A data analysis method for a distributed fiber optic vibration sensing system

Abstract

A kind of data analysis method for distributed vibration sensing system, using distributed optical fiber sensing system to constitute front-end system output detection signal, using circulator to inject detection signal into the optical fiber to be measured and using detector to directly detect backscattering signal, realize the receiving and analysis of the signal to be measured;The detection signal collected by the detector is composed of a series of backscattering / reflection curves, and each detection pulse obtains a backscattering / reflection curve;The length of each backscattering / reflection curve is proportional to the length of the distributed vibration sensing optical fiber;The record curve matrix is redrawn to provide two-dimensional matrix data of backscattering / reflection curve, and the output of specific position is determined as a function of time;The two-dimensional matrix data of Rayleigh backscattering / backreflection curve is Fourier transformed along the distance direction row by row, and the disturbance information of the optical fiber sensing along the line is obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to a data analysis method for a distributed vibration optical fiber sensing system, in particular to a distributed optical fiber vibration sensing data harmonic analysis method for improving the dynamic range and strain resolution capability of the system, and belongs to the field of optical fiber sensing. BACKGROUND

[0002] Compared with other types of sensing technology, distributed optical fiber sensing technology has a competitive advantage of measuring the spatial distribution of physical parameters along the entire fiber length (usually tens of kilometers), rather than just obtaining information at a single location. Based on Rayleigh scattering, Brillouin scattering and Raman scattering, a variety of distributed sensing technologies have been developed. Among different types of distributed sensing technologies, distributed vibration sensors have attracted increasing attention, especially in the application of the oil and gas industry. Multiplexing of distributed vibration sensors is achieved by using polarization, phase or intensity information to locate the vibration measurement along the fiber.

[0003] Polarization-based distributed vibration sensors (DVS) are susceptible to vibration events close to the light injection point if there are multiple vibration points at the same time. The information of the far subsequent vibration points occurring after the first vibration point along the fiber is easily affected by the vibration event close to the light injection point. This limitation is due to the process of the system extracting polarization information. Most DVS systems are based on phase information extraction, usually achieved by double pulse method, heterodyne detection or interference phase recovery method. Existing DVS schemes based on phase information usually have complex system structure and are expensive. For example, the double pulse scheme requires two acousto-optic modulators (AOMs) or electro-optic modulators (EOMs) to obtain the probe pulse, the heterodyne detection method requires a narrow linewidth laser and a high-speed data acquisition (DAQ) card, and the interference phase recovery DVS requires three photodetectors (PDs), which produces three times the data, thus requiring more computing requirements.

[0004] The present application proposes a method for quantitatively measuring vibration strain based on Rayleigh backscattering (RBS) intensity information. This method quantitatively analyzes the vibration intensity by the number of harmonics, so it can greatly improve the system strain dynamic range and strain resolution compared with only based on intensity or polarization information, and can greatly reduce the cost compared with the scheme based on narrow linewidth light source and high-speed data acquisition card. The algorithm of the present application is simple and easy to implement, and is suitable for wide engineering applications. SUMMARY

[0005] The present application aims to provide a method for quantitatively measuring vibration strain based on Rayleigh backscattering (RBS) / back reflection signal intensity information. By analyzing multiple harmonic signals in the sensing data, the system dynamic range is improved and the strain is quantitatively measured.

[0006] The application improves the spatial resolution and the physical quantity measurement accuracy by using the following technical solutions: a data analysis method for a distributed vibration sensing system, a Rayleigh backscattering signal is acquired by using a directly detected distributed optical fiber sensing system, a distributed vibration sensing optical fiber constitutes a front-end system to output a detection signal, a circulator is used to inject the detection signal into a to-be-measured optical fiber, and a detector is used to directly detect the backscattering Rayleigh signal, so that the measurement signal is received and analyzed; the detection signal collected by the detector is composed of a series of backscattering curves (the modulator uniformly or non-uniformly emits a detection pulse signal, and each detection pulse obtains a backscattering / reflection curve); the length of each backscattering / reflection curve is proportional to the length of the distributed vibration sensing optical fiber; a record curve matrix is redrawn to provide two-dimensional matrix data of the backscattering curve and allow the output of a specific position to be determined as a function of time; the two-dimensional matrix data of the Rayleigh backscattering / backreflection curve is subjected to Fourier transform row by row along the distance direction, and disturbance information of the optical fiber sensing along the line can be obtained; and the method can improve the spatial resolution and the physical quantity measurement accuracy of the vibration sensing system.

[0007] Specifically, the following steps are included:

[0008] Step 1: The direct current light signal output by the laser in the distributed optical fiber sensing system is modulated by the modulator into a periodic detection pulse (the detection pulse width can be femtosecond, picosecond, nanosecond, microsecond or even millisecond according to the spatial resolution requirement), and the detection pulse enters the sensing optical fiber through the circulator or is amplified by the EDFA and then enters the sensing optical fiber through the circulator again;

[0009] Step 2: The backscattering Rayleigh / reflection signal of the sensing optical fiber is received by the detector or is received by the detector after being amplified by the EDFA;

[0010] Step 3: The data acquisition card is used to collect the output signal of the detector, and the measured one-dimensional data is reshaped into two dimensions according to the period; assuming that the period of the detection pulse is T, the data acquisition card collects and processes n periods of time domain data each time, each period of time domain measurement curve contains N data points, the continuous n periods of time domain data are assigned to a matrix n*N by period, and then the matrix is transposed into N*n;

[0011] Step 4: Fourier transform is performed on the N*n matrix row by row along the distance direction (i.e. along the direction of N rows in the N*n matrix), and the disturbance information of the optical fiber sensing along the line is obtained;

[0012] Furthermore, harmonic analysis improves the dynamic range of the system by performing a Fourier transform on the intensity information of the collected back Rayleigh scattering / reflection signals to obtain the location of the external disturbance information. The disturbance information is then accurately located according to the formula z = c * x * Δt / (2n), where z is the position on the optical fiber, c is the speed of light, x represents any point from 0 to N among the N sampling points collected in one cycle (i.e., the sequential number of uniform sampling along the optical fiber (a natural number)), the specific position corresponding to z is proportional to x, Δt is the sampling time interval, and n is the refractive index in the optical fiber.

[0013] The spectrum of each row of the N*n matrix obtained by Fourier transform along the distance direction represents the impact of vibration on each point on the optical fiber during the data acquisition time. If the optical fiber is not affected by vibration, no frequency other than zero frequency can be seen in the spectrum obtained after row Fourier transform. If the optical fiber is affected by vibration, in addition to zero frequency, vibration frequencies will appear from the position affected by vibration to the end of the optical fiber. The non-zero frequency position is the position of the vibration source.

[0014] Furthermore, the response spectrum after Fourier transform changes with the strength of the vibration signal, and harmonic phenomena appear in the response spectrum, with the harmonic order increasing as the vibration signal strengthens.

[0015] Furthermore, given a sufficiently low noise floor, the strain change occurs for each additional harmonic in the system response. Within the range, i.e., strain resolution Where λ is the incident light wavelength, n is the fiber refractive index, and Δl is the system spatial resolution, i.e., the distance resolution of data acquisition. Ideally, when λ is 1550 nm, n is 1.45, and Δl is 0.5 m, the corresponding strain value is 1.07 με.

[0016] Furthermore, analyzing the intensity of the vibration signal based on the harmonic order in the response spectrum of the vibration sensing system can improve the dynamic range of the sensing system.

[0017] The harmonic data analysis method proposed in this invention is applicable to various fiber optic sensing systems that acquire vibration information based on Fourier transform. The fiber optics mentioned in the system can be single-mode fiber, few-mode fiber, polarization-maintaining fiber, multi-core fiber, multimode fiber, photonic crystal fiber, or special fiber. Meanwhile, the sensing unit can also be an arrayed fiber grating or an fiber optic unit containing an array of weak reflection units.

[0018] The working principle of harmonic analysis methods to improve the dynamic range of a system can be described as follows: Based on distributed fiber optic sensing technology, the disturbance information can be accurately located by using the intensity information of the Rayleigh backscattered signal and the time value of the external disturbance information. The signal acquired by the detector consists of a series of backscattering / reflection curves (one scattering / reflection curve is obtained for each detection pulse). The length of each curve is proportional to the length of the sensing fiber, and the repetition rate is equal to the repetition frequency of the pulse. The recorded curve matrix is ​​redrawn to provide two-dimensional matrix data of the backscattering curves, allowing the output at a specific location to be determined as a function of time. By performing a Fourier transform on the two-dimensional matrix data of the backscattering Rayleigh curves row by row along the distance direction, the disturbance information along the fiber optic sensing line can be obtained.

[0019] The present invention adopts the above technical solution and can produce the following technical effects: The present invention achieves distributed vibration sensing by using harmonic analysis method without increasing or even reducing the technical complexity of traditional distributed vibration sensing systems, and improves the dynamic range by analyzing the harmonic information in the spectrum response. Attached Figure Description

[0020] Figure 1 Block diagram of a preferred embodiment of a distributed vibration sensing system;

[0021] Figure 2 Data analysis and processing steps diagram; Figure 2 (a) Rayleigh backscattering curve, Figure 2 (b) is Figure 2 The data in (a) is periodically segmented and reshaped. Figure 2 (c) is correct. Figure 2 Transpose the data in (b);

[0022] Figure 3 (a) and (b) are the response spectra of the fiber optic sensing system when the PZT1 driving voltage is 0.8V and 20V, respectively. Figure 3 (b) The numbers 1-12 represent the harmonic order. Here, PZT1 is a piezoelectric ceramic that simulates vibration. Other vibration equipment or actual vibration sources can also be used.

[0023] Figure 4 The four figures show the spectral response of the fiber optic vibration measurement system when two vibrations exist along its path. The system responses of PZT1 and PZT2 (simulated vibration sources) under different driving voltages are shown. The driving frequencies of the two PZTs are 800Hz (PZT1) and 500Hz (PZT2). Figures (a) and (b) show the results when the driving voltage of PZT1 is 0.8V, and figures (c) and (d) show the results when the driving voltage of PZT2 is 20V. The spectral response diagrams show the system response when two vibrations exist along its path. Figure 4In figures (a) and (c), the spectral information is from the two vibration source locations (1632 m, 800 Hz; 1686 m, 500 Hz, respectively). Figure 4 In Figures (b) and (d), the spectral information is distributed along the optical fiber, corresponding to the response spectral information of the optical fiber from 1600m to 1700m, respectively, and contains the spectral information of two vibration sources.

[0024] Figure 5 The spectral response of the system to vibrations of different frequencies (taking a vibration source as an example); Figure 5 (a) shows the response spectrum of PZT1 at driving frequencies of 100Hz, 200Hz, and 500Hz. Figure 5 (b) shows the response spectrum of PZT1 at driving frequencies of 1 kHz, 2 kHz, and 5 kHz. Figure 5 In Figures (a) and (b), each frequency contains harmonic components. The experimental results systematically demonstrate that this scheme can achieve the measurement of different frequency components.

[0025] Figure 6 The harmonic orders and corresponding strains in the response spectrum of vibration source PZT1 under different driving voltages are shown. Figure 6 (a) represents the response harmonic number of the fiber optic sensing system. Figure 6 (b) The fiber strain response results calibrated by the grating demodulator.

[0026] Figure 7 Schematic diagram analyzing the causes of harmonic generation. Detailed Implementation

[0027] To describe the present invention more clearly, the following uses an optical time domain reflectometer-type sensing system as an example, but is not limited to such systems. The preferred embodiments of the present invention are described in conjunction with the accompanying drawings.

[0028] The data source for this invention is a distributed optical fiber sensing system. The specific implementation steps for data acquisition and processing in this invention are as follows:

[0029] This embodiment uses a direct-probe distributed fiber optic sensing system to acquire Rayleigh backscattered signals. The front-end system outputs a probe signal, a circulator injects the probe pulse into the fiber under test, and a detector directly detects the backscattered Rayleigh signal, thus achieving the reception and analysis of the measurement signal. The steps include:

[0030] Step 1: The DC optical signal output by the laser is modulated into a probe pulse by the modulator. The probe pulse is amplified by the EDFA and then enters the sensing fiber through the circulator (the use of the EDFA depends on the output power of the laser. When the output power of the laser is high enough, the EDFA can be omitted).

[0031] Step 2: The backscattered signal from the sensing fiber is amplified by an EDFA and then received by the detector. (Whether to use an EDFA depends on the power of the backscattered / reflected signal and the sensitivity of the detector; if the backscattered / reflected signal power is high enough or the detector sensitivity is high enough, an EDFA is not needed.) The system block diagram is as follows: Figure 1 As shown;

[0032] Step 3: Use a data acquisition card to acquire the detector output signal. The data acquisition card acquires and processes n cycles of time-domain data each time, and each time-domain measurement curve contains N data points. Figure 2 (a) contains N data points within each period T. The n consecutive periodic signals of the time-domain signal are assigned to a matrix n*N according to their periods, as shown below. Figure 2 As shown in (b), the matrix is ​​then transposed to N*n (e.g., Figure 2 (as shown in (c)); Figure 2 (c) Each data in each row of the matrix represents the intensity change of the Rayleigh backscattered signal at the same location on the fiber under test over time.

[0033] Step 4, Figure 2 (c) Each row of the matrix undergoes an FFT transformation, representing the impact of vibration on each point on the optical fiber during the data acquisition time. If the optical fiber is not affected by vibration, the spectrum obtained after row-by-row FFT transformation will show no frequencies other than zero. If the optical fiber is affected by vibration, the spectrum will show vibration frequencies from the affected location to the end of the optical fiber, in addition to zero. The non-zero frequency locations are the locations of the vibration source. A sample spectrum of the obtained spectrum in this embodiment is shown below. Figures 3-6 As shown.

[0034] Furthermore, harmonic analysis improves the system's dynamic range by performing a Fourier transform on the collected backscattered / reflected signal intensity information to obtain the time value of the external disturbance information. The disturbance information is then precisely located using the formula z = c * x * Δt / (2n), where z is the position on the optical fiber, c is the speed of light, x represents any point from 0 to N out of N sampling points in one period (the specific position corresponding to z is proportional to x), Δt is the sampling time interval, and n is the refractive index in the optical fiber; N represents... Figure 2 (b) a row of data or Figure 2 A column of data in (c).

[0035] In the embodiment, vibration source 1 (PZT1) is located at 1625m-1640m along the optical fiber, and vibration source 2 (PZT2) is located at 1680m-1691m along the optical fiber.

[0036] Furthermore, the response spectrum after Fourier transform changes with the strength of the vibration signal, and harmonic phenomena appear in the response spectrum, with the harmonic order increasing as the vibration signal strengthens.

[0037] Furthermore, given a sufficiently low noise floor, the strain change occurs for each additional harmonic in the system response. Within the range, i.e., strain resolution Where λ is the incident light wavelength, n is the fiber refractive index, and Δl is the system spatial resolution. Ideally, when λ is 1550 nm, n is 1.45, and Δl is 0.5 m, the corresponding strain value is 1.07 με.

[0038] Furthermore, analyzing the intensity of vibration signals based on the harmonic order in the response spectrum of a vibration sensing system can improve the dynamic range and strain resolution of the sensing system.

[0039] The harmonic data analysis method proposed in this invention is applicable to various optical fiber systems that obtain vibration information based on Fourier transform. The optical fiber mentioned in the sensing system can be a special optical fiber such as single-mode fiber, few-mode fiber, polarization-maintaining fiber, multi-core fiber, multimode fiber, photonic crystal fiber, etc. At the same time, the sensing unit can also be an arrayed fiber grating or an optical fiber containing an array of weak reflection units.

[0040] Based on the experimental results of this embodiment, the causes of harmonic generation can be briefly analyzed, such as... Figure 7 As shown, in coherent detection technology, frequency information is generally obtained through phase deconvolution. The information before deconvolution is as follows: Figure 7 As shown by the solid line, the information after unpacking is as follows: Figure 7 As shown by the dashed line. This invention is based on intensity modulation of the vibration source, without coherent detection. However, due to the sufficiently narrow linewidth of the light source, the self-coherent signal of the detection pulse itself is superimposed on the modulation information. The intensity modulation information acquired by the system is... Figure 7 The signal before unwinding is similar to the solid line in the figure. Assuming the original period of the signal is T and the corresponding frequency is f, based on the period information marked in the figure, it is easy to find that there are multiple periodic components T / 2, T / 4, etc. in the probe signal, which correspond to the frequencies of the harmonic signals 2f, 4f, etc. Therefore, it will be compared with... Figure 7 Data with similar solid lines will exhibit multiple orders of harmonic information after Fourier transform (i.e., Figure 7 The dashed lines in the diagram represent the fundamental frequency data, and the virtual signal also contains multiple harmonic information in addition to the fundamental frequency. The method of this invention allows the use of a low-cost light source on the system laser, eliminates the need for a high-frequency data acquisition card in the signal receiving section, and eliminates the need for electrical or digital filters, significantly reducing system costs and providing a powerful and feasible solution for the widespread adoption of distributed fiber optic sensing technology.

[0041] The above embodiments are preferred embodiments of the present invention, but the protection scope of the present invention is not limited to the above embodiments. Within the knowledge possessed by those skilled in the art, any modifications and partial substitutions made without departing from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. A data analysis method for distributed vibration sensing systems, characterized in that, A distributed optical fiber sensing system is used to construct the front-end system, which outputs a detection signal. A circulator injects the detection signal into the optical fiber under test, and a detector directly detects the backscattered Rayleigh signal to achieve the reception and analysis of the signal under test. The detection signal collected by the detector consists of a series of backscatter / reflection curves, with each detection pulse yielding one backscatter / reflection curve. The length of each backscatter / reflection curve is proportional to the length of the distributed vibration sensing optical fiber. The recording curve matrix is ​​redrawn to provide two-dimensional matrix data of the backscatter / reflection curves, and the output at a specific location is determined as a function of time. By performing a Fourier transform on the two-dimensional matrix data of the Rayleigh backscattering / reverse reflection curves row by row along the distance direction, the perturbation information along the fiber optic sensing line is obtained. Includes the following steps: Step 1: In the distributed fiber optic sensing system, the DC optical signal output by the laser is modulated into periodic probe pulses by the modulator. The probe pulses enter the sensing fiber through the circulator or are amplified by the EDFA and then enter the sensing fiber through the circulator. Step 2: The back Rayleigh scattering / reflection signal of the sensing fiber is received by the detector or amplified by the EDFA and then received by the detector. Step 3: Use a data acquisition card to acquire data from the detector output signal, and perform two-dimensional shaping on the measured one-dimensional data according to the period. Assuming the period of the detection pulse is T, the data acquisition card acquires and processes n periods of time-domain data each time. The time-domain measurement curve of each period contains N data points. Assign the n consecutive periods of time-domain data to a matrix n*N according to the period, and then transpose the matrix to N*n. Step 4: Perform a Fourier transform on the N*n matrix row by row along the distance direction to obtain the disturbance information along the fiber optic sensing line. Harmonic analysis is a method to improve the dynamic range of a system. It involves performing a Fourier transform on the intensity information of the acquired backscattered / reflected Rayleigh signals to obtain the location of external disturbances, based on the formula... z = c * x * ∆t / (2n) To achieve precise location of disturbance information, among which z It's the location on the optical fiber. c It's the speed of light. x Representing a cycle N Of the sampling points: any point from 0 to N refers to the sequential number of uniform sampling along the fiber optic line. z The corresponding specific location and x Proportional ∆t It is the sampling time interval. n It is the refractive index in optical fiber; The spectrum of each row of the N*n matrix obtained by Fourier transform along the distance direction represents the impact of vibration on each point on the optical fiber during the data acquisition time. If the optical fiber is not affected by vibration, no frequency other than zero frequency can be seen in the spectrum obtained after row Fourier transform. If the optical fiber is affected by vibration, in addition to the zero frequency signal, vibration frequencies will appear from the position affected by vibration to the end of the optical fiber. The non-zero frequency position is the position of the vibration source. The Fourier transform response spectrum changes with the intensity of the vibration signal, and harmonic phenomena appear in the response spectrum, with the harmonic order increasing as the vibration signal intensifies. When the noise floor is sufficiently low, for each additional harmonic in the system response, the strain change is... Within the range of με, i.e., strain resolution < με, where λ is the incident light wavelength, n is the fiber refractive index, and ∆l is the system spatial resolution.

2. The data analysis method for distributed vibration sensing systems according to claim 1, characterized in that, The detection pulse width can be in the femtosecond, picosecond, nanosecond, microsecond, or millisecond range depending on the spatial resolution requirements.

3. The data analysis method for distributed vibration sensing systems according to any one of claims 1-2, characterized in that, Analyzing the intensity of vibration signals based on the harmonic order in the response spectrum of a vibration sensing system can improve the dynamic range of the sensing system.

4. The data analysis method for distributed vibration sensing systems according to any one of claims 1-3, characterized in that harmonics... The data analysis method is applicable to various fiber optic sensing systems that acquire vibration information based on Fourier transform. The fiber optics mentioned in the system can be single-mode fiber, few-mode fiber, polarization-maintaining fiber, multi-core fiber, multimode fiber, photonic crystal fiber, and the sensing unit, arrayed fiber grating, or fiber optic containing arrayed weak reflection units.

Images