Signal processing device, vibration detection system, and signal processing method

Abstract

The purpose of the present disclosure is to remove the influence of uncorrected angle differences even in a case where the number of multiplexes divided for reducing fading noise is small and the angle differences between the main optical frequencies cannot be completely corrected. The present disclosure is a signal processing device that acquires a signal obtained by performing DAS-P (Distributed Acoustic Sensing-phase) by repeatedly incident a plurality of optical pulses having different optical frequencies to an optical fiber; calculates a phase change caused by vibration applied to an interval of the optical fiber using the acquired signal; removes a component of a period in which the same optical frequency is repeated among the plurality of optical pulses from the calculated phase change; and calculates vibration applied to the interval of the optical fiber using the phase change from which the component of the period is removed.

Description

Technical Field

[0001] This disclosure relates to a signal processing apparatus and a signal processing method for processing signals obtained by performing DAS-P, and a vibration detection system including the signal processing apparatus. Background Technology

[0002] As a means of measuring physical vibrations applied to an optical fiber in a distributed manner along its length, there is a known method called DAS (Distributed Acoustic Sensing) (Non-Patent Document 1), which involves incident a pulsed test light onto the optical fiber under test and detecting backscattered light caused by Rayleigh scattering.

[0003] In DAS (Digital Optical Array), changes in the optical path length of an optical fiber caused by physical vibrations applied to it are captured, and vibration is sensed and detected. By detecting vibrations, the movement of objects around the fiber being measured can be detected.

[0004] As a method for detecting backscattered light in DAS, there is a method called DAS-I (DAS-intensity) that measures the intensity of scattered light from various points in the fiber under test and observes the time change of the scattered light intensity. DAS-I is characterized by its simple device structure, but since it cannot quantitatively calculate the change in the optical path length of the fiber caused by vibration based on the scattered light intensity, it is a qualitative measurement method (Non-Patent Document 2).

[0005] On the other hand, DAS-P (DAS-phase) is also being researched and developed. This method measures the phase of scattered light from various locations in the fiber being measured and observes the temporal change of the phase. The structure and signal processing of the DAS-P device are more complex than those of DAS-I, but because the phase changes linearly with respect to the change in the optical path length of the fiber caused by vibration, and the rate of change is the same along the length of the fiber, quantitative measurement of vibration can be performed, and the vibration applied to the fiber being measured can be faithfully reproduced (e.g., non-patent literature 2).

[0006] In DAS-P measurements, a pulsed light is incident on the fiber under test. At the moment t of the incident pulse, the phase of the scattered light is measured distributed along the length of the fiber. That is, given a distance l from the incident end of the fiber, the phase θ(l, t) of the scattered light is measured. By repeatedly incidenting the pulsed light onto the fiber under test at time intervals T, the time change θ(l, nT) of the phase of the scattered light at time t = nT is measured at each point along the length of the fiber, where n is an integer. However, in practice, the time of measurement at a distance l is delayed only by the time it takes for the pulsed light to travel from the incident end to a distance l compared to the moment of the incident pulse. Furthermore, it is important to note that the time measured by the measuring instrument is delayed only by the time required for the scattered light to return to the incident end. It is known that the magnitude of the physical vibration applied to the interval from distance l to distance l+δl at each time nT is proportional to the difference θ(l, nT) between the phase θ(l+δl, nT) at distance l+δl and the phase θ(l, nT) at distance l. In other words, if time zero is taken as the reference, the following formula is satisfied.

[0007] [Mathematical Expression 1]

[0008]

[0009] As for the device structure for detecting the phase of scattered light, there are direct detection structures that directly detect backscattered light from the optical fiber being measured using a photodiode or the like, and coherent detection structures that use a combined wave with a separately prepared reference light (e.g., Non-Patent Document 1).

[0010] The mechanisms for performing coherent detection and calculating phase are further divided into two types: those that use Hilbert transform for software-based processing and those that use a 90-degree optical mixer for hardware-based processing. However, in either method, the in-phase component l (l, nT) and quadrature component Q (l, nT) of the scattered light are obtained, and the phase is calculated using the following formula.

[0011] [Mathematical Expression 2]

[0012]

[0013] The output value of the 4-quadrant arctan operator, in radians, is within the range (-π, π]. Let m be any integer. 2mπ + θ(l, nT) becomes the same vector direction in the xy-plane. Therefore, only the uncertainty of 2mπ exists in the calculated θ. cal In (l, nT), further signal processing such as phase expansion is performed as a method to more accurately evaluate θ(l, nT). In general phase expansion, if the expanded phase is set as θ... cal unwrapFor example, if the processing is performed in ascending order of time, then at the starting point of the phase expansion, θ is set... cal unwrap With θ cal Based on the same principle, the following is a sequential process according to θ. cal unwrap Calculate θ from (l, pT) cal unwrap (l,(p+1)T), where p is any integer.

[0014] [Mathematical Expression 2-1]

[0015]

[0016] If the above expression is greater than π radians, choose an appropriate integer q such that the following expression is less than π radians.

[0017] [Mathematical Expression 2-2]

[0018]

[0019] And the expanded phase θ cal unwrap (l, (p+1)T) are calculated sequentially using the following formula.

[0020] [Mathematical Expression 3]

[0021]

[0022] The superscript unwrap indicates the phase after unfolding. In addition, as a calculation step in actual distributed vibration measurement, most of the time, after calculating the difference between the phase values ​​between locations as in equation (1), the phase difference is unfolded.

[0023] In DAS measurements, there is noise from the measuring instrument, including thermal noise from the photodiode used to detect light, noise in the subsequent circuitry, and shot noise caused by light. Therefore, the intensity and phase of the measured scattered light are also affected by the noise of the measuring instrument.

[0024] In particular, when measuring the phase of scattered light, if the noise of the measuring instrument increases, not only will the uncertainty of the phase increase, but the probability of obtaining a measurement value that is significantly different from the ideal phase value in the absence of noise will also increase.

[0025] For example, in the case of coherent detection, regarding the vector of the scattered light measured with the in-phase component as the horizontal axis and the quadrature component as the vertical axis, the direction of the vector in the absence of noise corresponds to the phase to be measured. However, if the influence of noise is large, the direction of the vector is in the opposite direction. Compared with the ideal phase value in the absence of noise, the probability of the actual measured phase value being different by about π radians increases. This leads to the misconception that a large physical force is applied to the optical fiber when calculating the magnitude of the vibration according to Equation (1). Furthermore, if the influence of noise increases, the number of incorrectly selected integer q points increases in the expansion process shown in Equation (3), resulting in a phase value difference of more than 2π before and after the incorrectly selected points that does not actually exist. Such a phase value difference also leads to the misconception that a large physical force is applied to the optical fiber when calculating the magnitude of the vibration according to Equation (1).

[0026] To accurately measure phase, it is necessary to reduce the impact of instrument noise. The impact of instrument noise increases because the intensity of scattered light decreases when the instrument noise is considered to be the same at all locations and times; therefore, increasing the intensity of scattered light at all locations and times can reduce the impact of instrument noise.

[0027] The decrease in scattered light intensity is not solely due to losses caused by absorption and scattering as the probe pulse propagates through the fiber under test. Because a pulse with a finite time width is incident on the fiber under test, and the scattering of the pulse is detected, interference occurs from multiple very fine scatterers distributed along the fiber. As a result of this interference, depending on the distribution of the scatterers along the length of the fiber at each moment, locations where the scattered light intensity decreases are observed. This phenomenon is called fading (Non-Patent Document 3).

[0028] Therefore, when measuring the phase of the scattered light in DAS-P, in order to reduce the influence of the measurement instrument noise, there is a problem that the intensity of the scattered light decreases at various times due to fading.

[0029] One approach to address this problem is to simply increase the peak intensity of the incident light pulse. However, increasing the peak intensity introduces nonlinear effects, causing the characteristics of the pulsed light to change with transmission through the fiber being measured. Therefore, the peak intensity of the incident light pulse is limited and sometimes cannot adequately solve the aforementioned problem.

[0030] To address the aforementioned issues, a phase measurement method and signal processing device (Patent Document 1) are proposed, which can reduce the noise impact of the measuring instrument without increasing the peak intensity of the incident light pulse when measuring the phase of the scattered light in DAS-P.

[0031] To address the aforementioned issues, Patent Document 1 describes a method where, at time intervals that allow for neglecting changes in the fiber's state caused by vibration, pulses of different optical frequency components are arranged and wavelength-multiplexed before being incident on the fiber under test. A scattered light vector is created by plotting the scattered light from each wavelength of the fiber under test on a two-dimensional plane with the in-phase component as the horizontal axis and the quadrature component as the vertical axis. This scattered light vector is then rotated at various locations on the fiber under test according to each wavelength to align its direction. The vectors with aligned directions are then added together and averaged to generate a new vector. The phase is calculated using the values ​​of the in-phase and quadrature components of the generated new vector.

[0032] In DAS-P measurements, there is a trade-off between the measurement distance and the upper limit of measurable vibration frequencies. When using a single-frequency light pulse, if the measurement distance increases, the return time of scattered light from the distant end is delayed relative to the pulse incident time. Therefore, the scattered light from the distant end and the scattered light from near the incident end when the next light pulse is incident will not interfere with each other, thus there is an upper limit to the repetition frequency of the incident light pulse. Therefore, according to the sampling theorem, vibrations with frequencies larger than half the Nyquist frequency of the repetition frequency cannot be measured correctly due to aliasing.

[0033] Non-Patent Document 4 proposes a solution to the aforementioned problem. To address this issue, Non-Patent Document 4 involves multiplexing pulses of different optical frequency components at equal time intervals to obtain pulsed light; creating a scattered light vector by plotting the scattered light from each wavelength of the optical fiber on a two-dimensional plane with the in-phase component as the horizontal axis and the quadrature component as the vertical axis. The phase is then calculated using the obtained scattered light vector. If, in the case of a single optical frequency, the upper limit of the sampling rate determined based on the measurement distance is set to f... s By using N wavelength multiplexing, the upper limit of the sampling rate can be set to N×f. s In addition, the wavelength reuse number "N" is any natural number.

[0034] In the frequency multiplexing method described in Non-Patent Document 4, if the angle difference between each optical frequency is not corrected, and the phase change is simply calculated by connecting the angles of the scattered light vectors obtained at each optical frequency, the calculated phase change will be distorted relative to the actual phase change, making it impossible to measure the correct vibration waveform. To address this problem, Non-Patent Document 4 proposes the following method: First, the time phase difference of each optical frequency is calculated separately, and then the calculated phase differences of each optical frequency are connected together, so that even if the vibration frequency exceeds the Nyquist frequency f in the case of the single frequency, the calculated phase difference of each optical frequency can be connected together. vThe signal can also be correctly estimated in terms of frequency. That is, the frequency N×f can be estimated without aliasing. v However, in the aforementioned proposal, since the angular differences between the aforementioned optical frequencies were not determined, there is a problem that the vibration waveform cannot be measured.

[0035] As a countermeasure to the aforementioned problems, Non-Patent Document 5 proposes a measurement method that corrects the angle difference between the optical frequencies using a correction frequency, thereby increasing the upper limit of the sampling rate to N×f. s The vibration waveform is measured under the condition of [condition missing]. In the proposed method, the angle difference between the components of the main frequency and the components of the compensation frequency is corrected by using a compensation optical frequency different from the main optical frequency used to increase the sampling rate, and by using a sequence of probe pulses that are periodically incident on the fiber under test at regular intervals, treating the components of the main frequency and the components of the compensation frequency as being at the same moment, thereby correcting the angle difference between each main optical frequency.

[0036] Furthermore, in the trade-off between the measurement distance and the upper limit of the measurable vibration frequency, even stricter conditions are added because correct phase unfolding is required. Since phase unfolding cannot be uniquely performed when the absolute value of the phase change when sampling with adjacent light pulses is greater than π, it will lead to the failure of phase unfolding (Non-Patent Document 6).

[0037] Therefore, the upper limit of the absolute value of the phase change between adjacent sampling points creates the limitation of π. Thus, even below the Nyquist frequency, the higher the vibration frequency, the greater the phase change between adjacent sampling points; if the vibration amplitude increases, the upper limit of the measurable vibration frequency becomes even more restrictive. The proposed method described in Non-Patent Document 5 can measure vibration waveforms, thus effectively mitigating this limitation.

[0038] Furthermore, in Non-Patent Document 5, in order to simultaneously implement the frequency multiplexing method of incident pulses of different light frequencies at different times to increase the sampling rate and the frequency multiplexing method described in Patent Document 1 to cope with fading, a method for configuring optical frequency pulses and a method for processing received signals are also proposed.

[0039] Furthermore, the relationship between the magnitude of the phase change and the strain applied to the optical fiber due to vibration is explained, for example, in Non-Patent Document 7. According to Non-Patent Document 7, when the optical fiber of total length l is extended by Δ1 due to the strain ε, the increase in phase change Δφ caused by the extension of Δ1 during light transmission is shown in the following formula.

[0040] [Mathematical Expression 4]

[0041]

[0042] Where k = 2πn / λ is the propagation constant, n is the effective refractive index of the optical fiber, and μ p For Poisson's ratio, p 11 and p 12 For the strain-optical tensor components. For example, consider the case of λ = 1555 nm near a typical communication band, since n = 1.47, μ p =0.17, p 11 =0.121, p 12 The value of =0.27l is known and is shown in the following formula (Non-Patent Document 8).

[0043] [Mathematical Expression 5]

[0044]

[0045] Where, K = 4.6 × 10 6 m -1 Using this relationship, the condition for the magnitude of phase change can be replaced with the condition for the amount of distortion.

[0046] In the frequency multiplexing method described in Non-Patent Document 5, the total number of usable optical frequencies, which is the sum of the number of multiplexings allocated to increase the sampling rate and the number of multiplexings allocated to reduce fading noise, can be evaluated as the number of usable total bandwidths determined by the sampling rate of the A / D (Analog / Digital) converter (hereinafter sometimes called an AD board) that converts analog signals to digital signals, divided by the number of occupied bandwidths of each component determined by the pulse width of each optical frequency component. Therefore, the number of multiplexings allocated to reduce fading noise decreases in the following situations: when detecting high-frequency vibrations, the number of multiplexings allocated to increase the sampling rate increases; when an AD board with a low sampling rate is required, limiting the usable total bandwidth; and when high spatial resolution measurements are required, increasing the occupied bandwidth of each component to reduce the pulse width of each optical frequency component.

[0047] When the number of multiplexes allocated to reduce fading noise is reduced, the following problems arise: the angular differences between the main light frequencies recorded in the background cannot be completely corrected, resulting in residual angular differences between the main light frequencies and locations where the calculated phase change is still distorted relative to the actual phase change. When the residual angular differences between the main light frequencies are large, phase connection errors also increase. Furthermore, by using the incident compensation light frequency, there is a problem of crosstalk between the compensation light frequency and the main light frequency, leading to distortion and further causing phase connection errors.

[0048] Existing technical documents

[0049] Patent documents

[0050] Patent Document 1: Japanese Patent Application Publication No. 2020-169904

[0051] Non-patent literature

[0052] Non-patent document 1: Ali. Masoudi, TP Newson, "Contributed Rview: Distributed optical fiber dynamic strain sensing." Review of Scientific Instruments, vol. 87, pp011501 (2016).

[0053] Non-patent literature 2: Kenichi Nishiguchi, Zhexian Li, Kwata Koji, Mitsunori Yokoyama, and Yoshinori Masuda, “Prototype and Signal Processing of Fiber Optic Distributed Acoustic Sensor” Journal of Information Science and Technology, 115(202), pp29-34 (2015).

[0054] Non-patent literature 3: G. Yang et al., "Long-Range Distributed Vibration Sensing Based on Phase Extraction from Phase-Sensitive OTDR", IEEE Photonics Journal, vo1.8, no.3, 2016.

[0055] Non-patent document 4: D.Iida, K.Toge, T.Manabe, 'Distributed measurement of acoustic vibration location with frequency multiplexed phase-OTD', Opt.FiberTechnol., 2017, 36, pp19-25, DOI: 10.1016 / j.yofte.2017.02.005.

[0056] Non-patent document 5: Y. Wakisaka, D. Iida and H. Oshida, "Distortion-SuppressedSampling Rate Enhancement in Phase-OTDR Vibration Sensing with Newly DesignedFDM Pulse Sequence for Correctly Monitoring Various Waveforms," ​​2020 OpticalFiber Communications Conference and Exhibition (OFC), SanDiego, CA, USA, 2020, PP.1-3.

[0057] Non-patent literature 6: Maria Rosario Fernandez-Ruiz, Hugo F. Martins, "Steady-Sensitivity Distributed Acoustic Sensors", J. Lightwave Technol. 36, 5690-5696 (2018).

[0058] Non-patent literature 7: CDButter and GBHocker, "Fiber optics strain gauge", Appl. Opt. 17, 2867-2869 (1978)

[0059] Non-patent literature 8: AE Alekseev et al., "Fidelity of the dual-pulse phase-OTDRresponse to spatially distributed extema lperturbation," LaserPhys. 29, 055106 (2019) Summary of the Invention

[0060] The problem the invention aims to solve

[0061] The purpose of this disclosure is to remove the effects of uncorrected angle differences even when the number of multiplexings allocated to reduce fading noise is small and the angle differences between the main optical frequencies cannot be fully corrected.

[0062] The means used to solve the problem

[0063] This disclosure proposes a method to remove the influence of the residual angle difference when it is impossible to completely correct the angle difference between each main optical frequency and the angle difference between each main optical frequency remains. This method also reduces phase connection errors, thereby making the calculated phase change closer to the actual phase change.

[0064] Specifically, this disclosure relates to a frequency-reused phase OTDR, specifically the phase change δθ0(l, mT) generated in the interval from a distance l to l+δl from the incident end. N δθ0 after phase connection processing unwrap (l, mT) N ), from δθ0(l, mT N Subtract the period NT extracted using a multi-bandpass filter from the original data. N The components D(l, mT) N ).

[0065] Specifically, the signal processing apparatus and signal processing method disclosed herein...

[0066] The signal obtained by repeatedly incidenting multiple optical pulses of different frequencies onto an optical fiber and performing DAS-P is acquired.

[0067] Using the acquired signal, the phase change caused by the vibration applied to the section of the optical fiber is calculated;

[0068] From the calculated phase change, remove the components of the periodic repetition of the same optical frequency in the plurality of optical pulses;

[0069] The vibration of the region applied to the optical fiber is calculated using the phase change with the periodic component removed.

[0070] Specifically, the vibration detection system disclosed herein includes:

[0071] The measuring device incident multiple light pulses of different frequencies onto one end of an optical fiber and receives the scattered light of each wavelength returning to that end of the optical fiber; and,

[0072] The signal processing apparatus disclosed herein uses a signal from the measuring instrument to calculate the vibration of a region applied to the optical fiber.

[0073] The program disclosed herein is a program for enabling a computer to implement the various functional units of the signal processing apparatus of this disclosure, and a program for enabling a computer to execute the various steps of the signal processing method executed by the signal processing apparatus of this disclosure.

[0074] Invention Effects

[0075] According to this disclosure, even when the number of multiplexing divisions for reducing fading noise is small and the angle difference between each master optical frequency cannot be completely corrected, the influence of the uncorrected angle difference can still be removed. Attached Figure Description

[0076] Figure 1 This is a diagram illustrating the vibration detection device using DAS-P for vibration detection according to this embodiment.

[0077] Figure 2 Example of a structure representing a pulse mode.

[0078] Figure 3 This is an example of a correction method for distortion representing phase changes.

[0079] Figure 4 This is an example of a method for correcting distortion that accompanies crosstalk. Detailed Implementation

[0080] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. However, the present disclosure is not limited to the embodiments shown below. These examples are merely illustrations, and the present disclosure can be implemented in various modified and altered ways based on the knowledge of those skilled in the art. Furthermore, in this specification and the accompanying drawings, the same reference numerals denote identical constituent elements.

[0081] Furthermore, this disclosure is effective not only when distortion removal processing is performed in advance using the compensation light frequency described in Non-Patent Document 5 before performing the signal processing proposed in this disclosure, but also when the angle difference between the main light frequencies remains completely even without using the compensation light frequency. In addition, this disclosure also proposes a method that, even when using the compensation light frequency as described in Non-Patent Document 5, can reduce the distortion associated with crosstalk caused by the compensation light frequency.

[0082] (Implementation Method 1)

[0083] Figure 1 This diagram illustrates a vibration detection system using DAS-P for vibration detection according to this embodiment. The vibration detection system includes: a light source that incident a frequency-multiplexed sequence of light pulses onto one end of a fiber optic cable being measured; a light receiver that receives scattered light of various wavelengths returning to the one end of the fiber optic cable being measured; and a signal processing unit that observes the vibration of the fiber optic cable being measured as a time-varying phase component of the scattered light.

[0084] CW light source 1, coupler 2, and optical modulator 3 correspond to the light source. 90-degree optical mixer 7 and balanced detectors (13, 14) correspond to the light receiver. The light receiver uses the 90-degree optical mixer 7 for coherent detection. Signal processing device 17 corresponds to the signal processing unit. However, the receiving system does not necessarily need to use a 90-degree optical mixer; other devices and signal processing can be used as long as the in-phase and quadrature components of the scattered light can be measured. Furthermore, the signal processing device of this disclosure can be implemented by a computer or program, and the program can be stored in a storage medium or provided via a network.

[0085] The measuring device 31 measures the scattered light from the fiber 6 being measured as follows. A continuous beam of a single wavelength with a frequency of f0 is emitted from the CW light source 1 and branched into a reference beam and a probe beam by the coupler 2. The probe beam is shaped into a wavelength-multiplexed optical pulse 4 by the optical modulator 3. The optical pulse 4 can be a multiplexed pulse using the compensation optical frequency described in Non-Patent Document 5. The optical pulse 4 only needs to implement the compensation method described in Non-Patent Document 5, and its structure is exemplified in… Figure 2 As shown in the image.

[0086] Let the optical frequency components used in the main pulse be f1 to f2. NM There are a total of N×M groups, and N+1 groups (201) are prepared in sequence. The overall arrangement is shown in 202. Divide the group into groups of M (M is any natural number) to generate N(N+1) pulse pairs. For 202, a compensation optical frequency f is added to each N+1 pulse pair. NM+1 Pulse pairs 203 are generated. Based on pulse pairs 203, the sequence of actual incident light pulse pairs is arranged as shown in 204. Thus, a pulse pattern in which N(N+1) pulse pairs are arranged at a certain time period is generated.

[0087] Here, because the compensation optical frequency f NM+1 The pulse pairs numbered 1+k(N+1) (k=0, 1, …, (N-1)) are added to the pulse pairs. Therefore, for example, in the case of N=3 and M=1, pulse pairs with optical frequencies f1, f2, and f3 are repeatedly incident on the fiber 6 being measured. In this case, when k=0, the compensating optical frequency f4 is added to the pulse pair with optical frequency f1; when k=1, the optical frequency f4 is added to the pulse pair with optical frequency f2; and when k=2, the optical frequency f4 is added to the pulse pair with optical frequency f3.

[0088] If the interval between the pulse pairs is set to T N The period of the pulse mode is then N(N+1)T. N Compared to the case using a single-frequency pulse, the T value is influenced by the length of the measured fiber 6. NThe limitation on the minimum value to which it can be reduced is eased by a factor of 1 / N. Furthermore, in Figure 2 In the structure, M pulses present in each pulse pair can be used to reduce fading noise. In addition, although the compensation light frequency corresponds to one in the pulse pair 203, multiple compensation light frequencies can be multiplexed to suppress fading.

[0089] exist Figure 1 In this context, the type of optical modulator 3 is not specifically limited as long as it can generate optical pulses 4, and sometimes there are multiple modulators. For example, an SSB (Single Sideband) modulator or a frequency-variable AO (Acousto-Optics) modulator can be used. To further increase the extinction ratio in pulsed light, intensity modulation of an SOA (Semiconductor Optical Amplifier) ​​can also be performed. Furthermore, although the pulses of each optical frequency component shown in 204 are rectangular wave shapes, waveforms other than rectangular waves can also be used.

[0090] Optical pulse 4 is incident on the fiber 6 under test via circulator 5. The light scattered at various points along the length of fiber 6 returns to circulator 5 as backscattered light and is incident on one input of 90-degree optical mixer 7. The reference light branched by coupler 2 is incident on the other input of 90-degree optical mixer 7.

[0091] The internal structure of the 90-degree optical mixer 7 only needs to possess the functionality of a 90-degree optical mixer; it can be arbitrary. An example of its structure is shown below. Figure 1 As shown in the diagram, backscattered light is incident on coupler 8 with a 50:50 branch ratio, and the backscattered light after two branches is incident on the input sections of couplers 12 and 11 with a 50:50 branch ratio. Reference light is incident on coupler 9 with a 50:50 branch ratio, and one of the reference lights after two branches is incident on the input section of coupler 11, while the other is phase-shifted by π / 2 by phase shifter 10 and incident on the input section of coupler 12.

[0092] The two outputs of coupler 11 are detected by balance detector 13, which outputs the simulated in-phase component I. analog That is, electrical signal 15. The two outputs of coupler 12 are detected by balanced detector 14, and the output is an analog quadrature component Q. analog That is, electrical signal 16.

[0093] Electrical signals 15 and 16 are sent to a signal processing device 17 equipped with AD conversion elements 17a and 17b capable of sampling the optical frequency band of the signals without aliasing. In the signal processing device 17, the digitized in-phase component I output from AD conversion elements 17a and 17b is processed... digital and orthogonal components Q digital The signal is separated by the signal processing unit 17c to form the optical frequencies f0+f of the optical pulse 4. i The signal is located in the frequency band (i = 1, 2, ..., NM+1). The specific signal processing method only needs to be able to process the signal from I... digital and Q digital Correctly separating the signals of each frequency band is I i measure (i = 1, 2, ..., NM+1) and Q i measure (i = 1, 2, ..., NM+1) is sufficient; any method can be used. For example, one could consider making I... digital and Q digital Passing through the center frequency f0+f i Methods for calculating phase delay using bandpass filters, etc. If the pulse width of each optical frequency component is set to W, the passband can be set to 2 / W. Alternatively, after separating the in-phase and quadrature components of the analog electrical signal into optical frequency components using an analog electrical filter, AD conversion can be performed using AD conversion elements 17a and 17b, etc.

[0094] Signal processing unit 17d is based on I obtained by signal processing unit 17c i measure and Q i measure To calculate the phase, first, create a complex vector r on the xy plane with the in-phase component as the x-axis (real axis) and the orthogonal component as the y-axis (imaginary axis). i .

[0095] [Mathematical Expression 1-1]

[0096]

[0097] Let the start time of the incident pulse pair k be k×T. N +n×N×T N(n is any integer). Taking the initial optical frequency of each pulse pair as the reference wavelength, according to the method described in Patent Document 1, the vector calculated using equation (1-1) in the frequency bands of M different optical frequencies other than the compensation optical frequency constituting the pulse pair is averaged to calculate the phase at a distance z from the incident end. The state of the measured optical fiber 6 at a distance Z from the incident end in the longitudinal direction of the measured optical fiber 6 is considered at time k×T, taking into account the propagation time of the optical pulse. N +n×N×T N The measurement is performed using +z / v (n is any integer). Here, v is the speed of light in the measured fiber 6. Furthermore, considering the time it takes for the scattered light to propagate and return to the incident end, the measurement time is k×T. N +n×N×T N +2z / v (n is any integer). Therefore, the phase calculated from the location at a distance z, with the measurement time of the measuring instrument as the positive, is set as:

[0098] [Mathematical Expression 1-2]

[0099]

[0100] In this embodiment, mT is used N +2z / v=kT N +nNT N k and n for +2z / v are calculated as follows for the measurement time mT. N The phase θ(z, mT) of +2z / ν (m is an integer) N +2z / v).

[0101] [Mathematical Expressions 1-3]

[0102]

[0103] Then, the phase change caused by the vibration applied to the interval from distance z1 to distance z2 on the fiber 6 to be measured is calculated as the difference between mathematical formulas (1-3a) and (1-3b), namely mathematical formula (1-3c).

[0104] [Mathematical expression 1-3a]

[0105]

[0106]

[0107]

[0108] Furthermore, since the instant at which the state of the measured fiber 6 is measured does not include the time required for the scattered light to return to the incident end, as mentioned above, at a distance z1 and a time mT N +z1 / v, at a distance of z2 and time mT N +z2 / v, only the time difference (z1-z2) / v is different. However, since the difference in distance between z1 and z2 is equivalent to the spatial resolution, it is usually set to a few meters to tens of meters. Therefore, the time difference (z1-z2) / v is tens to hundreds of ns, which is very short relative to the scale of the time change of the usual vibration of the object being measured. Thus, the time difference in measuring the state of the fiber 6 can be ignored. Therefore, the vibration applied in this interval can be measured correctly.

[0109] However, θ(z, mT) N +2z / v) contains distortion terms caused by the angle difference between the starting optical frequencies of each pulse pair. Non-Patent Document 5 proposes a method for correcting the distortion terms caused by the angle difference when using a compensated optical frequency. In order to correct the distortion terms caused by the angle difference between different optical frequencies without omission, it is necessary to perform angle difference correction on the starting optical frequencies of any two pulse pairs. When arbitrarily choosing positive integers i and j that satisfy i < j, if the starting optical frequency of pulse pair j is set to f j pf Set the optical frequency of pulse pair i to f i pf Then you can use f NM+1 The difference in unfolding angles φ(z, f) j pf f i pf ),as follows.

[0110] [Mathematical Expressions 1-4]

[0111]

[0112] i and j are any positive integers; where i < j.

[0113] In the combination 203 of the light frequencies of the pulse pair used as an example, due to the light frequency f NM+1 Added to the pulse pair numbered 1+k(N+1) (k=0,1,…,(N-1)), therefore the optical frequency f NM+1 Other light frequencies in period N(N+1)T NEach of the frequencies f1, f2, and f3 must exist once in the same pulse pair. For example, when N=3 and M=1, the number of pulse pairs constituting the pulse pattern is 12. In this case, the first pulse pair contains optical frequencies f1 and f4, the fifth pulse pair contains optical frequencies f2 and f4, and the ninth pulse pair contains optical frequencies f3 and f4. Therefore, in the pulse pattern, each of the optical frequencies f4 and f1, f2, and f3 must exist once in the same pulse pair. Therefore, the terms on the right side of mathematical expression (1-4) can be calculated using the same principle as the method in Patent Document 1. Using the obtained φ(z, f j pf f i pf The value of ) is determined by θ(z, mT). N +2z / v) is used to calculate the phase. For example, in calculating from time m'T N +2z / v to mT N In the case of a phase change of +2z / ν, mathematical formula (1-5) can be used.

[0114] [Mathematical Expressions 1-5]

[0115]

[0116] Here, an integer i(m') is selected such that m'-i(m') is an integer multiple of N, and an integer i(m) is selected such that mi(m) is an integer multiple of N. As a practical calculation step, to calculate the phase change between two locations by calculating the phase difference between the two locations, the phase change between the distances from the incident end l and l+δl is shown in equation (1-3c). If time zero is taken as a reference, equation (1-6) can be used. On the left side of equation (1-6), δθ is taken as the difference between the two locations, hence the sign of the increment δ, and the subscript 0 indicates that time zero is taken as the reference. Furthermore, on the left side, the delay accompanying light propagation in the measured fiber 6 is simplified in a non-sunlit form.

[0117] [Mathematical Expressions 1-6]

[0118]

[0119] Up to this point, signal processing has been performed by the signal processing unit 17d.

[0120] The signal processing unit 17e calculates the final phase. In this disclosure, the steps of the signal processing unit 17e differ from conventional methods.

[0121] In previous methods, the above δθ0(l, mT) N The phase connection process is performed on δθ0(l, mT) to obtain the final vibrational change.N δθ0 obtained by phase connection processing unwrap (l, mT) N The final vibration waveform is obtained by dividing the signal into multiplexes (e.g., φ(z, f)). The superscript "unwrap" indicates the phase connection after processing. However, in cases such as detecting high-frequency vibrations, the number of multiplexes divided to increase the sampling rate increases. A low-sampling-rate AD board is required, limiting the total usable bandwidth. For measurements requiring high spatial resolution, the occupied bandwidth of each frequency component increases to reduce the pulse width of each component. In these cases, even with a reduced number of multiplexes to reduce fading noise, the following problem arises because the fading noise cannot be sufficiently reduced: the angle difference φ(z, f) j pf f i pf The deterioration of the estimation accuracy makes it impossible to completely correct the angle difference between the principal light frequencies recorded in the background. The distortion caused by the angle difference between the principal light frequencies remains, resulting in locations where the calculated phase change is still distorted relative to the actual phase change.

[0122] In this disclosure, as a countermeasure to the aforementioned problem, a calculation is performed to remove the distortion remaining after processing in the signal processing unit 17d. The steps are described in... Figure 3 middle.

[0123] The phase change δθ0(l, mT) generated in the interval from a distance l to l+δl from the incident end N Perform phase connection processing and calculate δθ0. unwrap (l, mT) N (S101).

[0124] Using a bandpass filter, for δθ0 unwrap (l, mT) N Extraction cycle NT N The components, as D(l,mT) N (S102).

[0125] Before phase connection, δθ0(l,mT) N Subtract D(l, mT) N ) after δθ0(l, mT N )-D(l,mT N ), updated to the new δθ0(l, mT) N (S103).

[0126] For the new δθ0(l, mT) N Phase connection processing is performed (S101).

[0127] If the distortion is not sufficiently removed (S104), repeat the steps.

[0128] The residual distortion is removed through these steps.

[0129] As a bandpass filter is used and the period NT is utilized N The components D(l, mT) N The reason for removing the distortion is that the distortion is periodically NT. N Changes have occurred. Period NT N It is the period of repetition of the same optical frequency in multiple optical pulses, for example, the frequency f at the beginning of pulse pair i(m). i(m) pf The switching period. If φ(z, f) in signal processing unit 17d... j pf f i pf The difference between the estimated value and the actual value of ) is expressed as δφ(z, f j pf f i pf According to equation (1-6), the δθ0(l, mT) calculated in the signal processing unit 17d is... N The residual distortion in the ) is:

[0130] [Mathematical Expression 7]

[0131]

[0132] Although the distortion shown in equation (7) depends on the frequency f at the beginning of the pulse pair i(m) i(m) pf However, the frequency f at the beginning of the pulse pair i(m) i(m) pf With period NT N Switching. Therefore, since the distortion shown in equation (7) is in period NT N The changing components, therefore, are derived from δθ0. unwrap (l, mT) N Extracting NT from the periodic N The changing components can be used to estimate the distortion shown in equation (7).

[0133] It should be noted that since the distortion term does not necessarily change in a sinusoidal waveform, all components that should pass through the bandpass filter must have a frequency of 1 / NT. N The higher harmonic components are integer multiples of the Nyquist frequency. Even for components exceeding the Nyquist frequency, aliasing should be considered, and bandpasses should be applied at the locations of components on the frequency axis that return to the Nyquist frequency range. However, due to the fact that 1 / NT in the frequency range does not cause aliasing... NThe higher harmonic components overlap with the positions on the frequency axis; therefore, using 1 / NT within a frequency range that does not substantially cause aliasing is appropriate. N All high-order harmonic components can be filtered through a multi-bandpass filter.

[0134] The phase-connected δθ0 is obtained by using a bandpass filter. unwrap (l, mT) N Instead of δθ0(l, mT) before phase connection N Extraction of periodic NT N The reason for the change in components is that, before the phase connection, δθ0(l, mT) N ) is convolved with the phase values ​​from -π to +π. Since the distortion shown in Equation (7) is added and convolved in the actual vibration change, the distortion shown in Equation (7) cannot be extracted by the bandpass filter.

[0135] Furthermore, for δθ0(l, mT) calculated by the signal processing unit 17d N To calculate D(l, mT) N The operation of obtaining the difference sometimes cannot fully remove the distortion in just one operation because: the distortion shown in factor (7) is large and δθ0 unwrap (l, mT) N Multiple phase connection errors occurred at various locations within the NT sequence, and the steep phase changes caused by these errors also exhibited periodicity. N The components, therefore even through a bandpass filter, δθ0 unwrap (l, mT) N Extraction cycle NT N The components cannot correctly presume the distortion shown by equation (7). Therefore, based on the following principle, namely, by means of... Figure 3 The iterative calculations shown can reduce the number of steps in δθ0. unwrap (l, mT) N The phase connection error that occurs in ) can improve the estimation accuracy of the distortion shown by equation (7).

[0136] In step S104, regarding the determination of whether the distortion has been sufficiently removed, if the number of phase connection errors can be counted, a time point where the number of phase connection errors no longer changes after the aforementioned steps can be selected. If this is not possible, since the steep phase change caused by phase connection errors increases the noise on the low-frequency side of the measurement band, a time point where the reduction in low-frequency noise cannot be observed after the aforementioned steps can also be selected. As a specific value for evaluation from verification experiments, setting the number of cycles to around 10 is sufficient.

[0137] In the above embodiments, an example of applying this disclosure using a compensated optical frequency has been shown. However, this disclosure can be applied according to the same principle even without using a compensated optical frequency. Without using a compensated optical frequency, regarding θ(z,mT) before distortion correction is performed on the signal using the compensated optical frequency... N The phase change of the difference between the two locations was calculated using +2z / v, and this disclosure is used.

[0138] When using a compensated optical frequency for measurement, although distortion due to crosstalk between the compensated optical frequency and the main optical frequency will still occur, it can also be considered for removal. The steps for removing distortion due to crosstalk between the compensated optical frequency and the main optical frequency are as follows: Figure 4 As shown in the image.

[0139] The phase change δθ0(l, mT) generated in the interval from a distance l to l+δl from the incident end N Perform phase connection processing and calculate δθ0. unwrap (l, mT) N (S201).

[0140] Using a bandpass filter, for δθ0 unwrap (l, mT) N Extraction cycle NT N The components, as D1(l,mT) N (S202). Additionally, a bandpass filter is used to filter from δθ0. unwrap (l, mT) N )-D1(l,mT N Extracting the period N(N+1)T from ) N The components, as D2(l,mT) N (S202).

[0141] δθ before phase connection o (l, mT) N Subtract D1(l, mT) N ) and D2(l,mT N ) after δθ0(l, mT N )-D1(l,mT N )-D2(l,mT N ), updated to the new δθ0(l, mT) N (S203).

[0142] For the new δθ0(l, mT) N Perform phase connection processing and calculate δθ0. unwrap (l, mT) N (S201).

[0143] If the distortion is not sufficiently removed (S204), repeat the steps.

[0144] These steps also allow for the removal of distortion caused by crosstalk between the compensation light frequency and the main light frequency.

[0145] After using D1(l, mT) N In the processing of ), and with the use of D(l, mT) N The processing principle is the same as that of D(l, mT), which uses the aforementioned D(l, mT) N The processing of D2(l,mT) is used to remove the distortion caused by the angle difference between the main light frequencies that was not removed during the processing of the signal processing unit 17d. On the other hand, D2(l,mT) is used. N The processing corresponds to removing the distortion associated with crosstalk between the compensation light frequency and the main light frequency. The distortion associated with crosstalk between the compensation light frequency and the main light frequency occurs with respect to the period of the pulse mode used. Since the period of the pulse mode is as follows... Figure 2 The figure shows N(N+1)T N Therefore, by using a bandpass filter to filter δθ0 unwrap (l, mT) N Extraction period N(N+1)T N The components can be used to infer the distortion that accompanies the crosstalk.

[0146] Furthermore, since the distortion accompanying the crosstalk is not a sinusoidal waveform, in actual bandpass filtering, it is necessary to use a frequency of 1 / {N(N+1)T}. N A multi-bandpass filter that allows all higher-order harmonic components corresponding to integer multiples of} to pass through. Among these, regarding 1 / {N(N+1)T N The higher harmonic components corresponding to integer multiples of} are because they contain 1 / (NT) N The higher harmonic components of D2(l, mT) are therefore required to be calculated using a multi-bandpass filter. N ), so that it will be related to 1 / {N(N+1)T N The higher harmonic components corresponding to integer multiples of} have had 1 / (NT) removed. N All components of the higher harmonics of δθ0 are bandpassed. Therefore, in the above steps, a bandpass filter is used to filter from δθ0. unwrap (l, mT) N )-D1(l,mT N Extracting the period N(N+1)T from ) N The components, as D2(l,mT) N Or, because in Figure 4 The steps shown do not use D1(l, mT) alone. N ) and D2(l,mTN Therefore, it is possible to use 1 / {N(N+1)T} N The multi-bandpass filter corresponding to all higher harmonic components of integer multiples of} is extracted as D1(l,mT) N )+D2(l,mT N In practical signal processing, this is a simpler method.

[0147] Regarding δθ0 unwrap (l, mT) N The reasons for performing bandpass filtering and the reasons for iterative calculations are related to... Figure 3 The steps described are the same.

[0148] Furthermore, this disclosure is not limited to the original embodiments described above, and the configuration elements can be modified and specified during the implementation phase without departing from its purpose.

[0149] Industrial applicability

[0150] This disclosure can be applied to the information and communication industry.

[0151] Explanation of reference numerals in the attached figures

[0152] l: CW light source

[0153] 2: Coupler

[0154] 3: Optical modulator

[0155] 4: Light Pulse

[0156] 5: Circulator

[0157] 6: The fiber being measured

[0158] 7: 90-degree optical mixer

[0159] 8, 9: Couplers

[0160] 10: Phase shifter

[0161] 11, 12: Couplers

[0162] 13, 14: Balance detector

[0163] 15: Analog in-phase component electrical signal

[0164] 16: Analog quadrature component electrical signals

[0165] 17: Signal processing device

[0166] 17a, 17b: AD conversion elements

[0167] 17c, 17d: Signal Processing Unit

[0168] 31: Measuring instrument.

Claims

1. Signal processing device, The signal is obtained by repeatedly incidenting multiple optical pulses of different frequencies onto an optical fiber and performing DAS-P (Distributed Acoustic Sensing-phase). Using the acquired signal, the phase change caused by the vibration applied to the section of the optical fiber is calculated; From the calculated phase change, remove the components of the periodic repetition of the same optical frequency in the plurality of optical pulses; The vibration of the region applied to the optical fiber is calculated using the phase change with the periodic component removed.

2. The signal processing apparatus according to claim 1, wherein the plurality of optical pulses are composed of a plurality of pulse pairs including a component of the main optical frequency and a component of the compensation optical frequency; The phase change in the optical fiber is calculated using the optical phase measured at the main optical frequency included in the plurality of pulse pairs. From the calculated phase change, remove the components of the periodic repetition of the same optical frequency in the plurality of optical pulses; The angle difference between the component of the main optical frequency (after removing the periodic component of the same optical frequency repetition in the plurality of optical pulses) and the component of the compensation optical frequency is corrected to the component of the main optical frequency (after removing the periodic component of the same optical frequency repetition in the plurality of optical pulses).

3. The signal processing apparatus according to claim 2, Remove the periodic component of the pulse mode of the pulse pair from the calculated phase change; The angle difference between the component of the main light frequency (after removing the periodic component of the pulse mode of the pulse pair) and the component of the compensation light frequency is corrected to the component of the main light frequency (after removing the periodic component of the pulse mode of the pulse pair).

4. Vibration detection system, equipped with: The measuring device incident multiple light pulses of different frequencies onto one end of an optical fiber and receives the scattered light of each wavelength returning to that end of the optical fiber; and, The signal processing apparatus according to any one of claims 1 to 3 uses a signal from the measuring instrument to calculate the vibration of a region applied to the optical fiber.

5. Signal processing methods, The signal is obtained by repeatedly incidenting multiple optical pulses of different frequencies onto an optical fiber and performing DAS-P (Distributed Acoustic Sensing-phase). Using the acquired signal, the phase change caused by the vibration applied to the section of the optical fiber is calculated; From the calculated phase change, remove the components of the periodic repetition of the same optical frequency in the plurality of optical pulses; The vibration of the region applied to the optical fiber is calculated using the phase change with the periodic component removed.

Images