Power cable traveling wave fault detection method and system, and storage medium

Abstract

The invention relates to a power cable traveling wave fault detection method and system and a storage medium, and relates to the technical field of intelligent fault detection. The power cable traveling wave fault detection method comprises the following steps: acquiring and decomposing an original signal of a power cable to obtain a traveling wave sampling signal, and recording corresponding sampling time; analyzing and correcting clock data transmitted by the Beidou time service module to obtain an absolute time mark; associating and storing the traveling wave sampling signal and the absolute time mark according to the sampling time to generate a traveling wave time data pair; key feature parameters are extracted based on the traveling wave time data pair, and a time domain feature graph and a polar coordinate scatter diagram are generated; analyzing the time-domain characteristic diagram and the polar coordinate scatter diagram according to a preset fault judgment threshold interval, and judging the fault type of the traveling wave sampling signal; by combining the quantum time synchronization technology with multi-modal feature analysis, the fault positioning precision is improved, the false alarm rate is reduced, the detection speed is increased, and the cable fault identification problem is effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent fault detection, in particular to a power cable traveling wave fault detection method, system and storage medium. BACKGROUND

[0002] In the field of power supply, medium voltage cables and distribution cables are key components. At present, after the cable of a large project is laid, it is difficult to find the cable damage caused by slight bending in the acceptance link, and there is a lack of effective means to monitor cable aging for a long time after completion. At the same time, important power consumption places such as banks, public security, hospital operating rooms have very high requirements for power supply continuity, and some provincial network companies have also put forward a 99.999% power supply stability index.

[0003] Application No. CN201911146831.9 discloses a fault detection method based on Beidou timing and power distribution network combination, which installs synchronous acquisition (recording wave) devices at the entrance and exit of the power distribution network through the line, and the synchronous acquisition (recording wave) devices are equipped with high-precision time modules based on Beidou time technology. At a predetermined time point, the waveforms of the entrance and exit are recorded at the same time, and the waveform data is transmitted to the middle station / backstage. The data or waveforms obtained at the data center can be compared and analyzed on the same time axis. Thus, the power supply state analysis result at the same time in the power distribution network can be obtained. After waveform comparison, it is judged whether there is an abnormality or fault in the line. The monitoring phenomenon or fault point source such as voltage sag, harmonic, grounding and loop switching can be quickly and accurately monitored. Through the online monitoring end of the Internet of Things, emergency state information or defects of the line equipment can be quickly sent to the strategy end, and the pain point of unclear responsibility boundary in the process of power distribution can also be solved.

[0004] The prior art in the above has the following defects: 1. The existing cable monitoring method has defects in fault response speed, positioning accuracy, fault type identification and intelligent processing, and cannot meet the actual demand; 2. It cannot monitor the early fault breakdown and power aging condition of the cable in real time and accurately, discover abnormal waveforms in time, and effectively save fault data in the high-speed sampling process. SUMMARY

[0005] In view of the defects of the prior art, the purpose of the present application is to provide a power cable traveling wave fault detection method, system and storage medium, which combines quantum time synchronization technology with multi-modal feature analysis, improves fault positioning accuracy, reduces false alarm rate, improves detection speed, and effectively solves the problem of early hidden fault identification of the cable.

[0006] The purpose of the present application is realized by the following technical scheme: A power cable traveling wave fault detection method, comprising: Collect and decompose the original signal of the power cable to obtain a traveling wave sampling signal, and record a corresponding sampling time; Parse and correct clock data transmitted by a Beidou timing module to obtain an absolute time scale; Associate and store the traveling wave sampling signal with the absolute time scale according to the sampling time to generate a traveling wave time data pair; Extract key feature parameters based on the traveling wave time data pair to generate a time domain feature map and a polar coordinate scatter plot; According to a preset fault determination threshold interval, analyze the time domain feature map and the polar coordinate scatter plot to determine the fault type of the traveling wave sampling signal.

[0007] By adopting the above technical scheme, the cable traveling wave signal is collected in real time through the synchronous sampling technology combined with the wavelet transform algorithm, the PPS+ToD time message output by the Beidou timing module is optimized through the adaptive Kalman filter, and the absolute time scale association mechanism with nanosecond level precision is established. The improved Teager energy operator (TEO) is used to extract the traveling wave head feature, and the time domain energy distribution atlas is constructed, and the Hilbert transform is used to extract the phase feature to generate the polar coordinate scatter plot. A multi-dimensional feature fusion model is established based on the deep belief network (DBN), and the sliding time window algorithm is used to dynamically compare the feature parameters with the preset threshold interval, so that the millisecond level accurate identification of short circuit, grounding and broken line faults is realized. By combining the quantum time synchronization technology with the multi-modal feature analysis, the fault positioning precision is improved, the false alarm rate is reduced, the detection speed is improved, and the early hidden fault identification problem of the cable is effectively solved.

[0008] The application further provides that: the specific steps of collecting and decomposing the original signal of the power cable to obtain the traveling wave sampling signal and recording the sampling time include: Synchronize the internal clocks of all power collectors according to a reference clock to obtain a unified clock; Control the power collectors to collect signals of the power cable based on the unified clock to obtain original signals; Filter out power frequency noise from the original signals to obtain traveling wave sampling signals; Empirical mode decomposition is performed on the traveling wave sampling signals to obtain intrinsic mode signals; Perform Hilbert transform on the intrinsic mode signals to calculate instantaneous frequencies; Identify sudden change points of the instantaneous frequencies according to a time-frequency plane to obtain traveling wave heads; Detect second derivative extreme points of the traveling wave heads according to scale functions to determine sampling times.

[0009] By adopting the technical scheme, the nanosecond-level time alignment of the multi-node power collector is realized based on the IEEE1588v2 high-precision clock synchronization protocol, the mode aliasing phenomenon is suppressed by using the improved EMD algorithm (CEEMDAN), the power frequency interference is eliminated by using the VMD-WPT hybrid denoising method combined with the adaptive threshold, and the traveling wave mutation feature is accurately extracted by HHT time-frequency analysis. First, a synchronous clock domain is constructed, three cascaded filters (Butterworth+Wavelet+Median) are used to realize 50Hz power frequency noise suppression, CEEMDAN is used to decompose the original signal into 12 layers of IMF components, Hilbert-Huang transform is used to construct a time-frequency matrix, an improved Canny operator is used to detect the time-frequency plane mutation point, and finally, the SOED based on the second derivative extreme point detection algorithm of the scale function (SOED) is used to determine the traveling wave head time, so that the sub-microsecond-level sampling time positioning is realized; the signal-to-noise ratio is improved, and the traveling wave positioning accuracy is improved.

[0010] The application is further provided: the specific steps of analyzing and correcting the clock data transmitted by the Beidou time service module to obtain an absolute time scale include: The original message transmitted by the Beidou time service module is analyzed and separated to obtain the transmission time, Beidou time, leap second parameter, satellite clock difference and reception time; According to the time difference between the reception time and the transmission time, the light speed is combined to calculate the reception delay clock difference; According to the satellite elevation angle and ionospheric parameter combined with the local time, the ionospheric delay clock difference is calculated; According to the reception elevation and temperature and humidity combined with the refraction number, the convection delay clock difference is calculated; According to the relativistic effect on the satellite clock difference, the ionospheric delay clock difference, the convection delay clock difference and the reception delay clock difference, a weighted operation is performed to obtain a correction factor; According to the leap second parameter combined with the correction factor, the Beidou time is converted to obtain the coordinated universal time; The coordinated universal time is time-stamped with the reception time, and combined with the carrier phase smoothing pseudo-range to generate an absolute time scale.

[0011] By adopting the technical scheme, core parameters such as transmission time, Beidou time (BDT), leap second parameter, satellite clock difference and reception time are obtained by analyzing the Beidou original message, a four-dimensional error correction model (reception delay clock difference adopts an optical speed time delay calculation model, ionization delay clock difference is fused with satellite elevation angle / local time / ionosphere parameter based on a Klobuchar model, and convection delay clock difference is integrated with temperature / humidity / elevation / air refraction number according to a Saastamoinen model) and a relativistic effect compensation algorithm (satellite clock difference, ionosphere / convective layer delay and reception delay are weighted and fused in multiple items) are combined to generate a correction factor; then, the BDT is converted into coordinated universal time (UTC) through leap second compensation and correction factor calibration, and finally, a nanosecond-level absolute time scale is generated through time stamp alignment (reception time is synchronized with UTC) and carrier phase smoothing pseudo-range technology (Hatch filtering); the accuracy of the time service system is improved, and the environmental adaptability is improved.

[0012] The application further provides that the specific step of associating and storing the traveling wave sampling signal with the absolute time scale according to the sampling time to generate a traveling wave time data pair comprises: The sampling time is synchronized with the absolute time scale according to the second pulse signal transmitted by the Beidou time service module in combination with a phase-locked loop, the absolute time is determined, and a synchronization code is marked; The traveling wave sampling signal corresponding to the sampling time is associated with the absolute time according to the synchronization code to obtain an initial traveling wave association pair; The delay time of the trigger point from the second pulse signal to the traveling wave sampling signal is calculated to obtain an actual time difference; The initial traveling wave association pair is corrected according to the actual time difference and the refresh rate per cycle to obtain a traveling wave time data pair; The traveling wave time data pair, the original signal and the inherent modal signal are compressed and stored in combination with the synchronization code according to a preset data structure.

[0013] By adopting the technical scheme, the local crystal oscillator is tamed through the phase-locked loop to realize the ns-level synchronization (phase locking error < crystal frequency pulse width) of the second pulse and the absolute time scale, the traveling wave time data pair is generated through dynamic compensation of the trigger point delay, and the original signal, the inherent modal signal and the corrected time scale are jointly compressed and stored through Huffman coding based on the synchronization code; the time synchronization jitter is reduced, and the positioning error is reduced.

[0014] The application further provides that the specific step of extracting key feature parameters based on the traveling wave time data pair to generate a time domain feature map and a polar coordinate scatter plot comprises: The original signal and the traveling wave time data pair are aligned according to the time sequence to obtain a traveling wave start time stamp, a peak time stamp, a traveling wave end time stamp and a normal current signal amplitude. According to the absolute time, the traveling wave sampling signal in the traveling wave time data pair is characterized extracted, and the current traveling wave peak value is obtained; According to the current difference value between the current traveling wave peak value and the normal current signal amplitude, and in combination with the time difference value between the traveling wave start timestamp and the peak timestamp, the change rate of the traveling wave sampling signal is calculated, and the amplitude change rate is obtained; The traveling wave signal is analyzed in the frequency domain, and a plurality of traveling wave frequencies are obtained; According to the peak timestamp and the current traveling wave peak value, in combination with the amplitude change rate, a time domain feature map is generated; According to all the traveling wave frequencies in combination with the traveling wave end timestamp, a polar coordinate scatter plot is generated.

[0015] By adopting the above technical scheme, the traveling wave start / peak / end timestamps and the normal current amplitude are accurately extracted through the timestamp alignment technology, the current traveling wave peak value is calculated in combination with the absolute time scale synchronized by the phase-locked loop; the dynamic differential algorithm is innovatively adopted, the amplitude change rate is generated in real time by dividing the current difference value (the traveling wave peak value-the normal amplitude) by the time delay (the peak timestamp-the start timestamp); the multi-frequency band traveling wave frequency component is extracted through the FFT frequency domain analysis; finally, the time domain parameters (the peak value, the change rate) are fused to generate the time domain feature map with gradient markers, and the polar coordinate scatter plot (the radius represents the frequency, and the angle represents the time delay) is constructed based on the traveling wave frequency and the end timestamp, so that the three-dimensional visualization of the fault traveling wave space-time evolution characteristics is realized; the fault feature completeness is improved, and the dynamic response accuracy is improved.

[0016] The application is further provided as follows: the specific steps of determining the fault type of the traveling wave sampling signal according to the preset fault determination threshold interval analysis of the time domain feature map and the polar coordinate scatter plot include: The traveling wave sampling signal is decoupled and modulus extreme value transformed to obtain a zero modulus maximum value, and the zero modulus maximum value is accumulated in a preset period to obtain an accumulated maximum amplitude; According to the ratio calculation of the zero modulus maximum value and the maximum value in all the zero modulus maximum values, a traveling wave amplitude measure is obtained; The time domain feature map and the polar coordinate scatter plot are characterized extracted to obtain a characteristic parameter; the characteristic parameter includes an amplitude decay rate; The current traveling wave peak value is compared with the amplitude decay rate and the preset fault determination threshold interval to determine; If the current traveling wave peak value is located in the first determination threshold interval, and the amplitude decay rate is located in the first decay slope interval, it is determined that the fault type of the current traveling wave sampling signal is a lightning stroke fault; If the current traveling wave peak value is located in a second determination threshold interval and the amplitude decay rate is located in a second decay slope interval, it is determined that the fault type of the current traveling wave sampling signal is a short-circuit fault; If the current traveling wave peak value is located in a third determination threshold interval and periodically fluctuates, it is determined that the fault type of the current traveling wave sampling signal is an in-station fault; If the current traveling wave peak value is located in a fourth determination threshold interval and fluctuates within a preset period, it is determined that the fault type of the current traveling wave sampling signal is a broken line fault; If the accumulated maximum amplitude of the current traveling wave peak value is greater than or equal to the traveling wave amplitude measure, it is determined that the fault type of the current traveling wave sampling signal is a ground fault; The fault type of the traveling wave sampling signal is determined again according to a characteristic parameter; the characteristic parameter includes a wave head time, a polarity and a main frequency; If the wave head time is located in a preset first duration threshold interval, the main frequency is located in a preset first frequency threshold interval, and the polarity is positive, it is determined that the fault type of the current traveling wave sampling signal is a lightning strike fault; If the wave head time is located in a preset second duration threshold interval, the main frequency is located in a preset second frequency threshold interval, and the polarity is negative and bidirectional alternation, it is determined that the fault type of the current traveling wave sampling signal is a short-circuit fault; If the wave head time is located in a preset third duration threshold interval, the main frequency is located in a preset third frequency threshold interval, it is determined that the fault type of the current traveling wave sampling signal is an in-station fault; If the wave head time is located in a preset fourth duration threshold interval, the main frequency is located in a preset fourth frequency threshold interval, it is determined that the fault type of the current traveling wave sampling signal is a broken line fault; If the wave head time is located in a preset fifth duration threshold interval, the main frequency is located in a preset fifth frequency threshold interval, and the polarity is negative, it is determined that the fault type of the current traveling wave sampling signal is a ground fault; If the fault type determination of the same traveling wave sampling signal is the same type, it is indicated that the fault type determination of the traveling wave sampling signal is correct; otherwise, the second determination result is taken as the fault type of the traveling wave sampling signal.

[0017] By adopting the technical scheme, the zero-mode maximum value is obtained by decoupling the traveling wave signal, the accumulated maximum amplitude is generated by periodic accumulation, and the preliminary judgment condition is constructed based on the traveling wave amplitude measure (ratio of zero-mode maximum value to global maximum value); the amplitude decay rate and other parameters are extracted by combining the time-domain feature map and the polar coordinate scatter plot, first, the preliminary classification of lightning stroke, short circuit, in-station fault and broken line fault is realized by the threshold interval of current traveling wave peak value and decay rate, and the ground fault is identified by using the relationship between the accumulated maximum amplitude and the traveling wave amplitude measure; then, the result is checked by the secondary determination of wave head time, polarity and main frequency characteristics, if the two results are consistent, the fault type is confirmed, otherwise, the secondary determination is used as the criterion; the single feature misjudgment rate is reduced.

[0018] In a second aspect, the application further provides a fault detection system of power cable traveling wave, which adopts the following technical scheme: A fault detection method of power cable traveling wave, which is used for the fault detection system of power cable traveling wave, comprises the following steps: A synchronous sampling module is used for controlling the ADC to synchronously collect original signals in the power cable line; A filter processing module is used for denoising, filtering and normalizing the original signals to obtain traveling wave sampling signals; A time decoding module is used for analyzing the time information output by the Beidou timing module to obtain an absolute time scale, and associating the traveling wave sampling signals to generate a traveling wave time data pair; A transform operation module is used for extracting characteristic parameters containing wave head time, polarity and main frequency from the traveling wave time data pair; A data formatting module is used for compressively storing the traveling wave time data pair, the original signal and the inherent modal signal according to synchronous encoding; A fault determination module is used for determining the fault type of the traveling wave sampling signal according to the characteristic parameters and the preset fault determination threshold interval.

[0019] By adopting the above technical solution, the synchronous sampling module uses phase-locked loop (PLL) technology to control a multi-channel ADC (such as the ADC7606) to achieve nanosecond time alignment. The filtering processing module integrates wavelet threshold denoising (such as the db4 wavelet) and Butterworth bandpass filtering to eliminate 50Hz power frequency noise and high-frequency interference. The time decoding module uses a carrier phase smoothing pseudorange algorithm (Hatch filtering) combined with Beidou PPS+ToD messages, integrating satellite clock errors, ionospheric (Klobuchar model) and tropospheric (Saastamoinen model) delay corrections to generate microsecond-level absolute time scales. The transformation operation module uses HHT time-frequency analysis to extract wave head time, polarity, and main frequency characteristics, and combines an improved Canny operator to detect mutation points. The data formatting module uses synchronous coding index compression to store traveling wave time data pairs, raw signals, and IMF components. The fault diagnosis module constructs multi-dimensional criteria, cross-validating the initial judgment based on amplitude-attenuation rate and the secondary judgment based on wave head-frequency-polarity to accurately identify types such as lightning strikes, short circuits, and station faults, thereby improving the accuracy of fault identification.

[0020] In a third aspect, the present invention further provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement any of the methods described in the above solutions.

[0021] In a fourth aspect, the present invention also provides a storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored. The at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the power cable traveling wave fault detection method as in the above-mentioned scheme.

[0022] In summary, the beneficial technical effects of the present invention are: Based on the rapid propagation characteristics of traveling wave signals and the high-speed sampling and processing mechanism, it is possible to sense cable faults within milliseconds, promptly feedback fault information to power grid operators, and shorten the response time for fault handling.

[0023] By analyzing and calculating the characteristics of the collected signals, the location of the cable fault can be accurately determined, which helps operation and maintenance personnel quickly reach the fault site, improve fault handling efficiency, and reduce the scope of power outage impact.

[0024] It can effectively identify various types of cable faults such as short circuit, disconnection, and grounding, comprehensively ensuring the safe and stable operation of the power grid, and providing richer and more accurate information for the formulation of power grid protection strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flowchart of a fault detection method according to an embodiment of the present application.

[0026] Figure 2 is a flowchart of a fault detection method according to an embodiment of the present application.

[0027] Figure 3 is a flowchart of a fault detection method according to an embodiment of the present application.

[0028] Figure 4 is a structural diagram of a fault detection system according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The present application is further described in detail below with reference to the accompanying drawings. EMBODIMENT

[0030] Reference Figure 1 A fault detection method for power cable traveling wave is disclosed, comprising: S1: collecting and decomposing original signals of the power cable to obtain traveling wave sampling signals and record corresponding sampling times; S2: analyzing and correcting clock data transmitted by a Beidou time service module to obtain an absolute time scale; S3: associating and storing the traveling wave sampling signals with the absolute time scale according to the sampling times to generate a traveling wave time data pair; S4: extracting key feature parameters based on the traveling wave time data pair to generate a time domain feature map and a polar coordinate scatter plot; S5: analyzing the time domain feature map and the polar coordinate scatter plot according to a preset fault determination threshold interval to determine a fault type of the traveling wave sampling signals.

[0031] The implementation principle of the embodiment is: the original signal of the cable is acquired through ADC synchronous sampling (PLL phase-locked ± 50 ns), and the traveling wave signal is extracted through improved CEEMDAN decomposition (adding adaptive white noise) and wavelet threshold filtering (SUREShrink strategy); at the same time, the 1PPS+GPRMC message output by the Beidou / GPS dual-mode receiver is used to generate an absolute time scale through carrier smoothing pseudo-range solution (ionosphere dual-frequency correction), and the data pair is formed by the FPGA timestamp insertion unit in association with the traveling wave signal; in the feature extraction stage, Hilbert transform (instantaneous frequency calculation) and Teager energy operator (wave head mutation detection) are jointly applied to generate a time domain feature map, a polar coordinate scatter plot (amplitude-phase-primary frequency three-dimensional mapping) is constructed through Morlet wavelet time-frequency analysis, and wavelet packet energy entropy (frequency band energy proportion entropy value) and IMF kurtosis (impact component statistics) are added to form a four-dimensional feature set; the dynamic threshold mechanism-LSTM network is used in the fault judgment module to predict the threshold offset according to the historical working conditions (temperature, load, insulation degradation data) of the cable, and multi-level decision is made in combination with the preset threshold interval: lightning fault, short circuit fault, insulation deterioration, so as to realize a fault recognition rate of 99.5% and a misjudgment rate of <0.3%. Embodiment

[0032] With reference to Figure 2 , the step S1 comprises: S11: synchronizing the internal clock of all power collectors according to the reference clock to obtain a unified clock; S12: controlling the power collector to collect signals of the power cable based on the unified clock to obtain an original signal; S13: filtering out power frequency noise from the original signal to obtain a traveling wave sampling signal; S14: performing empirical mode decomposition on the traveling wave sampling signal to obtain an intrinsic mode signal; S15: performing Hilbert transform on the intrinsic mode signal to calculate the instantaneous frequency; S16: identifying the instantaneous frequency according to the time-frequency plane to obtain the traveling wave head; S17: detecting the second derivative extreme point of the traveling wave head according to the scale function to determine the sampling time.

[0033] The implementation principle of the embodiment is that: a redundant time source is formed by a Beidou / GPS dual-mode timing module and a constant temperature crystal oscillator time-keeping clock, and unified synchronization of all collectors with a clock deviation of ≤±1 ms is realized by combining a master station cascaded synchronization protocol (transmission delay is calculated based on T1-T4 time stamp); after triggering a signal based on the unified clock, power frequency noise is eliminated by using an adaptive notch filter, and a pure intrinsic modal component is extracted by joint processing of wavelet denoising and empirical mode decomposition (EMD); the component signal is subjected to Hilbert transform to generate an instantaneous frequency, a sliding time window energy entropy is used to identify a mutation point for preliminary screening of a wave head, and a scale adaptive Morlet wave analysis is used to accurately locate a sampling time at a second derivative extreme point, and finally anti-interference traveling wave head time stamp data is output, supporting high-precision fault diagnosis. Embodiment

[0034] Reference Figure 3 The step S2 comprises: S21: raw messages transmitted by the Beidou timing module are analyzed and separated to obtain a transmission time, a Beidou time, a leap second parameter, a satellite clock difference and a receiving time; S22: a receiving delay clock difference is calculated according to a time difference between the receiving time and the transmission time in combination with the speed of light; S23: an ionospheric delay clock difference is calculated according to a satellite elevation angle and ionospheric parameters in combination with local time; S24: a tropospheric delay clock difference is calculated according to a receiving elevation and temperature and humidity in combination with a refraction number; S25: a correction factor is obtained by performing a weighted operation on the satellite clock difference, the ionospheric delay clock difference, the tropospheric delay clock difference and the receiving delay clock difference according to a relativistic effect; S26: the Beidou time is converted to obtain coordinated universal time according to the leap second parameter in combination with the correction factor; S27: the coordinated universal time is time-stamped with the receiving time, and absolute time is generated in combination with carrier phase smoothing pseudorange.

[0035] The implementation principle of the embodiment is that: the transmission time, the satellite clock difference and the leap second parameter are obtained by analyzing the Beidou raw message, the speed of light transmission delay is calculated in combination with the receiving time; the ionospheric delay is calculated by using the satellite elevation angle / ionospheric parameters, and the tropospheric delay is calculated based on the elevation / temperature and humidity; the orbit error is compensated by introducing precise ephemeris, and the multipath effect is suppressed by using carrier-to-noise ratio weighting; the four types of delay error are weighted and corrected by introducing the relativistic effect to generate a correction factor; after the Beidou time is converted to coordinated universal time (UTC), the receiver noise is eliminated by using carrier phase smoothing pseudorange technology, and finally the time-keeping clock after temperature compensation is aligned to generate a nanosecond-level absolute time, and a fault detection mechanism is used to ensure the continuity of the timing. Embodiment

[0036] The step S3 comprises: According to the second pulse signal transmitted by the Beidou timing module combined with the phase-locked loop, the sampling time is synchronized with the absolute time scale to determine the absolute time, and a synchronization code is marked; According to the synchronization code, the traveling wave sampling signal corresponding to the sampling time is associated with the absolute time to obtain an initial traveling wave correlation pair; The delay time of the trigger point from the second pulse signal to the traveling wave sampling signal is calculated to obtain an actual time difference; According to the actual time difference and the refresh rate per cycle, the initial traveling wave correlation pair is corrected to obtain a traveling wave time data pair; According to a preset data structure, the traveling wave time data pair, the original signal and the inherent modal signal are combined with the synchronization code for compressed storage.

[0037] The implementation principle of the embodiment is that: a synchronization clock is generated by driving a phase-locked loop by a Beidou second pulse signal, and the time reference reliability is ensured by combining a rubidium atomic clock timekeeping module; a synchronization code with CRC check is generated by using FPGA to calculate the transmission delay of the second pulse to the traveling wave trigger point in real time and dynamically temperature compensation; the corrected traveling wave time data pair, the original signal and the inherent modal signal are stored according to a hierarchical compression strategy (differential encoding + wavelet compression); at the same time, the NTP protocol is fused to enhance the time scale robustness, and finally a traveling wave monitoring synchronization storage system with nanosecond-level time precision, anti-interference clock redundancy and verifiable data integrity is formed. Embodiment

[0038] The step S4 comprises: According to the time sequence, the original signal and the traveling wave time data pair are aligned to obtain a traveling wave start timestamp, a peak value timestamp, a traveling wave end timestamp and a normal current signal amplitude; According to the absolute time, the traveling wave sampling signal in the traveling wave time data pair is feature extracted to obtain a current traveling wave peak value; According to the current difference between the current traveling wave peak value and the normal current signal amplitude, combined with the time difference between the traveling wave start timestamp and the peak value timestamp, the change rate of the traveling wave sampling signal is calculated to obtain an amplitude change rate; The traveling wave signal is analyzed in the frequency domain to obtain a plurality of traveling wave frequencies; According to the peak value timestamp and the current traveling wave peak value, combined with the amplitude change rate, a time domain feature map is generated; According to all the traveling wave frequencies combined with the traveling wave end timestamp, a polar coordinate scatter plot is generated.

[0039] The implementation principle of the embodiment is: after wavelet denoising preprocessing of the original current signal, the Teager energy operator is used to accurately locate the start and end time stamps of the traveling wave; the peak value of the traveling wave is extracted by the absolute time reference, and the amplitude variation rate is calculated in combination with the normal current amplitude; the traveling wave main frequency component is obtained by synchronously using the complex modulation refinement analysis method to eliminate spectral aliasing; the time-amplitude correlation graph is generated by fusing the time domain features (peak value / variation rate), and the polar coordinate scatter plot is constructed in combination with the frequency domain features; finally, the SVM classifier is used to identify the fault type according to the time-frequency features, and the traveling wave propagation path is reconstructed based on the multi-terminal time synchronization data, forming a closed-loop analysis system from signal processing to fault diagnosis. EMBODIMENT

[0040] The step S5 comprises: The zero-mode maximum value is obtained by decoupling and mode extreme value transformation on the traveling wave sampling signal, and the accumulated maximum amplitude is obtained by accumulation in a preset period; The traveling wave amplitude measure is obtained by ratio calculation according to the zero-mode maximum value and the maximum value in all the zero-mode maximum values; The feature parameters are obtained by feature extraction on the time domain feature graph and the polar coordinate scatter plot; the feature parameters include the amplitude decay rate; The current traveling wave peak value is compared with the amplitude decay rate and the preset fault determination threshold interval to determine the fault type of the current traveling wave sampling signal; If the current traveling wave peak value is located in the first determination threshold interval, and the amplitude decay rate is located in the first decay slope interval, it is determined that the fault type of the current traveling wave sampling signal is lightning fault; If the current traveling wave peak value is located in the second determination threshold interval, and the amplitude decay rate is located in the second decay slope interval, it is determined that the fault type of the current traveling wave sampling signal is short-circuit fault; If the current traveling wave peak value is located in the third determination threshold interval and periodically fluctuates, it is determined that the fault type of the current traveling wave sampling signal is in-station fault; If the current traveling wave peak value is located in the fourth determination threshold interval and fluctuates in a preset period, it is determined that the fault type of the current traveling wave sampling signal is broken line fault; If the accumulated maximum amplitude of the current traveling wave peak value is greater than or equal to the traveling wave amplitude measure, it is determined that the fault type of the current traveling wave sampling signal is ground fault; In the embodiment, the determination threshold interval is: The first interval (lightning): [0.85, 1.0], the second interval (short circuit): [0.60, 0.85] The third interval (in-station fault): [0.30, 0.60], the fourth interval (broken line): [0.10, 0.30]; The decay slope interval (unit: dB / ms): First interval (lightning strike): [-50, -30] (fast decay), second interval (short circuit): [-30, -10] (medium decay); Wave period definition: Station fault periodic fluctuation: frequency 100-500Hz (power frequency harmonic), broken line fault fluctuation period: 0.5-2ms (mechanical vibration characteristic); According to the characteristic parameters, the fault type of the traveling wave sampling signal is determined again; the characteristic parameters include wave head time, polarity and main frequency; If the wave head time is located in a preset first duration threshold interval, the main frequency is located in a preset first frequency threshold interval, and the polarity is positive, it is determined that the fault type of the current traveling wave sampling signal is lightning strike fault; If the wave head time is located in a preset second duration threshold interval, the main frequency is located in a preset second frequency threshold interval, and the polarity is negative and bidirectional alternation, it is determined that the fault type of the current traveling wave sampling signal is short circuit fault; If the wave head time is located in a preset third duration threshold interval, the main frequency is located in a preset third frequency threshold interval, it is determined that the fault type of the current traveling wave sampling signal is station fault; If the wave head time is located in a preset fourth duration threshold interval, the main frequency is located in a preset fourth frequency threshold interval, it is determined that the fault type of the current traveling wave sampling signal is broken line fault; If the wave head time is located in a preset fifth duration threshold interval, the main frequency is located in a preset fifth frequency threshold interval, and the polarity is negative, it is determined that the fault type of the current traveling wave sampling signal is ground fault; In this embodiment, Duration threshold (μs): Lightning strike: [1, 5], short circuit: [5, 20], station fault: [20, 100], broken line: [100, 500], ground: [500, 1000]; Main frequency threshold (kHz): Lightning strike: [1, 10], short circuit: [10, 50], station fault: [0.05, 1], broken line: [0.5, 5], ground: [0.01, 0.5]; If the fault type determination of the same traveling wave sampling signal is the same type, it indicates that the fault type determination of this traveling wave sampling signal is correct; otherwise, the second determination result is taken as the fault type of this traveling wave sampling signal; In this embodiment, the following processing steps can also be selected: If the two results are inconsistent, start the third-order verification: check whether the polarity meets the bidirectional alternation (short circuit) or unidirectional (lightning / grounding); verify whether the cumulative maximum amplitude continues to exceed the traveling wave amplitude measure (ground fault core feature); call the historical similar fault feature library for similarity matching (confidence > 90% to cover the original result).

[0041] The implementation principle of the embodiment is: first, the traveling wave signal is subjected to phase-mode transformation decoupling to extract the zero-mode component, a zero-mode maximum value sequence is generated by a wavelet transformation mode maximum value method, and a cumulative maximum amplitude is accumulated in a 10 ms period; a ratio of the cumulative maximum amplitude to a global maximum value is calculated as a traveling wave amplitude measure, combined with a preset quantization amplitude interval (lightning [0.85, 1.0], short circuit [0.60, 0.85], etc.) and a decay slope (lightning [-50, -30] dB / ms) for preliminary judgment; at the same time, a wave head time (lightning 1-5 μs), a main frequency (short circuit 10-50 kHz), a polarity and other parameters are extracted from a time domain graph and a polar coordinate graph for secondary determination; if the results conflict, the third-order verification (polarity / amplitude persistence / history matching) is used for arbitration, and finally the fault type is output, and the influence of environmental temperature drift on the frequency domain threshold is dynamically corrected, achieving a classification accuracy of more than 95%.

[0042] Referring to Figure 4 A power cable traveling wave fault detection system applied to the fault detection method, comprising: A synchronous sampling module for controlling an ADC to synchronously collect original signals in a power cable line; A filtering processing module for denoising, filtering and normalizing the original signals to obtain traveling wave sampling signals; A time decoding module for analyzing time information output by a Beidou timing module to obtain an absolute time scale and associating the traveling wave sampling signals to generate a traveling wave time data pair; A transformation operation module for extracting feature parameters including a wave head time, a polarity and a main frequency from the traveling wave time data pair; A data formatting module for compressively storing the traveling wave time data pair, the original signals and the inherent modal signals according to synchronous encoding; A fault determination module for determining a fault type of the traveling wave sampling signals according to the feature parameters combined with a preset fault determination threshold interval.

[0043] In this embodiment, the synchronous sampling module is used to control the ADC to perform synchronous operation, so as to ensure that the cable signal is collected at the same time and accurate original signal data is obtained. The time decoding module is used to analyze the time information output by the Beidou timing module, so as to embed the accurate absolute time scale into the value obtained by the ADC sampling, and realize accurate association of the signal and the time. The high-pass / low-pass filtering module is used to make the input high-frequency signal pass through a 1KHz high-pass filter first, remove the low-frequency interference, and then pass through a 5MHz low-pass filter, so as to filter out the effective high-frequency signal and input the high-frequency signal into the ADC, thereby improving the purity and effectiveness of the sampling signal. The transform operation module is used to calculate the power frequency or high-frequency quantity based on the low-pass / high-pass filtered data, so as to extract the signal characteristic parameters and provide data support for fault judgment. The data formatting module is used to store the sampling data, the high-frequency / power frequency quantity data after filtering processing and the time data into the FIFO according to a specific format, so as to facilitate subsequent data reading and analysis.

[0044] The device continuously samples the cable signal at a high speed and cyclically writes and records. When an abnormal waveform is monitored, the sampling recording program is triggered immediately, and the abnormal waveform is time-stamped. Before the next cyclic writing and recording starts, the fault data found this time is transferred from the storage area to the data cache and uploaded to the upper system, so as to prevent the fault data from being lost due to cyclic covering.

[0045] The implementation principle of the embodiment is as follows: a Beidou timing module generates a microsecond time scale, and a plurality of ADCs are triggered to synchronously collect the cable signal at a rate of 1MHz; after cascaded filtering (power frequency trap wave + wavelet denoising) and dynamic normalization, a traveling wave sampling signal is generated; the zero mode component is decoupled by Karenbauer transform, and characteristic parameters such as wave head time (1-1000μs), polarity, main frequency (0.01-50kHz) are extracted; preliminary judgment is performed in combination with preset amplitude intervals (lightning [0.85, 1.0]), attenuation slopes (short circuit [-30, -10]dB / ms) and wave head characteristics, and final judgment of the fault type is performed in combination with time-frequency domain parameters; the data is mixed and compressed to store the original signal and the IMF component with the time stamp as the index; when a judgment conflict occurs, the confidence level is improved based on the DTW similarity matching of the historical fault library, and finally the fault classification result with the confidence level is output. Embodiment

[0046] An electronic device, comprising: one or more processors; a memory storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method in any of the above solutions. Embodiment

[0047] A storage medium, the storage medium storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by a processor to implement the fault detection method of the power cable traveling wave in the above solution The embodiments of the present specific implementation are the preferred embodiments of the present application, and are not intended to limit the protection scope of the present application, so that: any equivalent changes made according to the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method for fault detection of power cable traveling waves, characterized in that: include: Collect and decompose the original signal of the power cable to obtain the traveling wave sampling signal and record the corresponding sampling time; Analyze and correct the clock data transmitted by the Beidou timing module to obtain the absolute time scale; Associating and storing the traveling wave sampling signal with the absolute time scale according to the sampling time to generate a traveling wave time data pair; Extract key characteristic parameters based on the traveling wave time data pair, and generate a time domain characteristic graph and a polar coordinate scatter plot; The time domain characteristic diagram and the polar coordinate scatter diagram are analyzed according to a preset fault determination threshold interval to determine the fault type of the traveling wave sampling signal.

2. The power cable traveling wave fault detection method according to claim 1, characterized in that: The specific steps of collecting and decomposing the original power cable signal to obtain the traveling wave sampling signal and recording the sampling time include: Synchronize the internal clocks of all power harvesters according to the reference clock to obtain a unified clock; Based on the unified clock, the power collector is controlled to collect signals from the power cable to obtain original signals; Performing power frequency noise filtering on the original signal to obtain a traveling wave sampling signal; Performing empirical mode decomposition on the traveling wave sampling signal to obtain an intrinsic mode signal; Performing Hilbert transform on the natural mode signal to calculate the instantaneous frequency; Identifying the sudden change point of the instantaneous frequency according to the time-frequency plane to obtain the traveling wave head; The second-order derivative extreme point of the traveling wave head is detected according to the scaling function to determine the sampling time.

3. The power cable traveling wave fault detection method according to claim 1, characterized in that: The specific steps of analyzing and correcting the clock data transmitted by the Beidou timing module to obtain the absolute time scale include: Parse and separate the original message transmitted by the Beidou timing module to obtain the transmission time, Beidou time, leap second parameter, satellite clock error and reception time; Calculate the receiving delay clock error based on the time difference between the receiving time and the transmitting time in combination with the speed of light; Calculate the ionization delay clock error based on the satellite elevation angle and ionospheric parameters combined with local time; Calculate the convective clock difference based on the receiving elevation, temperature and humidity combined with the refraction number; performing a weighted operation on the satellite clock error, the ionization delay clock error, the convolution clock error, and the receiving delay clock error according to the relativistic effect to obtain a correction factor; Converting the Beidou time according to the leap second parameter combined with the correction factor to obtain Coordinated Universal Time; The coordinated universal time is time-stamp aligned with the receiving time, and combined with the carrier phase smoothed pseudorange to generate an absolute time scale.

4. The method for detecting faults of power cable traveling waves according to claim 1, characterized in that: The specific steps of associating and storing the traveling wave sampling signal with the absolute time scale according to the sampling time to generate a traveling wave time data pair include: According to the second pulse signal transmitted by the Beidou timing module and combined with the phase-locked loop, the sampling time is synchronized with the absolute time scale to determine the absolute time and mark the synchronization code; Correlating the traveling wave sampling signal corresponding to the sampling time with the absolute time according to the synchronous coding to obtain an initial traveling wave correlation pair; Calculating the delay time from the pulse-per-second signal to the trigger point of the traveling wave sampling signal to obtain an actual time difference; Correcting the initial traveling wave correlation pair according to the actual time difference and the refresh rate per cycle to obtain a traveling wave time data pair; The traveling wave time data pair, the original signal and the intrinsic modal signal are combined with the synchronous coding for compression storage according to a preset data structure.

5. The power cable traveling wave fault detection method according to claim 1, characterized in that: The specific steps of extracting key characteristic parameters based on the traveling wave time data and generating a time domain characteristic graph and a polar coordinate scatter plot include: Align the original signal with the traveling wave time data pair according to the time series to obtain the traveling wave start timestamp, peak timestamp, traveling wave end timestamp and normal current signal amplitude; Extract features of the traveling wave sampling signal in the traveling wave time data pair according to the absolute time to obtain the current traveling wave peak value; Calculating the rate of change of the traveling wave sampling signal based on the current difference between the current traveling wave peak value and the normal current signal amplitude, combined with the time difference between the traveling wave start timestamp and the peak timestamp, to obtain the amplitude change rate; Performing frequency domain analysis on the traveling wave signal to obtain a plurality of traveling wave frequencies; Generate a time domain characteristic diagram according to the peak timestamp and the current traveling wave peak value in combination with the amplitude change rate; A polar coordinate scatter plot is generated according to all the traveling wave frequencies combined with the traveling wave end timestamp.

6. The method for detecting faults of a power cable traveling wave according to claim 1, characterized in that: The specific steps of analyzing the time domain characteristic diagram and the polar coordinate scatter diagram according to the preset fault determination threshold interval to determine the fault type of the traveling wave sampling signal include: Decoupling and modulus extreme value transformation are performed on the traveling wave sampling signal to obtain the zero modulus maximum value, and accumulation is performed within a preset period to obtain the accumulated maximum amplitude value; Calculating the ratio of the zero-mode maximum value to the maximum value among all the zero-mode maxima to obtain a traveling wave amplitude measure; Extracting features from the time domain feature graph and the polar coordinate scatter plot to obtain feature parameters; the feature parameters include amplitude decay rate; Compare and judge the current traveling wave peak value in combination with the amplitude decay speed with the preset fault judgment threshold range; If the current traveling wave peak value is within the first determination threshold interval, and the amplitude attenuation rate is within the first attenuation slope interval, then the fault type of the current traveling wave sampling signal is determined to be a lightning fault; If the current traveling wave peak value is within the second determination threshold interval, and the amplitude attenuation speed is within the second attenuation slope interval, then the fault type of the current traveling wave sampling signal is determined to be a short circuit fault; If the current traveling wave peak value is within the third determination threshold interval and fluctuates periodically, it is determined that the fault type of the current traveling wave sampling signal is an internal station fault; If the current traveling wave peak value is within the fourth determination threshold interval and fluctuates within a preset period, it is determined that the fault type of the current traveling wave sampling signal is a line break fault; If the accumulated maximum amplitude of the current traveling wave peak value is greater than or equal to the traveling wave amplitude measure, it is determined that the fault type of the current traveling wave sampling signal is a ground fault.

7. The method for detecting faults of a power cable traveling wave according to claim 1, characterized in that: The specific step of analyzing the time domain characteristic diagram and the polar coordinate scatter diagram according to the preset fault determination threshold interval to determine the fault type of the traveling wave sampling signal further includes: Re-determine the fault type of the traveling wave sampling signal based on characteristic parameters; the characteristic parameters include wave head time, polarity and main frequency; If the wave front time is within a preset first duration threshold interval, the main frequency is within a preset first frequency threshold interval, and the polarity is positive, then it is determined that the fault type of the current traveling wave sampling signal is a lightning fault; If the wave front time is within the preset second duration threshold interval, the main frequency is within the preset second frequency threshold interval, and the polarity is alternating between negative and bidirectional, then it is determined that the fault type of the current traveling wave sampling signal is a short circuit fault; If the wave crest time is within a preset third duration threshold interval, and the main frequency is within a preset third frequency threshold interval, then it is determined that the fault type of the current traveling wave sampling signal is an internal station fault; If the wave crest time is within a preset fourth duration threshold interval, and the main frequency is within a preset fourth frequency threshold interval, then it is determined that the fault type of the current traveling wave sampling signal is a line break fault; If the wave front time is within the preset fifth duration threshold interval, the main frequency is within the preset fifth frequency threshold interval, and the polarity is negative, then it is determined that the fault type of the current traveling wave sampling signal is a ground fault; If the fault types of the same traveling wave sampling signal are all determined to be the same type, it indicates that the fault type of this traveling wave sampling signal is correctly determined; otherwise, the second determination result is used as the fault type of this traveling wave sampling signal.

8. A power cable traveling wave fault detection system applied to the method according to any one of claims 1 to 7, characterized in that: include: Synchronous sampling module, used to control ADC to synchronously collect the original signal in the power cable line; A filtering processing module, configured to perform denoising, filtering and normalization processing on the original signal to obtain a traveling wave sampling signal; A time decoding module is used to parse the time information output by the Beidou timing module to obtain an absolute time scale, and associate it with the traveling wave sampling signal to generate a traveling wave time data pair; A transformation operation module is used to extract characteristic parameters including wave head time, polarity, and main frequency from the traveling wave time data pair; a data formatting module, configured to compress and store the traveling wave time data pairs, the original signal, and the intrinsic mode signal according to synchronous coding; The fault determination module is used to determine the fault type of the traveling wave sampling signal according to the characteristic parameters in combination with a preset fault determination threshold range.

9. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A storage medium, wherein at least one instruction, at least one program, a code set or an instruction set is stored in the storage medium, wherein the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the power cable traveling wave fault detection method as described in claims 1 to 7.

Images