A continuous chirp wave coherent optical time domain reflection cable partial discharge detection system and method

Abstract

The application discloses a continuous chirp wave coherent optical time domain reflection cable partial discharge detection system and method. Single-frequency continuous light output by a narrow linewidth laser is modulated into single-sideband continuous chirp detection light by an IQ-EOM, and after amplification, the detection light is injected into a common single-mode sensing optical fiber along a cable line. Backscattered Rayleigh scattering light is coherent beat with local light and collected by a data acquisition card. A computer extracts multiple sub-chirp beat signals from a complete chirp period, respectively performs optical field reconstruction and matched filtering, rotates corresponding complex results by using the normalized conjugate of the matched filtering complex results of each frequency interval, sums the rotated complex results of different frequency intervals to complete phase demodulation, and rearranges and splices in time according to the actual detection time, so that the equivalent sampling rate is improved, the long-distance detection bandwidth is expanded, and the coherent fading is suppressed. The system can realize continuous distributed detection and positioning of cable partial discharge acoustic signals.

Description

Technical Field

[0001] This invention belongs to the field of power cable insulation condition monitoring, distributed optical fiber sensing and partial discharge detection technology, specifically relating to a continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system and method. Background Technology

[0002] Power cables are crucial equipment in power transmission and distribution systems. During manufacturing, laying, and long-term operation, cables may experience localized electric field distortions due to factors such as insulation aging, mechanical damage, moisture, and air gaps, leading to partial discharge. Partial discharge is an early sign of insulation degradation and fault development; if not detected in time, it can further cause breakdown and power outages. In practical partial discharge detection, offline detection typically requires power outages, making long-term continuous monitoring difficult. Live-line inspection and point-based online monitoring are limited by inspection cycles, installation locations, and sensor coverage, making continuous detection of long-distance cable lines challenging, and also present problems such as complex deployment, high maintenance costs, and a high risk of missed detections. Rayleigh scattering-based phase-sensitive optical time-domain reflectometry (TDR) technology can utilize the phase changes of backscattered Rayleigh light along the sensing fiber to achieve distributed detection and localization of acoustic and vibration signals, offering advantages such as passive operation, resistance to electromagnetic interference, and long-distance continuous coverage. However, traditional phase-sensitive optical time-domain reflectometry (TDR) employs a single-frequency narrow-pulse probe, whose detection bandwidth is limited by the fiber round-trip time and pulse repetition frequency. Furthermore, sensing distance, spatial resolution, signal-to-noise ratio, and detection sensitivity are mutually constrained, and the technology is susceptible to coherent fading, making it difficult to accurately detect weak partial discharge acoustic signals over long distances. To overcome this bandwidth limitation over long distances, this invention uses a continuous chirped wave as the probe. By segmenting and temporally reconstructing sub-chirped signals of different frequencies within a single fiber round-trip cycle, a higher equivalent sampling rate and detection bandwidth can be achieved without increasing the overall probe repetition frequency. Simultaneously, by utilizing the complementary characteristics of Rayleigh scattering signals at different frequencies for coherent superposition, coherent fading can be suppressed, improving the detection sensitivity and reliability of weak partial discharge signals.

[0003] A search revealed Chinese patent CN121613267A, which discloses a chirped distributed acoustic wave sensing device and a cable abnormal discharge monitoring device using the same. This scheme injects linearly frequency-modulated chirped detection light into the sensing fiber and performs coherent reception and pulse compression on the returned scattered light to achieve distributed localization and detection of cable abnormal discharge acoustic signals. Simultaneously, it employs scattering-enhancing fiber to improve backscattered light intensity, detection sensitivity, and signal-to-noise ratio. The main feature of this scheme is the combination of chirped pulse compression and scattering-enhancing fiber to balance detection energy, spatial resolution, and detection sensitivity. However, it still uses the chirped pulse as the basic detection unit, and the dynamic sampling frequency is limited by the pulse repetition period and sensing distance, without further proposing a method to improve the equivalent sampling rate. Furthermore, this scheme relies on scattering-enhancing fiber with a specific microstructure, which is not conducive to achieving universality using ordinary communication fiber. Therefore, the accuracy of this type of scheme in reconstructing high-frequency acoustic signals of partial discharge under long-distance sensing conditions remains limited.

[0004] A search revealed Chinese patent CN121933776A, which discloses an online monitoring probe, system, and method for partial discharge in cable joints. This scheme involves multiple strip-shaped fiber optic probes connected in series on the transmission optical fiber and mounted on the surface of the cable joint's cold-shrink sleeve, using frequency-division multiplexed chirped pulse light for detection. The receiving end uses matched filtering to achieve pulse compression and frequency channel separation, sets distance gates according to the preset positions of each probe, and then reconstructs the partial discharge ultrasonic signal through orthogonal demodulation and polarization synthesis. This scheme can improve the detection signal-to-noise ratio and spatial resolution, and reduce the impact of polarization state changes on monitoring stability. However, this scheme still uses chirped pulses as the basic detection unit, and frequency division multiplexing is mainly used to separate frequency channels with overlapping time domains. It does not disclose methods for using different sub-chirps to perform staggered sampling of the same spatial location and for reconstructing the signal according to the actual detection time. Therefore, it does not further address the problem of limited bandwidth for detecting high-frequency acoustic signals of partial discharge under long-distance sensing conditions. Furthermore, its polarization diversity reception is mainly used to suppress polarization fading, but it is still difficult to eliminate the local low signal-to-noise ratio locations formed by Rayleigh scattering coherent fading; at the same time, this scheme relies on pre-set discrete strip probes, acoustic coupling layers and distance gates, and can only perform quasi-distributed monitoring of known cable joint locations.

[0005] Therefore, existing technologies still lack a continuous chirped wave distributed partial discharge detection method suitable for long-distance cable lines. This method should be able to achieve continuous coverage using ordinary single-mode optical fiber, and improve the equivalent sampling rate through sub-chirped extraction, coherent superposition, and time-series reconstruction to overcome the problem of limited detection bandwidth under long-distance conditions. Simultaneously, it should suppress Rayleigh scattering coherent fading, achieving accurate detection and localization of weak partial discharge high-frequency acoustic vibration signals. Summary of the Invention

[0006] This invention uses the acoustic vibration signal generated by partial discharge in cable joints or the cable itself as the measurement object, and provides a continuous chirped wave coherent optical time-domain reflectometry (TDDR) cable partial discharge detection system and method. This addresses the problems of existing chirped pulse schemes, such as insufficient equivalent sampling rate under long-distance conditions, limited ability to reconstruct high-frequency partial discharge acoustic signals, significant coherent fading effects, and reliance on dedicated reinforcing optical fibers or discrete probes. The objective of this invention can be achieved through the following technical solutions.

[0007] The aforementioned continuous chirped wave coherent optical time-domain reflectometry (CTDR) cable partial discharge detection system and method includes a narrow linewidth laser, an optical isolator, a first fiber coupler, an electro-optic modulator (IQ-EOM), an IQ-EOM driver, a first arbitrary waveform generator (AWG1), an erbium-doped fiber amplifier (EDFA), an optical circulator, a sensing fiber, a second fiber coupler, a balanced photodetector (BPD), a data acquisition card (DAQ), a second arbitrary waveform generator (AWG2), and a computer.

[0008] The narrow linewidth laser is used to output single-frequency continuous light with a center operating wavelength of around 1550nm, providing a coherent light source for continuous chirped modulation, coherent beat frequency, and backscattering Rayleigh phase detection.

[0009] The optical isolator is used to suppress reflected light from the subsequent optical path from returning to the narrow-linewidth laser, thereby reducing the impact of optical feedback on the laser output frequency, linewidth, and power stability.

[0010] The first fiber coupler is used to split single-frequency continuous light into probe light and local oscillator light. Preferably, the splitting ratio of the first fiber coupler is 95:5, wherein the light with 95% optical power is used as probe light and the light with 5% optical power is used as local oscillator light.

[0011] The first arbitrary waveform generator AWG1 is used to generate an in-phase I-channel sweep frequency signal and a quadrature Q-channel sweep frequency signal. The two linear sweep frequency signals have the same sweep period T, start frequency f0, and cutoff frequency f1, with a sweep bandwidth of B = f1 - f0 and a sweep slope of K = B / T. The initial phase of the I-channel sweep frequency signal is 0°, and the phase difference between the Q-channel sweep frequency signal and the I-channel sweep frequency signal is +90° or -90°. Within one sweep period, the instantaneous frequency of the linear sweep frequency signal can be expressed as: Where t is the time within a complete chirp cycle. The complete chirp cycle is determined based on the target sensing distance L and satisfies: Where, n eff denoted as the effective refractive index of the sensing fiber, and c as the speed of light in a vacuum.

[0012] The IQ-EOM driver is used to amplify the amplitude of the in-phase I-channel swept frequency electrical signal and the quadrature Q-channel swept frequency electrical signal generated by the first arbitrary waveform generator AWG1.

[0013] The electro-optic modulator IQ-EOM is used to perform in-phase and quadrature modulation of single-frequency continuous light under the drive of the I-channel and Q-channel swept frequency electrical signals to obtain single-sideband continuous chirped probe light. Preferably, the two sub-Mach-Zehnder modulators of the electro-optic modulator IQ-EOM are biased at the minimum transmission point, and the main Mach-Zehnder modulator is biased at the quadrature point; by setting the phase difference between the Q-channel swept frequency electrical signal and the I-channel swept frequency electrical signal to +90° or −90°, the upper sideband or the lower sideband is selected as the probe light, and the optical carrier suppression degree and the non-target sideband suppression degree are both not less than 20dB. When the lower sideband is selected as the probe light, the complex optical field of the single-sideband continuous chirped probe light can be expressed as: Among them, A p To detect the light amplitude, f c This is the carrier frequency for the output light of a narrow linewidth laser.

[0014] The erbium-doped fiber amplifier (EDFA) is used to amplify the optical power of a single-sideband continuously chirped probe light to compensate for power loss caused by long-distance transmission and Rayleigh backscattering, and to improve the detectability of the Rayleigh backscattering signal at the far end of the sensing fiber.

[0015] The optical circulator is used to guide the amplified single-sideband continuously chirped probe light into the sensing fiber and to separate and output the back Rayleigh scattered light returned by the sensing fiber from the probe light propagation channel.

[0016] The sensing fiber is used to generate backscattered Rayleigh light distributed along the fiber and to sense acoustic disturbances caused by partial discharge in the cable joint or cable body. The partial discharge acoustic disturbances cause axial strain and optical path changes in the sensing fiber at the corresponding location, which in turn induces a phase change in the backscattered Rayleigh light. The Rayleigh scattering process within a spatial resolution cell is equivalent to M scattering points, where the distance from the m-th equivalent scattering point to the fiber input end is z. m The complex scattering coefficient is Γ m The total backscattered Rayleigh field obtained at the receiver can be expressed as: Each equivalent scattering point has a different round-trip propagation delay and random scattering phase; the distance position z on the sensing fiber and the round-trip propagation time τ satisfy the following:

[0017] The second fiber coupler is used to coherently superimpose the back Rayleigh scattered light with the local oscillator light to form a beat frequency optical signal, so as to convert the amplitude and phase information of the back Rayleigh scattered light into the beat frequency component between the local oscillator light and the scattered light.

[0018] The balanced photodetector (BPD) is used to convert the beat frequency optical signal into a beat frequency electrical signal and suppress common-mode noise in the two photocurrents. The effective beat frequency term in the beat frequency electrical signal photocurrent can be expressed as: Among them, E LO (t) represents the complex optical field of the local oscillator, and the superscript "*" indicates complex conjugate.

[0019] The second arbitrary waveform generator AWG2 is used to output a trigger signal synchronized with the complete chirped cycle, providing a unified time reference for determining the starting position of each complete chirped cycle, segmenting the beat frequency electrical signal, and subsequent timing rearrangement.

[0020] The data acquisition card DAQ is used to continuously acquire beat frequency electrical signals under the control of the synchronization trigger signal and transmit the acquired data to the computer.

[0021] The computer divides the broadband beat frequency electrical signal corresponding to a complete chirp cycle into N different frequency intervals, and uses a bandpass filter with the corresponding passband to extract N segments of sub-chirp beat frequency signals. The bandwidth B of the nth sub-chirp segment... n and the corresponding time window width T n They are respectively:

[0022] The computer performs a Hilbert transform on the beat frequency signal of the nth sub-chirp to construct an analytical signal, and reconstructs the backscattered Rayleigh light field corresponding to the sub-chirp based on the local oscillator light field. The computer uses a reference signal matching the nth sub-chirp for matched filtering to compress the long-duration chirp response into a narrow-pulse distance response. The matched filter output of the nth sub-chirp can be expressed as: Where, r n (t) represents the reconstructed light field of the nth segment, s n (t) represents the matched filter reference signal corresponding to the sub-chirp. For each sub-chirp frequency band, a complex reference signal with the same sweep direction and sweep slope as the sub-chirp is generated. The conjugate time-reversed signal of the complex reference signal is used as a matched filter to compress the long-duration sub-chirp response into a narrow pulse distance response. Its theoretical spatial resolution Δz is mainly determined by the sub-chirp bandwidth B. sub A decision can be expressed as:

[0023] The computer is used to perform phase alignment and coherent superposition of the matched filter complex results corresponding to different frequency sub-chirps. For each frequency interval, the matched filter complex result corresponding to a preset reference detection time is selected at each distance position in the sensing fiber as a reference complex result. The conjugate value of the reference complex result is taken and divided by its magnitude to obtain a normalized conjugate reference quantity with a unit magnitude. The normalized conjugate reference quantity is multiplied by the matched filter complex result corresponding to other detection times in the same frequency interval to obtain the rotated complex result after eliminating the initial reference phase. The rotated complex results corresponding to two or more different frequency intervals are summed. Then, the phase is extracted from the summed complex result, and initial phase elimination, spatial difference, and temporal difference are performed sequentially to obtain the dynamic phase result distributed along the sensing fiber. The representative frequency f of the nth sub-chirp is... n The corresponding time position t within the complete chirping cycle n satisfy:

[0024] The computer is used to rearrange and stitch together the time sequence according to the actual detection time corresponding to each effective phase sampling point, and to identify the phase anomaly location and recover the time-domain waveform and frequency characteristics of the partial discharge acoustic-vibration signal from the rearranged dynamic phase time sequence. Assuming that G equally timed effective phase sampling points are ultimately obtained within a complete chirp cycle, then the equivalent sampling period T is... e Equivalent sampling rate f s,eq and the theoretical detection frequency upper limit f max They are respectively: When each sub-chirp independently forms an effective phase sampling point, G=N; when every two sub-chirps are summed by a rotation vector to form an effective phase sampling point, G=N / 2. The computer also determines the location of the partial discharge on the sensing fiber based on the round-trip propagation delay corresponding to the matched filter output, and, combined with the laying correspondence between the sensing fiber and the cable under test, converts the fiber distance position into the actual location of the partial discharge on the cable under test.

[0025] Compared with the prior art, the present invention has the following beneficial effects: First, the present invention uses ordinary single-mode sensing optical fiber laid along the cable line to realize continuous distributed detection of partial discharge acoustic vibration signal. It does not require the use of scattering enhancement optical fiber with specific microstructure, nor does it require setting up special strip optical fiber probes, acoustic coupling layers and corresponding distance gates at each cable joint. This can reduce the manufacturing requirements of sensing optical fiber and detection probe, and simplify the on-site deployment and maintenance process. Second, this invention uses a continuous chirped wave as the probe light, dividing a complete chirp into multiple sub-chirped signals, and performing optical field reconstruction, matched filtering, and phase demodulation on each sub-chirp. By utilizing the time difference of different sub-chirs passing through the same spatial position, the phase results of each group are rearranged and spliced ​​according to the actual detection time. This can improve the equivalent sampling rate without increasing the repetition frequency of the complete chirp, and reduce the limitation of sensing distance and fiber round-trip propagation time on the detection bandwidth. Third, this invention utilizes the characteristic that the Rayleigh scattering fading positions corresponding to different frequency sub-chirps are different. It uses the normalized conjugate of the complex results of the reference matched filter in each frequency interval to rotate the complex results of the matched filter in the corresponding frequency interval, and sums the rotated complex results in different frequency intervals. Then, it extracts and demodulates the phase from the summation result, so that the effective Rayleigh scattering signals at different frequencies are coherently superimposed, reducing the local low signal-to-noise ratio, phase noise spikes and detection blind zones caused by coherent fading at a single frequency, and improving the detection sensitivity, waveform integrity and detection reliability of weak partial discharge signals. Fourth, this invention can form continuous coverage of long-distance cable lines using a single sensing optical fiber, and is not limited to the location of pre-installed discrete probes; by establishing the correspondence between propagation delay and optical fiber distance through matched filtering results, and combining the laying relationship between the sensing optical fiber and the cable line, long-distance detection, waveform reconstruction and spatial positioning of partial discharge acoustic vibration signals can be realized simultaneously. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the structure of a continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system according to the present invention.

[0027] Figure 2 This is a flowchart of the partial discharge signal processing procedure for a continuous chirped wave coherent optical time-domain reflectometry cable according to the present invention.

[0028] Among them, 101 is a narrow linewidth laser; 102 is an optical isolator; 103 is a first fiber coupler; 104 is an electro-optic modulator-EOM; 105 is an IQ-EOM driver; 106 is a first arbitrary waveform generator (AWG1); 107 is an erbium-doped fiber amplifier (EDFA); 108 is an optical circulator; 109 is a sensing fiber; 110 is a second fiber coupler; 111 is a balanced photodetector (BPD); 112 is a data acquisition card (DAQ); 113 is a second arbitrary waveform generator (AWG2); and 114 is a computer. Detailed Implementation

[0029] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Where there is no conflict, the technical features of the various embodiments of the present invention can be combined with each other.

[0030] Example 1: like Figure 1 As shown, this embodiment provides a continuous chirped wave coherent optical time-domain reflectometry (TDDR) cable partial discharge detection system. The connection method is as follows: the output end of the narrow linewidth laser (101) is connected to the first fiber coupler (103) via an optical isolator (102). The optical isolator (102) is used to suppress the reflected light in the subsequent optical path from returning to the narrow linewidth laser (101) to reduce the impact of reflected light on the stability of the light source. The first fiber coupler (103) splits the single-frequency continuous light into probe light and local oscillator light, wherein the first output end is connected to the optical input end of the electro-optic modulator IQ-EOM (104), and the second output end is connected to one input end of the second fiber coupler (110). The splitting ratio of the first fiber coupler (103) is 95:5, wherein 95% of the light is used as probe light and 5% of the light is used as local oscillator light. The first arbitrary waveform generator AWG1 (106) outputs two linear sweep frequency electrical signals with the same sweep frequency period and sweep frequency range and maintaining an orthogonal phase relationship. The two electrical signals are amplified by the IQ-EOM driver (105) and then loaded onto the in-phase modulation branch and the quadrature modulation branch of the electro-optic modulator IQ-EOM (104), respectively. By adjusting the amplitude and phase of the two driving signals and the bias voltage of the electro-optic modulator IQ-EOM (104), the optical carrier and non-target sidebands are suppressed, and the upper sideband or the lower sideband is selected as the single-sideband continuous chirped probe light. The continuously chirped probe light output from the electro-optic modulator IQ-EOM (104) is amplified by the erbium-doped fiber amplifier EDFA (107), input through the first port of the optical circulator (108), and injected into the sensing fiber (109) through the second port. The sensing fiber (109) is continuously laid along the cable under test and forms acoustic coupling with the outer surface of the cable at the test area of ​​the cable joint or the cable body. When the continuously chirped probe light propagates along the sensing fiber (109), backscattered Rayleigh light is generated at different positions of the fiber. The backscattered Rayleigh light returned by the sensing fiber (109) enters through the second port of the optical circulator (108) and is output through the third port to the second fiber coupler (110). The second fiber coupler (110) coherently superimposes the backscattered Rayleigh light with the local oscillator light to form a beat frequency optical signal. The beat frequency optical signal is converted into a beat frequency electrical signal by the balanced photodetector BPD (111) and then input into the data acquisition card DAQ (112). The second arbitrary waveform generator AWG2 (113) provides a trigger signal synchronized with the complete chirp cycle to the data acquisition card DAQ (112). Under the control of the synchronous trigger signal, the data acquisition card DAQ (112) acquires the beat frequency electrical signal and transmits the acquired data to the computer (114). During the detection process, the acoustic vibration signal generated by partial discharge in the cable acts on the adjacent sensing fiber (109), causing a change in the length of the sensing fiber at the corresponding location, which in turn causes a phase change in the back Rayleigh scattering light. The computer (114) obtains the dynamic phase change distributed along the sensing fiber (109) based on the collected beat frequency electrical signal, and determines the distance position of the disturbance on the sensing fiber based on the round-trip propagation delay of the back Rayleigh scattering signal. Combining the laying correspondence between the sensing fiber (109) and the cable under test, the actual location of the partial discharge on the cable joint or the cable body can be determined. In one specific embodiment, the sensing fiber is a single-mode fiber with an operating wavelength of 1550nm, the target sensing distance L is 20km, the sweep period T is 240μs, the starting frequency f0 is 10MHz, the cutoff frequency f1 is 110MHz, the sweep bandwidth B is 100MHz, and the sweep slope K is 0.4167MHz / μs. The instantaneous frequency shift of the single-sideband continuous chirped probe light relative to the single-frequency continuous light varies linearly with time in the range of 10-110MHz. The electrical bandwidth of the balanced photodetector BPD is not less than 120MHz to fully preserve the beat frequency components in the range of 10-110MHz. The period of the synchronization trigger signal is 240μs. Preferably, the sampling rate of the data acquisition card DAQ is not less than 250MS / s, and the analog bandwidth is not less than 120MHz to meet the acquisition requirements of the 10-110MHz broadband beat frequency electrical signal. The computer divides the 10–110 MHz broadband beat frequency electrical signal into 10 adjacent and non-overlapping sub-chirped frequency bands. The passband ripple of each bandpass filter is no greater than 1 dB, the stopband attenuation is no less than 40 dB, and the single-sided transition band width is no greater than 1 MHz. For each 10 MHz sub-chirped frequency band, a complex reference signal with a duration of 24 μs and a sweep direction and sweep slope consistent with that sub-chirp is generated. The conjugate time-reversed signal of the complex reference signal is used as a matched filter to compress the long-duration sub-chirped response into a narrow pulse distance response with a pulse width of approximately 0.1 μs. The effective refractive index n of the sensing fiber... eff Take 1.468, sub-chirp bandwidth B sub When the frequency is 10MHz, the theoretical spatial resolution is approximately 10.2m.

[0031] Example 2: like Figure 2 As shown, this embodiment provides a method for processing the partial discharge signal of a continuous chirped wave coherent optical time-domain reflectometry cable. The computer (114) sequentially performs sub-chirped extraction, optical field reconstruction, matched filtering, rotation vector summation, phase demodulation, time sequence rearrangement and splicing, and disturbance signal reconstruction on the complete chirped beat frequency signal acquired by the data acquisition card DAQ (112). In this embodiment, the partial discharge signal of the cable can be processed according to the following steps: In this embodiment, taking a sensing distance of 20km as an example, the complete chirp period is 240μs, the sweep frequency range is 10~110MHz, and the total sweep frequency bandwidth is 100MHz. The complete chirp beat frequency signal is divided into 10 sub-chirp frequency bands with a bandwidth of 10MHz each. Two adjacent sub-chirps are summed by rotating vector to form an effective phase sampling point. The first step is to acquire the complete chirped beat frequency signal. The 240μs periodic trigger signal output by the second arbitrary waveform generator AWG2 (113) is used as the time reference to determine the starting position of each complete chirped cycle. The electrical bandwidth of the balanced photodetector BPD (111) is not less than 120MHz. The data acquisition card DAQ (112) continuously acquires the beat frequency electrical signal at a sampling rate of not less than 250MS / s and an analog bandwidth of not less than 120MHz. The 10-110MHz broadband beat frequency electrical signal corresponding to each complete chirped cycle is extracted from the continuously acquired data. The second step is sub-chirp extraction. Based on the continuous chirped probe light's 10MHz start frequency, 110MHz cutoff frequency, 0.4167MHz / μs sweep slope, and 240μs complete chirp cycle, the broadband beat frequency electrical signal corresponding to one complete chirp cycle is divided into 10 frequency bands, each with a passband width of 10MHz and adjacent passbands not overlapping. The passband ripple of each bandpass filter is no greater than 1dB, the stopband attenuation is no less than 40dB, and the single-sided transition band width is no greater than 1MHz; the effective duration of each 10MHz sub-chirped frequency band is 24μs. The third step is to reconstruct the optical field. Hilbert transforms are performed on the beat frequency signals of each sub-chirp to construct analytic signals, and combined with the complex optical field information of the local oscillator, the backscattered Rayleigh optical field corresponding to each sub-chirp is reconstructed. The fourth step is matched filtering. For each 10MHz sub-chirped frequency band, a complex reference signal with a duration of 24μs and a sweep direction and sweep slope of 0.4167MHz / μs consistent with that sub-chirped frequency is generated. The conjugate time-reversed signal of this complex reference signal is used as a matched filter to perform matched filtering on the corresponding backscattered Rayleigh complex optical field, compressing the long-duration sub-chirped response into a narrow pulse distance response with a pulse width of approximately 0.1μs. When the refractive index of the sensing fiber is 1.468, a theoretical spatial resolution of approximately 10m is obtained. The fifth step is to perform rotational vector summation. For the sub-chirp beat frequency signal corresponding to each frequency interval, the matched filter complex result corresponding to the preset reference detection time is selected at each distance position of the sensing fiber as the reference complex result. The conjugate value of the reference complex result is taken and divided by its magnitude to obtain the normalized conjugate reference quantity. The normalized conjugate reference quantity corresponding to each frequency interval is multiplied by the matched filter complex result corresponding to other detection times of the same frequency interval to obtain the rotational complex result after eliminating the reference initial phase. The rotational complex results of two adjacent sub-chirps are grouped together for phase alignment and complex vector summation, so that the 10 sub-chirps form 5 sets of effective phase sampling results. Since the position of Rayleigh scattering coherent fading is different at different optical frequencies, rotational vector summation can utilize the complementarity between sub-chirps of different frequencies to reduce the phase noise spikes and local detection blind spots caused by coherent fading. The sixth step is to perform phase demodulation. The phase is extracted from the complex result after the summation of the rotation vector, and the initial phase elimination, spatial difference and temporal difference are performed in sequence to obtain the dynamic phase result distributed along the sensing fiber (109) at the corresponding detection time. Step 7: Time sequence rearrangement and splicing. Based on the linear frequency-time correspondence of the continuously chirped probe light, the actual detection times corresponding to the five sets of effective phase sampling results within a complete 240μs chirped period are determined. The dynamic phase results are arranged sequentially according to the actual detection times, and the processing results of adjacent complete chirped periods are continuously spliced ​​to form a dynamic phase time sequence with an equivalent sampling period of 48μs. Therefore, the system's equivalent phase sampling rate is approximately 20.83kHz, the theoretical Nyquist frequency is approximately 10.42kHz, and the effective detection bandwidth of the partial discharge acoustic vibration signal can be set to 10kHz. Step 8: Restore the disturbance signal. Identify the distance locations where the phase changes from the distributed dynamic phase results after time-series rearrangement and splicing, and extract the dynamic phase time series at these locations to obtain the time-domain waveform and frequency characteristics of the partial discharge acoustic vibration signal. Determine the location of the partial discharge on the sensing optical fiber (109) of not less than 20km based on the round-trip propagation delay corresponding to the matched filtering results, and convert the optical fiber distance location into the actual location of the partial discharge on the cable under test. Through the above processing, the detection system can process 10 sub-chirps within a complete chirp cycle of 240μs into 5 effective phase sampling points at a sensing distance of 20km, so that the equivalent phase sampling rate reaches about 20.83kHz, the effective detection bandwidth reaches 10kHz, and a theoretical spatial resolution of about 10m is obtained. At the same time, Rayleigh scattering coherent fading is suppressed by rotating vector summation, and finally the continuous distributed detection, disturbance waveform reconstruction and spatial positioning of partial discharge acoustic vibration signals are realized. Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A continuous chirped wave coherent optical time-domain reflectometry (TDDR) cable partial discharge detection system, characterized in that, It includes a narrow linewidth laser (101), an optical isolator (102), a first fiber coupler (103), an electro-optic modulator based on in-phase / quadrature modulation IQ-EOM (104), an IQ-EOM driver (105), a first arbitrary waveform generator AWG1 (106), an erbium-doped fiber amplifier EDFA (107), an optical circulator (108), a sensing fiber (109), a second fiber coupler (110), a balanced photodetector BPD (111), a data acquisition card DAQ (112), a second arbitrary waveform generator AWG2 (113), and a computer (114). The output end of the first arbitrary waveform generator AWG1 (106) is connected to the electrical signal input end of the electro-optic modulator IQ-EOM (104) via the IQ-EOM driver (105); the first arbitrary waveform generator AWG1 (106) generates an in-phase I-channel linear frequency-swept electrical signal and a quadrature Q-channel linear frequency-swept electrical signal, and the full chirp period (T), start frequency (f0), cutoff frequency (f1), frequency sweep bandwidth (B) and frequency sweep slope (K) of the two paths of linear frequency-swept electrical signals respectively satisfy B=f1−f0 and K=B / T; the time (t) and instantaneous frequency (f(t)) within the full chirp period satisfy f(t)=f0+Kt, 0≤t<T; the full chirp period (T), the effective refractive index (n of the sensing optical fiber (109) eff ), the target sensing distance (L) and the speed of light in vacuum (c) satisfy T≥2n eff L / c; the target sensing distance L is not less than 20km, T is 240μs, f0 is 10MHz, f1 is 110MHz, B is 100MHz, and K is 0.4167MHz / μs; The initial phase of the I-line linear sweep frequency electrical signal is 0°, and the phase difference between the Q-line linear sweep frequency electrical signal and the I-line linear sweep frequency electrical signal is +90° or −90°. The electro-optic modulator IQ-EOM (104) modulates the single-frequency continuous light into a single-sideband continuous chirped probe light under the drive of the two linear sweep frequency electrical signals. The narrow linewidth laser (101) is connected to the first fiber coupler (103) via the optical isolator (102). The first fiber coupler (103) splits the single-frequency continuous light into probe light and local oscillator light. The single-sideband continuous chirped probe light is amplified by the erbium-doped fiber amplifier (EDFA) (107) and injected into the sensing fiber (109) by the optical circulator (108). The sensing fiber (109) is a common single-mode fiber continuously laid along the cable line to be tested, with a total sensing length of not less than 20km. The backscattered Rayleigh light generated by the sensing fiber (109) is output through the optical circulator (108) and coherently superimposed with the local oscillator light in the second fiber coupler (110) to form a beat frequency optical signal. The output of the second fiber optic coupler (110) is connected to the balanced photodetector BPD (111), which converts the beat frequency optical signal into a beat frequency electrical signal with an electrical bandwidth of not less than 120MHz. The electrical signal output of the balanced photodetector BPD (111) is connected to the data acquisition card DAQ (112), which has a sampling rate of not less than 250MS / s and an analog bandwidth of not less than 120MHz. The second arbitrary waveform generator AWG2 (113) provides the data acquisition card DAQ (112) with a trigger signal of period 240μs that is synchronized with the complete chirp period. Under the control of the synchronous trigger signal, the data acquisition card DAQ (112) continuously acquires beat frequency electrical signals in the range of 10 to 110MHz and transmits the acquired data to the computer (114). The computer (114) is configured to: extract 10 sub-chirp beat frequency signals, each with a bandwidth of 10MHz and with non-overlapping adjacent frequency bands, from the 10-110MHz broadband beat frequency electrical signal corresponding to each 240μs complete chirp cycle; remove the DC component of each sub-chirp beat frequency signal and perform Hilbert transform to construct a complex analytical signal; reconstruct the backscattered Rayleigh scattering complex optical field corresponding to each sub-chirp by combining the complex optical field information of the local oscillator; and construct a complex reference signal (s) for the nth sub-chirp with the same sweep direction, sweep slope, and duration as the sub-chirp. n (t)), and its conjugate time-reversed signal is used as a matched filter for the nth reconstructed complex optical field (r). n (t) is subjected to matched filtering, and the round-trip propagation delay (τ) of the nth sub-chirp and the matched filter output (y) are calculated. n (τ) satisfies ;Chirp bandwidth (B sub The theoretical spatial resolution (Δz) and the theoretical spatial resolution (Δz) satisfy Δz=c / (2n) eff B sub ); The computer (114) is further configured to: for each frequency range, select the matched filter complex result corresponding to a preset reference detection time at each distance position of the sensing fiber (109) as a reference complex result, take the conjugate value of the reference complex result and divide it by its modulus value to obtain a normalized conjugate reference quantity with a unit modulus value. The normalized conjugate reference is multiplied by the matched filter complex result corresponding to the frequency range at other detection times to obtain the rotated complex result after eliminating the initial reference phase. The rotated complex results corresponding to two adjacent sub-chimes of different frequencies are summed using a complex vector, resulting in 5 effective phase sampling results from 10 sub-chimes. The phase is extracted from the summed complex vector result, and initial phase elimination, spatial difference, and temporal difference are performed sequentially. The representative frequency (f) of the nth sub-chime... n ), starting frequency (f0), sweep slope (K), and actual detection time (t) n ) satisfies t n =(f n -f0) / K, arrange the effective phase sampling results sequentially according to the actual detection time and continuously splice the results of adjacent complete chirp periods; complete chirp period (T), number of effective phase sampling points (G), equivalent sampling period (T e ), equivalent sampling rate (f s,eq ) and the theoretical detection frequency limit (f max ) respectively satisfy T e =T / G, f s,eq =G / T, f max =G / 2T, when T=240μs and G=5, T e For 48 μs, f s,eq Approximately 20.83 kHz, f max The effective detection bandwidth is approximately 10.42 kHz; the computer (114) identifies the phase anomaly location and recovers the partial discharge acoustic vibration signal from the dynamic phase time series; the sensing fiber distance position (z), round-trip propagation delay (τ), and effective refractive index (n) are also considered. eff The speed of light in a vacuum (c) and the speed of light in a vacuum (c) satisfy z = cτ / 2n eff By combining the laying correspondence between the sensing optical fiber (109) and the cable under test, the actual location of the partial discharge on the cable under test is determined.

2. The continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system according to claim 1, characterized in that, The narrow linewidth laser (101) operates at a wavelength around 1550nm; the first fiber coupler (103) has a splitting ratio of 95:5, wherein the first light with a power ratio of 95% is used as the probe light and the second light with a power ratio of 5% is used as the local oscillator light.

3. The continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system according to claim 1, characterized in that, The two sub-Mach-Zehnder modulators of the electro-optic modulator IQ-EOM (104) are biased at the minimum transmission point, and the main Mach-Zehnder modulator is biased at the quadrature point. By setting the phase difference between the Q-line linear sweep signal and the I-line linear sweep signal to +90° or −90° to select the upper or lower sideband, the optical carrier suppression degree and the non-target sideband suppression degree are both not less than 20dB.

4. The continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system according to claim 1, characterized in that, Ten sub-chirped beat frequency signals with a bandwidth of 10MHz and no overlap between adjacent frequency bands are extracted from the 10-110MHz broadband beat frequency electrical signal corresponding to each complete 240μs chirped cycle. Ten sub-chirped beat frequency signals are then extracted using bandpass filters corresponding to the ten frequency ranges. The passband ripple of the ten bandpass filters is no greater than 1dB, the stopband attenuation is no less than 40dB, and the width of the single-sided transition band is no greater than 1MHz.

5. The continuous chirped wave coherent optical time-domain reflectometry cable partial discharge detection system according to claim 1, characterized in that, The duration of the complex reference signal is 24 μs, and the matched filter compresses each sub-chirp response into a narrow pulse distance response with a pulse width of about 0.1 μs; The effective refractive index (n) of the sensing fiber (109) eff Take 1.468, sub-chirp bandwidth (B sub When the MHz frequency is 10, the theoretical spatial resolution (Δz) is calculated as Δz = c / (2n). eff B sub The theoretical spatial resolution obtained from the calculation is approximately 10.2m.

6. A method for detecting partial discharge in a continuous chirped wave coherent optical time-domain reflectometry cable, characterized in that, Includes the following steps: S1: The first arbitrary waveform generator AWG1 (106) generates an in-phase I-channel linear sweep frequency signal and an orthogonal Q-channel linear sweep frequency signal. The complete chirp period of the two linear sweep frequency signals is 240μs, the starting frequency is 10MHz, the cutoff frequency is 110MHz, the sweep bandwidth is 100MHz, the sweep slope is 0.4167MHz / μs, the initial phase of the I-channel is 0°, and the phase difference between the Q-channel and the I-channel is +90° or −90°. The two linear sweep frequency signals are used to drive the electro-optic modulator IQ-EOM (104) to modulate the single-frequency continuous light output from the narrow linewidth laser into a single-sideband continuous chirped probe light. S2: The computer (114) divides the 10-110MHz broadband beat frequency electrical signal corresponding to each complete chirp cycle into 10 frequency intervals, and uses bandpass filters corresponding to the 10 frequency intervals to extract 10 sub-chirp beat frequency signals. S3: The computer (114) removes the DC component from each sub-chirp beat frequency signal and performs Hilbert transform on each sub-chirp beat frequency signal after removing the DC component to construct the corresponding complex analytical signal; The computer (114) combines the complex optical field information of the local oscillator light to reconstruct the backscattered Rayleigh optical field corresponding to each sub-chirp. S4: For each 10MHz sub-chirped frequency band, a complex reference signal with a duration of 24μs and a sweep direction and sweep slope of 0.4167MHz / μs consistent with that sub-chirped frequency is generated. The conjugate time-reversed signal of this complex reference signal is then used as a matched filter to perform matched filtering on the corresponding backscattered Rayleigh complex optical field, compressing the long-duration sub-chirped response into a narrow pulse distance response with a pulse width of approximately 0.1μs. With a refractive index of 1.468 in the sensing fiber, a theoretical spatial resolution of approximately 10m is obtained. S5: For each frequency range, the computer (114) selects the matched filter complex result corresponding to the preset reference detection time at each distance position of the sensing fiber (109) as the reference complex result, and takes the conjugate value of the reference complex result and divides it by the modulus value of the reference complex result to obtain the normalized conjugate reference quantity with unit modulus; the computer (114) multiplies the normalized conjugate reference quantity with the matched filter complex result corresponding to the corresponding frequency range at other detection times to obtain the rotated complex result after eliminating the reference initial phase, and performs phase alignment and complex vector summation on the rotated complex results corresponding to two adjacent different frequency sub-chirps as a group, so that the 10 sub-chirps form 5 groups of effective phase sampling results; S6: The computer (114) extracts the phase from the complex result after summing the rotation vectors of each group, and performs initial phase elimination, spatial difference and temporal difference on the extracted phase in sequence to obtain the dynamic phase result distributed along the sensing fiber (109) at the detection time corresponding to each effective phase sampling result. S7: Based on the linear frequency-time correspondence of the continuous chirped probe light, determine the actual detection time corresponding to each effective phase sampling result within the complete chirped cycle; arrange the 5 effective phase sampling results within the same complete chirped cycle in sequence according to the actual detection time, and continuously splice the processing results of adjacent complete chirped cycles to form a dynamic phase time series with an equivalent sampling period of 48μs and an equivalent sampling rate of approximately 20.83kHz, and an effective detection bandwidth of 10kHz for the partial discharge acoustic vibration signal; S8: Identify the distance position where the phase change occurs from the dynamic phase time series and extract the dynamic phase signal at that distance position to obtain the time-domain waveform and frequency characteristics of the partial discharge acoustic vibration signal; The location of the partial discharge on the sensing fiber is determined based on the round-trip propagation delay corresponding to the matched filter output and the effective refractive index of the sensing fiber. Then, by combining the laying correspondence between the sensing fiber and the cable under test, the distance position on the sensing fiber is converted into the actual location of the partial discharge on the cable under test.

7. The method for detecting partial discharge in a continuous chirped wave coherent optical time-domain reflectometry cable according to claim 6, characterized in that, In step S1, the second arbitrary waveform generator AWG2 (113) outputs a trigger signal with a period of 240μs that is synchronized with the complete chirp cycle; the electrical bandwidth of the balanced photodetector BPD (111) is not less than 120MHz, the sampling rate of the data acquisition card DAQ (112) is not less than 250MS / s and the analog bandwidth is not less than 120MHz; the data acquisition card DAQ (112) continuously acquires beat frequency electrical signals under the control of the synchronous trigger signal, and the computer (114) uses the synchronous trigger signal as a time reference to determine the starting position of each complete chirp cycle, and extracts the 10~110MHz broadband beat frequency electrical signal corresponding to each complete chirp cycle from the continuously acquired beat frequency electrical signals.

8. The method for detecting partial discharge in a continuous chirped wave coherent optical time-domain reflectometry cable according to claim 6, characterized in that, In step S2, the frequency bands of adjacent sub-chirped beat frequency signals do not overlap; the passband fluctuation of each bandpass filter is no greater than 1dB, the stopband attenuation is no less than 40dB, and the width of the single-sided transition band is no greater than 1MHz.

9. The method for detecting partial discharge in a continuous chirped wave coherent optical time-domain reflectometry cable according to claim 6, characterized in that, The sensing fiber (109) is a common single-mode fiber with a working wavelength of 1550nm and a total sensing length of not less than 20km.

Images