the effectiveness of cable defect positioning.

[0123] The aforementioned cable defect location method transforms the cable's reflection coefficient spectrum based on generalized S-transform estimation to obtain the cable's time-frequency spectrum, providing a data foundation for determining the defect location. Furthermore, peak filtering and spurious peak elimination are performed on the time-frequency spectrum to obtain the target signal of the cable, improving signal accuracy by eliminating interference. Next, frequency domain summation of the target signal yields the cable's location curve, further distinguishing the defect signal from the normal signal. Finally, the defect location can be determined based on the location curve and signal propagation speed. This scheme lays the foundation for improving signal resolution by introducing generalized S-transform estimation; furthermore, the peak filtering and spurious peak elimination processes effectively suppress interference from noise and other signals, ultimately achieving accurate cable defect location.

[0124] Based on the above embodiments, the present invention provides a detailed explanation of S202. Specifically, the present invention relates to the process of obtaining the target signal of the cable, as follows: Figure 4 As shown, the specific steps include:

[0125] S401. Extract local peak points from the time spectrum to determine the first set of peak points corresponding to the time spectrum.

[0126] The first set of peak points includes at least one peak point.

[0127] For example, all peak points (t) can be extracted from the time-spectrum graph by peak search. i f i A i ), and extract all peak points (t) i f i A i Construct the first set of peak points. Where i represents the sequence number of the peak point, t... i f represents the time corresponding to the i-th peak point; i A represents the signal frequency corresponding to the i-th peak point; i This represents the amplitude corresponding to the i-th peak point.

[0128] Local peak extraction can accurately pinpoint the energy concentration region of a fault signal from a complex time-frequency spectrum, providing microscopic feature information for distinguishing between real faults and background interference. The time-frequency spectrum is the energy distribution map of a signal in a two-dimensional plane of time and frequency. The energy of a fault reflection signal typically forms local peaks at specific time-frequency locations (such as energy spikes caused by defect reflection), while background noise manifests as dispersed low-energy fluctuations. By setting local peak detection algorithms (such as gradient-based extreme point identification, neighborhood energy comparison, etc.), points in the time-frequency spectrum that satisfy the condition of "highest local energy and higher than surrounding neighboring points" can be identified one by one. These points are the microscopic feature carriers of the fault signal in the time-frequency domain. For example, the reflected signal of an insulation defect may form a peak at 100kHz frequency and 500ns time, while simultaneously forming another associated peak at an adjacent 80kHz and 510ns time. These peaks together constitute the time-frequency characteristic trajectory of the defect. By extracting these local peak points, the fault signal can be separated from the complex time-frequency background, avoiding feature loss due to energy dispersion, and providing a microscopic feature basis for subsequent pseudo-peak identification and real fault location.

[0129] Secondly, the multi-dimensional characteristics (time, frequency, and energy) of the first peak point set provide rich criteria for subsequent spurious peak elimination and spectral peak screening, improving the accuracy of fault feature identification. Each peak point in the first peak point set contains three core parameters: time (the arrival time of the reflected signal), frequency (the frequency components of the signal), and energy (signal amplitude). The combination pattern of these parameters is closely related to the fault type and location. For example, the peak points of a real fault should match the theoretical reflection time (estimated based on cable length and propagation speed) in the time dimension, exhibit a continuous or correlated distribution in the frequency dimension (due to dispersion effects, the reflected signal of the same fault may form peaks at multiple frequency points), and be higher than the noise threshold and have a certain degree of stability in the energy dimension. In contrast, the peak points of spurious peaks may exhibit characteristics such as temporal disorder, frequency jumps, and fluctuating energy. By performing statistical analysis on the parameters of each point in the first peak point set (such as calculating frequency continuity, time consistency, and energy distribution entropy), the peak points corresponding to spurious peaks can be quickly identified and eliminated, ensuring that the remaining peak points are all related to the real fault. This multi-dimensional parameter-based screening mechanism is more accurate than a single energy threshold judgment, effectively reducing the probability of false peak misjudgment.

[0130] Furthermore, the first peak point set provides individualized feature labels for fault differentiation in multi-defect scenarios, solving the localization confusion problem caused by the superposition of multiple defect signals in traditional methods. When a cable has multiple defects, the reflected signals from different defects will form their own independent peak point clusters in the time spectrum—each cluster corresponds to the time-frequency characteristics of a defect. For example, a joint defect 100 meters from the test end may form a group of low-frequency (50 to 150 kHz) peak points around 200 ns, while an insulation failure 200 meters away may form a group of high-frequency (200 to 300 kHz) peak points around 400 ns. By extracting the first peak point set, these peak points belonging to different defects can be distinguished, and then clustering algorithms (such as time- or frequency-based clustering) can be used to group peak points of the same defect into one category, thereby achieving individualized feature extraction for multiple defects. This individualized processing method avoids mutual interference between multiple defect signals in the overall analysis, allowing the features of each defect to be independently identified and located, improving the localization accuracy in complex fault scenarios.

[0131] Furthermore, the energy parameters of the first set of peak points provide a quantitative basis for assessing the severity of defects, expanding the application value of fault location. The energy amplitude of the peak points is positively correlated with the severity of the defects—the more severe the defect, the stronger the reflected signal energy, and the higher the corresponding peak point energy parameter. By statistically analyzing the energy of each point in the first set of peak points (e.g., calculating average energy and maximum energy), the severity of the defects can be preliminarily determined, providing a reference for prioritizing maintenance. For example, if the maximum energy of one set of peak points is five times that of another set, it indicates that the reflected signal of its corresponding defect is stronger and may be closer to the critical fault state, requiring priority handling. This combination of location and quantitative assessment allows the fault location results to not only guide defect location investigation but also assist in formulating maintenance strategies, improving the efficiency of fault handling.

[0132] In summary, by extracting local peak points from the time spectrum and determining the first set of peak points, the micro-time-frequency characteristics of the fault are accurately captured, multi-dimensional screening criteria are provided, multi-defect differentiation is supported, and the severity of defects is quantitatively assessed. This provides high-quality feature data for subsequent fault location, thereby improving the accuracy, robustness, and application value of fault location from the source.

[0133] S402. Based on the first peak point set, perform pseudo-peak elimination on the time spectrum to obtain the target signal of the cable.

[0134] One possible approach involves determining the predicted defect time interval for the cable based on its structural parameters; performing a first pseudo-peak elimination on the time spectrum based on the predicted defect time interval and the peak points in the first peak point set to obtain a second time spectrum and a second peak point set corresponding to the second time spectrum; performing a second pseudo-peak elimination on the second time spectrum based on the peak points in the historical cable defect frequency interval and the second peak point set to obtain a third time spectrum and a third peak point set corresponding to the third time spectrum; and performing a third pseudo-peak elimination on the isolated pseudo-peaks in the third time spectrum and the third peak point set to obtain the target signal of the cable.

[0135] The structural parameters include cable length and signal propagation speed; the predicted defect time interval is a reasonable time range for limiting the defect signal calculated based on the structural parameters, which can be used to initially screen out spurious peaks in the time dimension; the second peak point set represents the set of peak points after removing spurious peaks based on the time dimension; the historical cable defect frequency interval can represent the typical frequency range of the defect signal obtained by summarizing historical defect data, which can be used to screen out spurious peaks from the frequency dimension. Optionally, different defects (such as insulation degradation and wire breakage) correspond to different frequency intervals; the third peak point set represents the set of peak points after removing spurious peaks based on the time and frequency dimensions.

[0136] For example, the predicted defect time interval can be calculated by combining the cable's structural parameters (such as cable length L0 and signal propagation speed v): t valid ∈[0, 2L0 / v], and thus, based on the predicted defect time interval, the frequency spectrum and the time of the first peak point exceeding t can be eliminated. valid The pseudo-peak portion is removed, and the removed time spectrum is smoothed (such as by filtering or smoothing operations) to obtain the second time spectrum and the second peak point set.

[0137] Furthermore, it is possible to base this on specific frequency ranges typically corresponding to historical cable defects (such as insulation degradation) [f] d1 f d2 (where f) d1 <f d2 f d1 f is the minimum value within a specific frequency range. d2 (the maximum value within a specific frequency range), for the frequency f of each peak point in the second peak point set. i With [f d1 f d2 Compare the frequencies and assign frequencies not within the specified frequency range [f] d1 f d2 The peak points of the second time spectrum and the second peak point set are removed, and the removed second time spectrum is smoothed to obtain the third time spectrum and the third peak point set.

[0138] Among them, fd1 =f i -2.5%×f i f d2 =f i +2.5%×f i .

[0139] Furthermore, based on morphological filtering, an opening operation (erosion + dilation) can be performed on the third time-frequency spectrum to eliminate isolated spurious peaks and retain continuous defect signal energy, thereby obtaining the target signal of the cable. The formula is:

[0140] Formula (8);

[0141] Where Erode() is the erosion operation function; Dilate() is the dilation operation function; B r It is a circular structuring element with radius r, where the parameter r is determined by the signal resolution (e.g., r takes the value 2 or 3).

[0142] By determining the predicted defect time interval based on cable structural parameters and performing the first spurious peak elimination, false signals that do not conform to the actual cable layout can be filtered out at the physical level, significantly reducing the source risk of positioning deviation. The cable's structural parameters (such as length, laying path, and joint location) determine the maximum time threshold for signal propagation in the cable—for example, for a 10-kilometer-long cable, the round-trip propagation time is approximately 100 microseconds (based on propagation speed). Peaks exceeding the predicted time interval (calculated in meters per second) obviously cannot correspond to real defects (they could be due to external electromagnetic interference or equipment noise). By eliminating peaks in the first peak set whose time parameters exceed the predicted interval, spurious peaks caused by test system mis-triggers or long-distance interference can be directly eliminated. For example, if a peak has a time parameter of 200 microseconds, far exceeding the theoretical maximum time of 100 microseconds for a 10-kilometer cable, it can be identified as a spurious peak and removed from the second time spectrum. This screening based on physical laws fundamentally avoids misjudging interference signals unrelated to the cable as fault characteristics, providing a more reliable second time spectrum and second peak set for subsequent processing.

[0143] Secondly, by combining historical cable defect frequency ranges with the second time-spectrum for a second round of spurious peak elimination, the fault patterns of similar equipment can be used to further focus on the true characteristics and improve the signal's relevance. Different types of cable defects (such as loose joints, insulation aging, and mechanical damage) often correspond to specific frequency response ranges—for example, the reflected signals from joint defects are mostly concentrated between 50 and 200 kHz, while insulation breakdown may have a significant response between 200 and 500 kHz. These patterns can be derived statistically from historical fault data. Removing peaks whose frequency parameters exceed the historical defect frequency range (e.g., a peak frequency of 1 MHz, far exceeding the frequency range of similar historical defects) from the second peak point concentration can effectively filter out abnormal frequency spurious peaks caused by dispersion effects or harmonics from the test equipment. For example, historical data for a certain cable shows that 90% of the defect frequencies are concentrated between 100 and 300 kHz. If an isolated peak of 400 kHz appears in the second peak point concentration, it is likely a spurious peak. The resulting third time-spectrum and third peak point set will more closely match the frequency characteristics of the true fault, reducing misjudgments caused by frequency anomalies.

[0144] Furthermore, a third elimination process is performed on isolated spurious peaks in the third time-frequency spectrum and the third peak point set. This process removes scattered spurious peaks caused by random noise or local interference, enhancing the continuity and consistency of fault characteristics. The reflected signals of genuine faults typically exhibit a continuous cluster of peak points in the time-frequency spectrum (because the same defect reflects at different frequencies), while isolated spurious peaks are mostly single discrete points (such as instantaneous electromagnetic pulse interference), without any associated peak points around them. By calculating the neighborhood density of each point in the third peak point set (e.g., no other peak points within 5 frequency / time units around a certain peak point), these isolated spurious peaks can be identified and eliminated. For example, a point in the third peak point set may have a frequency of 150kHz and a time of 50 microseconds, with no associated peak points around it, while the peak points corresponding to other genuine defects form continuous trajectories. After eliminating this isolated point, the time-frequency characteristics of the target signal are clearer, avoiding interference from single spurious peaks on the localization algorithm.

[0145] Furthermore, the three-step pseudo-peak elimination method, through layer-by-layer focusing, endows the final target signal with the triple attributes of "reasonable timing, typical frequency, and continuous features," providing highly pure input data for subsequent peak extraction and location calculations. Reasonable timing ensures the defect location is within the physical range of the cable; typical frequency improves the matching degree with the actual fault; and continuous features enhance the identifiability of the fault signal. For example, after three elimination steps, the peak clusters of the target signal strictly fall within the predicted time interval and historical frequency interval, forming a continuous time-frequency trajectory. The defect location error calculated by the localization algorithm based on such signals can be controlled at the centimeter level, far superior to the results without layered processing. Simultaneously, this layered processing also preserves independent feature clusters in multi-defect scenarios, ensuring that each real defect can be accurately identified and avoiding localization confusion caused by mutual interference between multiple defect signals.

[0146] In summary, the three-step pseudo-peak elimination method based on the first peak point set systematically removes various pseudo-peaks with time anomalies, frequency anomalies, and isolated existences through multiple verifications of physical laws, historical data, and characteristic morphology. The final target signal has high purity and clear characteristics, providing a reliable signal foundation for accurate fault location and significantly improving the accuracy, robustness, and stability of the location results.

[0147] In this embodiment of the invention, by eliminating spurious peaks in the time spectrum, interference from noise and other factors on the fault signal can be reduced, thereby improving the accuracy of fault location.

[0148] Based on the above embodiments, the embodiments of the present invention provide a detailed explanation of the expressions for the reflection coefficient spectrum in the above embodiments. For example... Figure 5 As shown, the distributed parameter model of the cable is presented. This model is constructed based on four core distributed parameters per unit length of the cable. The four core distributed parameters are the resistance R (Ω / m), inductance L (H / m), conductance G (S / m), and capacitance C (F / m) per unit length of the cable. Figure 5 In this diagram, I(x) is the input current (A) at a distance x meters from the cable end; U(x) is the input voltage (V) at a distance x meters from the cable end; Δx is the infinitesimal length; ΔI(x) is the current along the infinitesimal length; and ΔU(x) is the voltage along the infinitesimal length. Based on the fundamental principle of traveling wave reflection, any impedance mismatch will trigger signal reflection. For a cable with an open circuit at one end and no defects, during cable testing, the signal emitted from the test end will generate a reflected wave with a reflection coefficient of 1 at the end. The transmission process of this test signal can be combined with… Figure 6 (Defect-free cable model) Understanding. Considering that the signal energy will be greatly attenuated after multiple reflections, and for the sake of calculation convenience, this embodiment only focuses on the single reflected wave. Figure 6In this context, 0 represents the propagation coefficient of the cable body, and Z0 represents the inherent characteristic impedance of the cable. This is the propagation coefficient of the cable itself. Based on the principles of transmission line theory, the expression for the reflection coefficient ρ at the end of a healthy cable of length l can be derived as follows:

[0149] Formula (9);

[0150] Among them, Z L Z0 is the impedance value at the load end; Z0 is the inherent characteristic impedance of the cable.

[0151] The reflection coefficient at the cable start end can be represented by ρ0, and the input impedance at the start end is Z. in The mathematical expression can be represented as:

[0152] Formula (10);

[0153] Formula (11);

[0154] based on Figure 6 In the signal transmission model of a defect-free cable, if only a single reflection is considered, the signal will only be reflected once at the end of the cable. Based on this setting, the reflection coefficient spectrum under the defect-free cable signal transmission model can be derived. :

[0155] Formula (12);

[0156] in, Angular frequency is a key parameter that describes how fast an alternating current signal changes.

[0157] Because local defects cause reflections, these reflected waves superimposed at the beginning. Therefore, to further analyze this effect, a reflected signal transmission model of a cable of length L0 containing multiple defects was designed, such as... Figure 3 As shown in the (multi-defect cable model), this model intuitively presents the impact of n (n>1) defect locations in the cable on the signal transmission path and characteristics.

[0158] Through analysis Figure 3 The presented multi-defect cable structure leads to the conclusion that when there are n different defects inside the cable, the reflected waveform recorded at the beginning of the cable is actually a composite waveform formed by the superposition and mutual interference of the reflected waveforms generated independently by the n defects distributed on the cable and the waveform reflected back from the end of the cable in time and space.

[0159] Furthermore, the reflected wave H at the nth defectn It can be represented as:

[0160] Formula (13);

[0161] Where U0 is the voltage of the test signal; ρ k ρ is the reflection coefficient of the k-th defect; n γ is the reflection coefficient of the nth defect; k It is the propagation coefficient of the k-th defect; l (k+1)1 It is the distance from the (k+1)th defect starting point to the beginning; l k2 It is the distance from the termination point of the k-th defect to the beginning; l k1 It is the distance from the starting point of the k-th defect to the beginning.

[0162] At the same time, the reflected wave H generated at the end of the cable can be obtained. l The expression is:

[0163] Formula (14);

[0164] Furthermore, the superposition H of all reflected waves obtained at the cable head end can be obtained. sum The expression is:

[0165] Formula (15);

[0166] in, It is the reflected wave at the k-th defect.

[0167] Finally, the reflection coefficient spectrum at the cable head end can be derived based on the relationship between the reflection coefficient and the traveling wave voltage. The expression is formula (1).

[0168] It should be noted that the propagation coefficient of the k-th defect can be expressed as: That is, the propagation coefficient γ of the k-th defect. k It is a complex number; where, It is the attenuation constant of the k-th reflected wave. It is the phase constant of the k-th reflected wave, and f is the frequency of the signal, v k It is the propagation speed of the signal within the k-th defect; therefore, the cable reflection coefficient spectrum It is also a plural number.

[0169] In this embodiment of the invention, a detailed analysis of the cable distributed parameter model ensures the physical authenticity of the derived reflection coefficient spectrum. The cable's distributed parameters (such as resistance, inductance, capacitance, and conductance per unit length) are fundamental factors determining signal transmission characteristics, and they vary with cable material, structure, and operating environment. By clarifying the quantitative representation of these parameters in the reflection coefficient spectrum through the derivation process, the attenuation characteristics of the signal propagating in the cable can be accurately captured. This ensures that the derived reflection coefficient spectrum truly reflects the intrinsic relationship between the signal and cable characteristics, providing a reliable "benchmark template" for subsequent identification of reflected signals caused by defects, and avoiding distortion of reflection characteristics due to model simplification or missing parameters. Secondly, in-depth analysis of the defect-free cable model constructs a "zero reference frame" for defect identification, significantly improving the identification accuracy of defect features in reflected signals. The reflection coefficient spectrum of a defect-free cable is essentially an ideal response of a signal propagating in a uniform, intact medium, and its spectral characteristics exhibit clear regularity (such as the amplitude and phase stability range at specific frequencies). Furthermore, the detailed derivation process provides rigorous theoretical support for the mapping relationship between the reflection coefficient spectrum and the defect location, directly improving the accuracy of the localization algorithm. The core of defect localization lies in inferring the distance between the defect and the test point by analyzing the propagation time or frequency characteristics of the reflected signal, and the accuracy of this mapping relationship depends on the rigor of the reflection coefficient spectrum derivation. During the derivation process, by establishing mathematical relationships between the distributed parameters and signal propagation speed and attenuation coefficient, the propagation laws of different frequency components in the cable were clarified, leading to the derivation of a quantitative formula for the characteristic frequency in the reflection coefficient spectrum and the defect distance. This quantitative relationship based on theoretical derivation avoids the limitations of empirical formulas or fitting algorithms under complex working conditions, making the defect localization results traceable and reliable. In addition, this derivation process provides flexible expansion space for subsequent algorithm optimization and scenario adaptation, indirectly ensuring the accuracy of defect localization under different working conditions. In practical applications, cables may face complex conditions such as temperature changes and differences in laying methods (e.g., direct burial, conduit installation), which can affect the reflection coefficient spectrum by changing the distributed parameters. Because the derivation process clearly demonstrates the correlation mechanism between parameters and spectral characteristics, the model can be specifically modified for different operating conditions (such as introducing a temperature correction coefficient to adjust the resistance parameters), ensuring that the reflection coefficient spectrum always remains consistent with the actual cable condition. In summary, the detailed explanation of the cable reflection coefficient spectrum derivation process, by ensuring the physical authenticity of the spectrum, constructing a defect identification benchmark, establishing accurate mapping relationships, and supporting scenario adaptation, provides comprehensive support for accurate defect location results from theoretical foundation to practical application, significantly improving the reliability and practicality of cable defect diagnosis.

[0170] The above text combined Figures 1 to 6 The cable defect location method provided in the embodiments of the present invention has been described in detail. The apparatus and equipment provided in the embodiments of the present invention will be described below with reference to the accompanying drawings.

[0171] like Figure 7 As shown in the figure, this is a schematic diagram of a cable defect location device provided in an embodiment of the present invention. The cable defect location device 500 includes: a transformation module 501, a processing module 502, a summation module 503, and a location module 504, wherein:

[0172] Transformation module 501 is used to transform the reflection coefficient spectrum of the cable based on the generalized S-transform estimation to obtain the time spectrum of the cable;

[0173] Processing module 502 is used to perform peak filtering and spurious peak elimination on the time spectrum to obtain the target signal of the cable;

[0174] The summation module 503 is used to perform frequency domain summation on the target signal to obtain the cable positioning curve;

[0175] The positioning module 504 is used to determine the location of cable defects based on the positioning curve and signal propagation speed.

[0176] In one embodiment, the processing module 502 is specifically used for:

[0177] Local peak points are extracted from the time spectrum to determine the first set of peak points corresponding to the time spectrum; wherein the first set of peak points includes at least one peak point; based on the first set of peak points, pseudo-peak elimination is performed on the time spectrum to obtain the target signal of the cable.

[0178] In one embodiment, the processing module 502 is specifically used for:

[0179] Based on the cable's structural parameters, the predicted defect time interval is determined. These structural parameters include cable length and signal propagation speed. Based on the predicted defect time interval and the peak points in the first peak point set, a first pseudo-peak elimination is performed on the time spectrum to obtain a second time spectrum and a second peak point set corresponding to it. Based on the historical cable defect frequency interval and the peak points in the second peak point set, a second pseudo-peak elimination is performed on the second time spectrum to obtain a third time spectrum and a third peak point set corresponding to it. A third pseudo-peak elimination is performed on the isolated pseudo-peaks in the third time spectrum and the third peak point set to obtain the cable's target signal.

[0180] In one embodiment, the summing module 503 is specifically used for:

[0181] For the target signal, energy accumulation is performed in the frequency domain to obtain the cable location curve; the location curve is used to characterize the defect signal intensity at different locations of the cable.

[0182] In one embodiment, the positioning module 504 is specifically used for:

[0183] Based on the positioning curve, determine the location of the main peak in the positioning curve; based on the location of the main peak and the signal propagation speed, determine the location of the cable defect.

[0184] In one embodiment, the positioning module 504 is specifically used for:

[0185] The product of the main peak position and the signal propagation speed is determined as the first defect position; half of the first defect position is determined as the cable defect position.

[0186] The cable defect locating device 500 according to an embodiment of the present invention can correspond to performing the method described in the embodiment of the present invention, and the other operations and / or functions of each module / unit of the cable defect locating device 500 described above are respectively for implementing Figures 2-6 For the sake of brevity, the corresponding processes of each method in the illustrated embodiments will not be described in detail here.

[0187] This invention also provides a computing device. This computing device can be a local computing device or an application server.

[0188] like Figure 8 As shown in the figure, this is a schematic diagram of a computing device provided in an embodiment of the present invention. The computing device 700 includes a bus 701, a processor 702, a communication interface 703, and a memory 704. The processor 702, the memory 704, and the communication interface 703 communicate with each other via the bus 701.

[0189] The 701 bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 8 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0190] The processor 702 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).

[0191] Communication interface 703 is used for external communication. For example, communication interface 703 can be used to communicate with terminal 102. Communication interface 703 is used to send the defect location of the cable to terminal 102 so that terminal 102 can display the defect location result of the cable.

[0192] Memory 704 may include volatile memory, such as random access memory (RAM). Memory 704 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0193] The memory 704 stores executable code, and the processor 702 executes the executable code to perform the aforementioned cable defect location method.

[0194] Specifically, in achieving Figure 7 In the case of the illustrated embodiment, and Figure 7 When the modules or units of the cable defect location device described in the embodiment are implemented by software, the following steps are performed: Figure 7 The software or program code required for the functions of each module / unit can be partially or entirely stored in the memory 704. The processor 702 executes the program code corresponding to each unit stored in the memory 704 to perform the aforementioned cable defect location method.

[0195] This invention also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that a computing device can store, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct a computing device to perform the aforementioned cable defect location method.

[0196] This invention also provides a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this invention are generated.

[0197] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0198] When the computer program product is executed by a computer, the computer performs any of the aforementioned cable defect location methods. The computer program product can be a software installation package; when any of the aforementioned cable defect location methods needs to be used, the computer program product can be downloaded and executed on the computer.

[0199] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0200] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for locating cable defects, characterized in that, The method includes: Based on the generalized S-transform estimation, the reflection coefficient spectrum of the cable is transformed to obtain the time spectrum of the cable; Local peak points are extracted from the time spectrum to determine a first set of peak points corresponding to the time spectrum; wherein, the first set of peak points includes at least one peak point; Based on the structural parameters of the cable, the predicted defect time interval for the cable is determined; wherein, the structural parameters include cable length and signal propagation speed; Based on the predicted defect time interval and the peak points in the first peak point set, the time spectrum is subjected to the first pseudo-peak elimination to obtain the second time spectrum and the second peak point set corresponding to the second time spectrum; Based on the frequency range of historical cable defects and the peak points in the second peak point set, a second pseudo-peak elimination is performed on the second time spectrum to obtain the third time spectrum and the third peak point set corresponding to the third time spectrum. A third pseudo-peak elimination process is performed on the isolated pseudo-peaks in the third time spectrum and the third peak point set to obtain the target signal of the cable; The target signal is summed in the frequency domain to obtain the positioning curve of the cable; Based on the positioning curve and signal propagation speed, the location of the defect in the cable is determined.

2. The method according to claim 1, characterized in that, The step of summing the target signal in the frequency domain to obtain the positioning curve of the cable includes: For the target signal, an energy accumulation operation is performed in the frequency domain to obtain the positioning curve of the cable; wherein, the positioning curve is used to characterize the defect signal intensity at different locations of the cable.

3. The method according to claim 1, characterized in that, Determining the location of the cable defect based on the positioning curve and signal propagation speed includes: Based on the positioning curve, determine the position of the main peak in the positioning curve; Based on the location of the main peak and the signal propagation speed, the location of the defect in the cable is determined.

4. The method according to claim 3, characterized in that, Determining the location of the cable defect based on the main peak position and signal propagation speed includes: The product of the main peak position and the signal propagation speed is determined as the first defect position; The defect location of the cable is determined by half of the first defect location.

5. The method according to claim 1, characterized in that, The transformation of the reflection coefficient spectrum of the cable based on the generalized S-transform estimation to obtain the time spectrum of the cable includes: Based on the generalized S-transform estimation, the real part of the reflection coefficient spectrum of the cable is transformed to obtain the time spectrum of the cable.

6. A cable defect location device, characterized in that, The device includes: The transformation module is used to transform the reflection coefficient spectrum of the cable based on the generalized S-transform estimation to obtain the time spectrum of the cable; The processing module is used to perform peak filtering and spurious peak elimination on the time spectrum to obtain the target signal of the cable; The summation module is used to perform frequency domain summation on the target signal to obtain the positioning curve of the cable; A positioning module is used to determine the location of the defect in the cable based on the positioning curve and the signal propagation speed. The processing module is configured to extract local peak points from the time spectrum to determine a first set of peak points corresponding to the time spectrum; wherein the first set of peak points includes at least one peak point; determine a predicted defect time interval corresponding to the cable based on the structural parameters of the cable; wherein the structural parameters include cable length and signal propagation speed; perform a first pseudo-peak elimination on the time spectrum based on the predicted defect time interval and the peak points in the first set of peak points to obtain a second time spectrum and a second set of peak points corresponding to the second time spectrum; perform a second pseudo-peak elimination on the second time spectrum based on the historical cable defect frequency interval and the peak points in the second set of peak points to obtain a third time spectrum and a third set of peak points corresponding to the third time spectrum; and perform a third pseudo-peak elimination on the isolated pseudo-peaks in the third time spectrum and the third set of peak points to obtain the target signal of the cable.

7. A computing device, characterized in that, Including memory and processor; The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program for performing the method as described in any one of claims 1 to 5.

Images