Distributed vibration sensing method based on optical frequency domain reflection based on Pearson correlation coefficient

Abstract

The present invention relates to a distributed vibration sensing method based on Pearson correlation coefficient optical frequency domain reflection, comprising the following steps: a first step of measuring a reference signal and a measurement signal; a second step of compensating for nonlinear phase noise and calculating a Rayleigh backscattering spectrum; a third step of obtaining a local RBS in the optical frequency domain; a fourth step of calculating a distributed Pearson correlation coefficient (PCC); based on the principle that vibration will cause a step-like jump in the distributed Pearson correlation coefficient (PCC), and that the size of the jump is proportional to the vibration amplitude within a certain range, the PCC of the local reference RBS and the local measurement RBS in each window is calculated to obtain the distributed PCC of the reference signal and the measurement signal at various positions of the optical fiber to be measured; a fifth step of detecting a change point in the PCC using a pruned exact linear time (PELT) algorithm based on dynamic programming to determine a possible vibration position; a sixth step of determining the vibration position; and a seventh step of determining the vibration amplitude based on the differential PCC.

Description

Technical Field

[0001] The present invention belongs to the technical field of distributed optical fiber sensing instruments, and in particular relates to a vibration demodulation method based on the Pearson correlation coefficient, which can be applied to long-distance distributed vibration sensing based on optical frequency domain reflection. Background Art

[0002] Distributed fiber optic vibration sensing technology is widely used in large-scale and wide-range information perception such as oil and gas pipelines, perimeter security, and stratum structure. In the distributed fiber optic vibration sensor, the entire optical fiber is a sensing point. It has the advantages of ultra-large capacity, ultra-high resolution, ultra-high sensitivity, ultra-long distance, strong adaptability and anti-electromagnetic interference, and has great potential in long-distance vibration sensors. In order to expand the sensing distance, the most common method is to use a two-end amplifier, install a relay amplifier, or use a weak grating to enhance the sensitivity of the optical fiber. Commonly used pumped distributed optical amplifiers are Raman amplifiers, Brillouin amplifiers, or Raman-Brillouin hybrid amplifiers. Wang et al. first applied pump amplifiers to distributed fiber optic vibration sensors and achieved vibration positioning of 175km and 25m spatial resolution [1]. Chen et al. used Raman amplifiers in the TGD-OFDR system and pre-distorted the optical pulses into a standard Hanning window to suppress interference and polarization fading, successfully achieving a sensing distance of 108km and a spatial resolution of 5m. Vibration sensing with strain sensitivity [2]. Fan et al. designed a distributed vibration sensing system based on six-span coherent detection and recovered the vibration waveform at 300.2 km [3]. The two-end system has good sensing performance in ultra-long DOFVS, but its complex structure makes it difficult to apply to existing optical networks. At the same time, power supply in extreme environments is another challenge. For single-end systems, Rayleigh scattering enhanced fiber, weak FBG, etc. are used to improve the signal-to-noise ratio in the optical fiber to be tested, thereby improving vibration sensitivity and increasing the sensing distance, but this will result in high cost and quasi-distributed problems.

[0003] In Optical Frequency Domain Reflectometry (OFDR), traditional long-distance distributed fiber vibration sensing systems are all based on external modulation, using a carrier-suppressed single-sideband modulator and an ultra-narrow linewidth laser as the system light source. The vibration information is demodulated based on the similarity analysis of Rayleigh backscattering. Rayleigh backscattering can be used as the "fingerprint" of the FUT in OFDR. The vibration on the FUT will cause changes in the "fingerprint", and the change in the "dissimilarity level" can be used to locate the vibration event. The distance domain cross-correlation similarity analysis (CCSA) method [4], optical frequency domain CCSA [5], the multi-characteristic analysis method combining CCSA with the "V"-shaped characteristics of Rayleigh backscattering [6], CCSA based on cross-correlation peak amplitude [7], and CCSA based on Euclidean distance [8] have all been used for vibration positioning in long-distance OFDR. Using the CCSA method based on the cross-correlation peak amplitude [7], the test distance can reach 120km-150km, and the disturbance positioning accuracy can reach 1m. However, it requires the amplitude of the vibration-free cross-correlation peak as the vibration judgment standard, and this value will change with the transmission distance. In addition, using the CCSA method based on Euclidean distance [8], multiple vibrations at 92km were successfully located, but the vibration amplitude was not determined. In order to simplify the complex external modulation system, the internal modulation TLS based on silicon-based photonic chips was used for OFDR. An arbitrary signal generator was used to control the external cavity temperature of the silicon-based photonic chip to achieve linear frequency modulation, successfully extending the sensing distance to 100km [9]. However, the above methods can only locate vibrations but cannot determine the vibration amplitude, which limits their judgment of vibration levels and the sensitivity is not mentioned.

[0004] References:

[0005] [1]Z.Wang,J.Zeng,J.Li,M.Fan,H.Wu,F.Peng,et al.,"Ultra-long phase-sensitive OTDR withhybrid distributed amplification,"Opt.Lett.,vol.39,no.20,pp.5866-5869,2014.

[0006] [2]D.Chen,Q.Liu and Z.He,"108-km Distributed Acoustic Sensor With220-pε / √Hz StrainResolution and 5-m Spatial Resolution,"J.LightwaveTechnol.,vol.37,no.18,pp.4462-4468,2019.

[0007] [3]C.Fan,H.Li,K.Zhang,H.Liu,Y.Sun,H.Liu,et al.,"300km ultralong fiberoptic DAS systembased on optimally designed bidirectional EDFA relays,"Photon.Res.,vol.11,no.6,pp.968-977,2023.

[0008] [4]Z.Ding,X.Yao,T.Liu,Y.Du,K.Liu,Q.Han,et al.,"12km-Long-rangevibration sensor based oncorrelation analysis of optical frequency-domainreflectometry signals,"Opt.Express,vol.20,

[0009] no.27,pp.28319-28329,2012.

[0010] [5]T.Liu,Y.Du,Z.Ding,K.Liu,Y.Zhou,J.Jiang,"40-km OFDR-BasedDistributed DisturbanceOptical Fiber Sensor,"IEEE Photon.Technol.Lett.,vol.28,no.7,pp.771–774,2016.

[0011] [6]Z.Ding,D.Yang,K.Liu,J.Jiang,Y.Du,B.Li,et al.,"Long-range OFDR-based distributedvibration optical fiber sensor by multi characteristics ofRayleigh scattering,"IEEE Photonics J.,

[0012] vol.9,no.5,art no.6804410,pp.1-10,2017.

[0013] [7] Liu Tiegen, Liu Kun, Ding Zhenyang, et al. A demodulation method for distributed disturbance sensing device based on Rayleigh scattering spectrum correlation coefficient.

[0014] Law[P]. China: CN201210100999.8, 2014-09-17.

[0015] [8] Ding Zhenyang, Zhang Teng, Liu Tiegen, et al. Distributed disturbance sensing and demodulation method based on Rayleigh scattering spectrum dissimilarity[P].

[0016] China: CN202210334748.X, 2023-09-15.

[0017] [9] Y. Yue, J. Zeng, Z. Ding, T. Zhang, H. Guo, T. Liu, "Long-range distributed vibration sensing based on internal-modulation OFDR," Nanotechnology and Precision Engineering, vol.7, no.4, art no.

[0018] 043002,pp.1-8,2024.

[0019]

[10] R.Killick, P.Fearnhead, I.Eckley, "Optimal detection of changepoints with a linear computational cost," J Am StatAssoc., vol.107, no.500, pp.1590–1598, 2012. Summary of the Invention

[0020] This paper provides a Pearson correlation coefficient-based optical frequency domain reflection vibration sensing demodulation method. This method utilizes a single-ended, unamplified, internally modulated OFDR system and utilizes the dynamic programming-based PELT algorithm for PCC change point detection, achieving high-sensitivity vibration location and amplitude determination over long distances. The internally modulated OFDR system based on PCC demodulation achieves long-range vibration sensing over 100 kilometers. The method involves the following steps:

[0021] A distributed vibration sensing method based on optical frequency domain reflection of Pearson correlation coefficient includes the following steps:

[0022] The first step is to measure the reference signal and the measurement signal: the OFDR system is used to measure the beat frequency signals of the additional interferometer and the main interferometer without applying vibration, and the beat frequency signal of the main interferometer in the vibration-free state is recorded as the original reference signal; vibration is applied to the optical fiber, and the beat frequency signals of the additional interferometer and the main interferometer under the vibration condition are measured, and the beat frequency signal of the main interferometer under the vibration condition is recorded as the original measurement signal;

[0023] The second step is to compensate for nonlinear phase noise and calculate the Rayleigh backscattering spectrum. The original reference signal and the original measurement signal are de-skewing filtered using the phase information extracted by the additional interferometer to compensate for nonlinear phase noise, obtaining the reference signal and measurement signal in the optical frequency domain. The reference signal and measurement signal in the optical frequency domain are converted to the range domain by performing an FFT to obtain the Rayleigh backscattering spectrum (RBS) of the reference signal and measurement signal.

[0024] The third step is to obtain the local RBS in the optical frequency domain: a sliding window is taken in the RBS of the reference signal and the measured signal to obtain the local RBS within each window. The local RBS is converted to the optical frequency domain by inverse fast Fourier transform to obtain the local reference RBS and local measured RBS at each position.

[0025] Step 4: Calculate the distributed Pearson correlation coefficient (PCC). Based on the principle that vibration can cause a step-like jump in the distributed Pearson correlation coefficient (PCC), and that the magnitude of the jump is proportional to the vibration amplitude within a certain range, calculate the PCC of the local reference RBS and the local measurement RBS in each window to obtain the distributed PCC of the reference signal and the measurement signal at each position of the fiber under test.

[0026] In the fifth step, the dynamic programming-based pruned exact linear time (PELT) algorithm is used to detect the change points in the PCC and determine the possible vibration locations.

[0027] Step 6: Determine the vibration location. Perform PCC differential processing on all change point locations detected by the PELT algorithm and set a differential PCC threshold, which is related to the noise level. Compare the differential PCC with the threshold, and determine locations above the threshold as vibration.

[0028] Step 7: Determine the vibration amplitude based on the differential PCC.

[0029] Furthermore, in the fourth step, the range is limited by the phase change of the beat frequency signal caused by the vibration, and the proportional relationship is satisfied when the phase change is within 2π.

[0030] Furthermore, the method of the fifth step is as follows: the residual sum of squares of the PCC sequence is calculated as the cost function, and a penalty value is set to prevent overfitting. Iteration is performed based on PELT, and observation points that meet the pruning conditions are deleted in each iteration. At the end of the iteration, all change point positions are obtained to determine the possible vibration positions.

[0031] Furthermore, the method of the seventh step is as follows: using vibration sources to make the optical fiber vibrate with different strain amplitudes, calculating the differential PCC of each vibration amplitude, obtaining the differential PCC and the vibration amplitude calibration coefficient; and determining the vibration amplitude by combining the calibration coefficient with the differential PCC.

[0032] The beneficial effects of the technical solution provided by the present invention are:

[0033] 1. Compensation for the nonlinear tuning effect of the light source in optical frequency domain reflection is realized.

[0034] 2. Realize long-distance optical fiber vibration positioning and amplitude setting.

[0035] 3. Realized long-distance optical fiber multi-point vibration measurement.

[0036] 4. The strain resolution in OFDR-based distributed optical fiber vibration sensing has been greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Flowchart of OFDR long-distance vibration sensing based on PCC demodulation;

[0038] Figure 2 Diagram of the optical frequency domain reflectometry system

[0039] Figure 3 Results of single-point vibration positioning experiments: (a) Rayleigh scattering spectrum in the distance domain; (b) distributed PCC; (c) differential PCC; (d) local magnification of the vibration position in (b).

[0040] Figure 4 is a schematic diagram of the calibration curve;

[0041] Figure 5 (a) Measured distributed PCC for multi-point vibration test results, (b) Local magnification of PCC at the vibration position, (c) Vibration demodulation results, (d) Local magnification of the vibration position in Figure (c)

[0042] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0043] 1: Arbitrary signal generator; 2: Silicon-based photonic chip narrow linewidth laser;

[0044] 3: variable optical attenuator; 4: 1:99 polarization-maintaining beam splitter;

[0045] 5: First circulator; 6: 50:50 coupler;

[0046] 7: First delay fiber; 8: First Faraday rotator mirror;

[0047] 9: Second Faraday mirror; 10: First balanced detector;

[0048] 11:20:80 polarization-maintaining coupler; 12: reference arm;

[0049] 13: test arm; 14: second circulator;

[0050] 15: second delay fiber; 16: first vibration source;

[0051] 17: Second vibration source; 18: Optical fiber to be tested;

[0052] 19: optical mixer; 20: second balanced detector;

[0053] 21: Third balanced detector; 22: Acquisition card

[0054] 23: Computer; 24: Trigger line;

[0055] 25: Additional interferometer; 26: Main interferometer. DETAILED DESCRIPTION

[0056] The present invention uses a single-ended internal modulation OFDR system and applies the Pearson correlation coefficient (PCC) to long-distance distributed optical fiber vibration sensing demodulation. Vibration will cause the Pearson correlation coefficient to decrease and the amount of decrease is proportional to the vibration amplitude. Vibration positioning can be achieved by locating the step-type change point in the distributed PCC, and vibration amplitude can be determined by calculating the differential PCC. We use the pruned exact linear time (PELT) algorithm based on dynamic programming

[10] to detect the change point in the PCC. In order to avoid noise interference, we set a differential PCC threshold and compare the differential PCC of all change point positions with the threshold. Positions above the threshold are judged as vibrations. The vibration amplitude can be determined by combining the differential PCC with the strain calibration coefficient. Compared with the traditional two-ended amplification and external modulation system, the system cost and operation complexity are effectively reduced; compared with the previously proposed internal modulation system, the system strain resolution is effectively improved and vibration amplitude measurement is achieved. This method can achieve micro-vibration sensing of >100km, with a spatial resolution of 15.3m and a strain sensitivity of 868pε. This method can also achieve multi-point vibration sensing.

[0057] The present invention will be described below with reference to the accompanying drawings and embodiments.

[0058] Example 1:

[0059] This example includes a distributed optical fiber vibration sensing system device based on optical frequency domain reflection

[0060] The optical frequency domain reflection distributed optical fiber vibration sensing system includes: an arbitrary signal generator 1, a silicon-based photonic chip narrow linewidth laser 2, a variable optical attenuator 3, a 1:99 optical beam splitter 4, an acquisition card 22, a computer 23, a trigger line 24, an additional interferometer 25 and a main interferometer 26.

[0061] The additional interferometer 25 is a polarization-insensitive Michelson interferometer structure, which is used to extract the optical frequency to estimate the nonlinear phase noise information of the light source, including: a first circulator 5, a first 50:50 coupler 6, a first delay fiber 7, a first Faraday rotator mirror 8, a second Faraday rotator mirror 9 and a balanced detector 10.

[0062] The main interferometer 26 is a modified Mach-Zehnder structure, comprising a 20:80 polarization-maintaining coupler 11, a reference arm 12, a test arm 13, a second circulator 14, a fiber under test 18 comprised of a second delay fiber 15, a first vibration source 16, and a second vibration source 17, an optical hybrid 19, a second balanced detector 20, and a third balanced detector 21. The main interferometer is used to measure vibration signals and is the core structure of the system.

[0063] The a port of the arbitrary signal generator 1 is connected to the a port of the silicon-based photonic chip narrow linewidth laser 2; the b port of the arbitrary signal generator 1 is the trigger out interface, which is connected to the trigger of the acquisition card 22 through the trigger line 24. in port; the b port of the silicon-based photonic chip narrow linewidth laser 2 is connected to the a port of the variable optical attenuator 3; the b port of the variable optical attenuator 3 is connected to the a port of the 1:99 optical beam splitter 4; the b port of the 1:99 optical beam splitter 4, i.e., the 1% splitting port, is connected to the a port of the first circulator 5; the c port of the 1:99 optical beam splitter 4, i.e., the 99% splitting port, is connected to the a port of the 20:80 polarization-maintaining coupler 11; the b port of the first circulator 5 is connected to the a port of the 50:50 coupler 6; the c port of the first circulator 5 is connected to the input end of the balanced detector 10; the b port of the 50:50 coupler 6 is connected to the input end of the balanced detector 10; the c port of the 50:50 coupler 6 is connected to the first Faraday rotator 8 through the first delay fiber 7; the d port of the 50:50 coupler 6 is connected to the second Faraday rotator 9; the output of the balanced detector 10 The output end is connected to a channel of the acquisition card 22; the c port of the 20:80 polarization-maintaining coupler 11, i.e., the 20% splitting port, is connected to the a port of the optical hybrid 19 through the reference arm 12; the d port of the 20:80 polarization-maintaining coupler 11, i.e., the 80% splitting port, is connected to the a port of the second circulator 14 through the test arm 13; the b port of the second circulator 14 is connected to the optical fiber 18 to be tested; the c port of the second circulator 14 is connected to the b port of the optical hybrid 19; the c port and the d port of the optical hybrid 19 are connected to the input end of the second balanced detector 20; the e port and the f port of the optical hybrid 19 are connected to the input end of the third balanced detector 21; the output end of the second balanced detector 20 is connected to a channel of the acquisition card 22; the output end of the third balanced detector 21 is connected to a channel of the acquisition card 22; and the output end of the acquisition card 22 is connected to the input end of the computer 23.

[0064] When the device is working, the modulated signal output from port a of the arbitrary signal generator 1 enters port a of the silicon-based photonic chip narrow linewidth laser 2, and the light emitted from port b of the silicon-based photonic chip narrow linewidth laser 2 enters port a of the variable optical attenuator 3, and enters port a of the 1:99 polarization-maintaining beam splitter 4 from port b of the variable optical attenuator 3, and enters port a of the first circulator 5 from port b of the 1:99 polarization-maintaining beam splitter 4, i.e., the 1% splitting port, and enters port a of the first circulator 5. Through port b of the first circulator 5, the light enters port a of the 50:50 coupler 6, 50% of which is emitted from port c and 50% from port d. The outgoing light from the c port of the coupler 6 is reflected by the first Faraday rotator 8 through the first delay fiber 7 and returns to the c port of the 50:50 coupler 6. The outgoing light from the d port of the 50:50 coupler 6 is reflected by the second Faraday rotator 9 and returns to the d port of the 50:50 coupler 6. The two beams of light interfere with each other in the 50:50 coupler 6 and are output from the a and b ports. The light from the a port of the 50:50 coupler 6 is input by the b port of the first circulator 5 and output from the c port. The output light and the output light from the b port of the first 50:50 coupler 6 are heterodyned and converted into analog electrical signals in the balanced detector 10. The output light from the c port of the 1:99 polarization-maintaining optical beam splitter 4, i.e., the 99% splitting port, is input by the a port of the 20:80 polarization-maintaining coupler 11, 20% of the output light from the c port enters the reference arm 12, and 80% of the output light from the d port enters the test arm 13. The light in the test arm 13 is input by the a port of the second circulator 14 and enters the optical fiber to be tested 18 from the b port of the second circulator 14, while the back Rayleigh scattered light of the optical fiber to be tested 18 enters from the b port of the second circulator 14 and is output from the c port. The reference light output from the reference arm 12 enters the a port of the optical hybrid 19 and is combined with the reference light entering the b port of the optical hybrid 19 from the c port of the second circulator 14 to form beat frequency interference. The light is heterodyned by the c port and d port of the optical hybrid 19 in the second balanced detector 20 and converted into an analog electrical signal, which is transmitted to the acquisition card 22; the output light from the e port and f port of the optical hybrid 19 is heterodyned by the third balanced detector 21 and converted into an analog electrical signal, which is transmitted to the acquisition card 22. The arbitrary signal generator 1 modulates and starts to send out a trigger signal to trigger the acquisition card 22 via the trigger line 24 , which converts the acquired analog electrical signal into a digital electrical signal and transmits it to the computer 23 .

[0065] The arbitrary signal generator 1 and the silicon-based photonic chip narrow linewidth laser 2 form a tunable laser, which provides a light source for the optical frequency domain reflection system, and its optical frequency can be linearly scanned.

[0066] The first circulator 5 prevents the light reflected from the port a of the 50:50 coupler 6 in the additional interferometer from entering the laser. The 50:50 coupler 6 is used for light splitting and light interference.

[0067] The first Faraday rotator mirror 8 and the second Faraday rotator mirror 9 are used to provide reflection for the additional interferometer and can eliminate the polarization fading phenomenon of the additional interferometer.

[0068] The first delay fiber 7 is used to realize the beat frequency interference of non-equal arms, and the optical frequency can be obtained according to the beat frequency and the length of the delay fiber.

[0069] The balanced detector 10 is used to collect the outgoing light from the b port of the 50:50 coupler 6 and the outgoing light from the c port of the first circulator 5 , that is, the time domain interference beat signal of the additional interferometer.

[0070] The optical mixer 19 completes polarization splitting of the signal, so that the light intensity of the reference light and the test light in two orthogonal directions during polarization splitting is basically consistent, eliminating the influence of polarization fading noise, and realizing the combination of the reference light and the test light to form beat frequency interference.

[0071] Acquisition card 22: The digital signal after analog-to-digital conversion of the wavelength domain beat frequency signal generated by the interferometer is transmitted to the computer via a bus such as USB and PCIE.

[0072] Computer 23: performs data processing on the interference signal after analog-to-digital conversion by the acquisition card to realize distributed vibration sensing of the optical fiber.

[0073] Example 2:

[0074] This example provides a long-distance distributed fiber optic vibration sensing system and demodulation method based on optical frequency domain reflectometry, achieving fiber optic distributed vibration sensing over distances greater than 100 km. The steps are as follows:

[0075] The first step is to measure the reference and measurement signals. The OFDR system measures the beat frequency signals of the supplementary interferometer and the main interferometer without vibration. The beat frequency signal of the main interferometer without vibration is recorded as the original reference signal. A vibration source is used to apply vibration to the fiber tail end. The beat frequency signals of the supplementary interferometer and the main interferometer with vibration are measured, and the beat frequency signal of the main interferometer under vibration is recorded as the original measurement signal.

[0076] The second step is to compensate for nonlinear phase noise and calculate the Rayleigh backscattering spectrum. The original reference signal and the original measurement signal are de-skewing filtered using the phase information extracted by the additional interferometer, respectively, to obtain the reference and measurement signals in the optical frequency domain. This de-skewing filtering algorithm effectively compensates for the nonlinear tuning effects of the tunable light source in the optical frequency domain, suppressing the energy diffusion of the distance-domain reflection peak of the measured fiber. Without compensation, sensing performance can be degraded and even masked by disturbance points, rendering sensing ineffective. The reference and measurement signals in the optical frequency domain are converted to the distance domain using an FFT, yielding their Rayleigh backscattering spectra (RBS).

[0077] The third step is to obtain the local RBS in the optical frequency domain. A sliding window, with N number of window points, is taken in the RBS of the reference and measured signals, obtaining the local RBS within each window. The local RBS is then converted to the optical frequency domain using an inverse fast Fourier transform to obtain the local reference RBS and local measured RBS at each location.

[0078] The fourth step is to calculate the distributed PCC. In OFDR, the key to vibration demodulation is to identify the difference between the local RBS in the reference signal and the measured signal, which can be determined by PCC. We assume that the local reference RBS and the local measured RBS at a certain location are X i and Y i , X i and Y i are all discrete sequences of N points, and their means are and Their standard deviations are σ X and σ Y , then PCC can be expressed as

[0079]

[0080] Vibration causes a step-like jump in the PCC, and the magnitude of the jump is proportional to the vibration amplitude within a certain range. This range is limited by the phase change of the beat frequency signal caused by vibration, which satisfies the proportional relationship within 2π. Based on the above analysis, we calculate the PCC of the local reference RBS and the local measurement RBS within each window, obtaining the distributed PCC of the reference and measurement signals at various locations on the fiber under test, and thus obtaining the PCC sequence.

[0081] In the fifth step, the Pruned Exact Linear Time (PELT) algorithm based on dynamic programming is used to detect the change points in the PCC

[10] . We use the PELT algorithm to locate all the change points in the PCC and thus determine the possible vibration locations. The PELT algorithm determines the optimal change point locations and number of change points by segmenting the PCC to minimize the cost function. The PELT algorithm has higher segmentation accuracy, and its computational cost is proportional to the number of observation points, making it suitable for processing large amounts of data. We calculate the residual sum of squares of the PCC sequence as the cost function and set a penalty value to prevent overfitting. We iterate according to the literature

[10] and delete the points that meet the pruning conditions in each iteration. At the end of the iteration, the optimal change point location is obtained. However, not all change points are caused by vibration. Changes in the similarity of the light source and noise in the optical fiber transmission link can all produce change points in the PCC. However, the PCC jump caused by noise is continuous, while the PCC jump caused by vibration is step-like.

[0082] Step 6: Determine the vibration location. To distinguish between change points arising from vibration and noise, we use the PELT algorithm to differentiate the PCCs at all change point locations detected in step 5. A threshold for the differential PCC is set, which is related to the noise level. The differential PCC is compared with the threshold, and locations above the threshold are identified as vibration.

[0083] Step 7: Determine the vibration amplitude. Apply different voltage waveforms to the vibration source to produce vibrations with different strain amplitudes in the optical fiber. Calculate the differential PCC for each vibration amplitude, obtaining the differential PCC and the vibration amplitude calibration coefficient. Combined with the calibration coefficient, the vibration amplitude is determined from the differential PCC.

[0084] Example 3

[0085] The feasibility of the long-distance distributed optical fiber vibration sensing system and demodulation method in Examples 1-2 is verified by combining specific experiments, as described below:

[0086] The fiber 18 under test used in the verification experiment for this embodiment of the present invention consisted of approximately 100.9 km of single-mode fiber connected to a first vibration source. The first vibration source was then connected to a second vibration source via approximately 950 m of single-mode fiber, and the second vibration source was then connected to approximately 500 m of single-mode fiber. The measurement objective was to locate and determine the amplitude of vibrations in the fiber under test.

[0087] In our experiment, we used an OFDR system. A narrow-linewidth laser from a silicon-based photonic chip has a linewidth of approximately 1 kHz and a central wavelength of 1550 nm. Modulated by an arbitrary signal generator, it produces linearly tuned light with a tuning rate of 84 GHz / s and an effective tuning range of 6.72 GHz. The output power of the light source is approximately 8 mW, and the tuning time is 0.08 s. The additional interferometer delay fiber is 4989 m long, and the acquisition card has a sampling rate of 200 MS / s and 16M sampling points. The voltage gauge factors generated in the optical fiber by the first and second vibration sources are 3.26 nε / V and 3.11 nε / V, respectively. The vibration sources apply rectangular pulse waveforms with a pulse width of 20 μs and a period of 1 ms. The number of data points used for the sliding window is 1000, corresponding to a spatial resolution of 15.3 m.

[0088] Only the first vibration source is subjected to a peak voltage of 0.8V. The distance domain RBS after nonlinear compensation using de-skewing filtering is as follows: Figure 3 As shown in (a), the tail end is located at 102.42 km and the signal-to-noise ratio is 10 dB. Using formula (a), the PCC in each sliding window is calculated to obtain the distributed PCC along the fiber under test. Figure 3(b) shows that PCC decreases with distance due to transmission loss and phase noise accumulation. We use the PELT algorithm to detect the change points in PCC, and set the penalty value to 0.06 and the number of iterations to the number of PCC data points. After the iteration, all the change points detected are marked on Figure 3 (b) Figure 3 (d) Figure 3 (b) The local magnification of the vibration position shows that there is a step jump in the vibration position. In order to determine that the change point comes from vibration rather than noise, we perform PCC difference processing on all the change points to obtain Figure 3 (c), The differential PCC threshold is set to 0.0197, which is related to the noise level of PCC. The position above the threshold, 100.936 km, is judged as vibration, which is consistent with the position of the first vibration source we set.

[0089] In the first vibration source, the voltage of 0.8V to 2.4V was applied with a step size of 0.2V. The strain in the optical fiber was 2.608nε to 7.824nε. A total of 9 groups of experiments were completed. Repeat the above processing steps and calculate the differential PCC at each vibration amplitude. The strain and differential PCC calibration curves are as follows: Figure 4 The calibration coefficient of the vibration amplitude represented by the strain magnitude and the differential PCC is 0.0227, which determines the strain resolution to be 868 pε. 2 The linearity is 0.9993, which indicates good linearity.

[0090] At the same time, a voltage of 1V is applied to the first vibration source and the second vibration source, and the strains generated in the optical fiber are 3.26nε and 3.11nε respectively. The distributed PCC is calculated according to the processing flow of the first experiment as follows: Figure 5 As shown in (a), the PELT algorithm marks all change point locations. Figure 5 (b) Figure 5 The local enlarged image of (a) shows that there are two obvious step-type jumps at the two vibration source positions. In order to locate the vibration and avoid misjudgment caused by noise, we calculated the differential PCC and combined it with the calibration coefficient to obtain the vibration amplitude as follows: Figure 5 (c) is shown. Figure 5 In (c), vibrations at locations 100.930 km and 101.897 km, exceeding the sensitivity of 868 pε, were identified. The amplitudes were 3.39 nε and 3.17 nε, respectively, similar to the applied vibration amplitude. This demonstrates that this method can effectively identify multiple small vibrations over 100 km with a spatial resolution of 15.3 m and determine the vibration amplitude with a strain resolution of 868 pε.

[0091] Unless otherwise specified, the embodiments of the present invention do not limit the models of the components. Any component that can perform the above functions may be used.

[0092] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and that the serial numbers of the embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A distributed vibration sensing method based on optical frequency domain reflectometry using the Pearson correlation coefficient, comprising the following steps: The first step is to measure the reference signal and the measurement signal: the OFDR system is used to measure the beat frequency signals of the additional interferometer and the main interferometer without applying vibration, and the beat frequency signal of the main interferometer in the vibration-free state is recorded as the original reference signal; vibration is applied to the optical fiber, and the beat frequency signals of the additional interferometer and the main interferometer under the vibration condition are measured, and the beat frequency signal of the main interferometer under the vibration condition is recorded as the original measurement signal; The second step is to compensate for nonlinear phase noise and calculate the Rayleigh backscattering spectrum. The original reference signal and the original measurement signal are de-skewing filtered using the phase information extracted by the additional interferometer to compensate for nonlinear phase noise, obtaining the reference signal and measurement signal in the optical frequency domain. The reference signal and measurement signal in the optical frequency domain are converted to the range domain by performing an FFT to obtain the Rayleigh backscattering spectrum (RBS) of the reference signal and measurement signal. The third step is to obtain the local RBS in the optical frequency domain: a sliding window is taken in the RBS of the reference signal and the measured signal to obtain the local RBS within each window. The local RBS is converted to the optical frequency domain by inverse fast Fourier transform to obtain the local reference RBS and local measured RBS at each position. Step 4: Calculate the distributed Pearson correlation coefficient (PCC). Based on the principle that vibration can cause a step-like jump in the distributed Pearson correlation coefficient (PCC), and that the magnitude of the jump is proportional to the vibration amplitude within a certain range, calculate the PCC of the local reference RBS and the local measurement RBS in each window to obtain the distributed PCC of the reference signal and the measurement signal at each position of the fiber under test. In the fifth step, the dynamic programming-based pruned exact linear time (PELT) algorithm is used to detect the change points in the PCC and determine the possible vibration locations. Step 6: Determine the vibration location. Perform PCC differential processing on all change point locations detected by the PELT algorithm and set a differential PCC threshold, which is related to the noise level. Compare the differential PCC with the threshold, and determine locations above the threshold as vibration. Step 7: Determine the vibration amplitude based on the differential PCC.

2. The distributed vibration sensing method based on Pearson correlation coefficient optical frequency domain reflectometry according to claim 1 is characterized in that: In the fourth step, the range is limited by the phase change of the beat frequency signal caused by the vibration, and the proportional relationship is satisfied when the phase change is within 2π.

3. The distributed vibration sensing method based on Pearson correlation coefficient optical frequency domain reflectometry according to claim 1 is characterized in that: The method of the fifth step is as follows: calculate the residual sum of squares of the PCC sequence as the cost function, and set the penalty value to prevent overfitting. Iterate based on PELT and delete the observation points that meet the pruning conditions in each iteration. At the end of the iteration, all change point positions are obtained to determine the possible vibration position.

4. The distributed vibration sensing method based on Pearson correlation coefficient optical frequency domain reflectometry according to claim 1, characterized in that: The seventh step is as follows: using vibration sources to make the optical fiber vibrate with different strain amplitudes, calculating the differential PCC of each vibration amplitude, obtaining the differential PCC and the vibration amplitude calibration coefficient; and determining the vibration amplitude from the differential PCC in combination with the calibration coefficient.

Images