JP5493089B2 - Distributed optical fiber sensor - Google Patents

Description

  The present invention relates to a distributed optical fiber sensor that uses an optical fiber as a sensor and can measure strain and temperature with high accuracy in the longitudinal direction.

  Conventionally, as a technique for measuring strain and temperature, there is a method based on the Brillouin scattering phenomenon that occurs in an optical fiber. In this method, the optical fiber is used as a medium for detecting strain and / or temperature in the environment (measurement object) in which the optical fiber is installed.

  The Brillouin scattering phenomenon is a phenomenon in which power moves through an acoustic phonon in an optical fiber when light enters the optical fiber, and two lights having different frequencies are incident on the optical fiber. There is a stimulated Brillouin scattering phenomenon caused by the interaction of light, and a natural Brillouin scattering phenomenon caused by the interaction of this light with acoustic phonons caused by thermal noise in the optical fiber. The Brillouin frequency shift seen during this Brillouin scattering phenomenon is proportional to the speed of sound in the optical fiber, and the speed of sound depends on the strain and temperature of the optical fiber. For this reason, strain and / or temperature is measured by measuring the Brillouin frequency shift.

  As a typical method for measuring strain and temperature distribution using the Brillouin scattering phenomenon, there is a BOTDR (Brillouin Optical Time Domain Reflectometer) described in Patent Document 1.

  In BOTDR, one laser beam is incident as pump light (inspection pulse light) from one end of an optical fiber, light related to the natural Brillouin scattering phenomenon emitted from this one end is detected, and the detected natural Brillouin scattering is detected. The light intensity of the light related to the phenomenon is measured in the time domain. In this BOTDR, acoustic phonons generated by thermal noise are used, and pulse light having a rectangular light intensity is usually used as the pump light.

  Such measurement is performed for each frequency while sequentially changing the frequency of the pump light, that is, by scanning the frequency, the power spectrum at each position along the longitudinal direction of the optical fiber (hereinafter, simply referred to as “power spectrum”). (Referred to as “Brillouin spectrum”), and the strain distribution and / or temperature distribution along the longitudinal direction of the optical fiber is measured from the Brillouin frequency shift amount.

Japanese Patent No. 2575794

  However, in the above BOTDR, it is difficult to improve both spatial resolution and frequency resolution. That is, in order to improve the spatial resolution, the pulse width of the pump light must be shortened, but the spectrum of light related to the Brillouin scattering phenomenon observed using the pump light with a shortened pulse width is expanded (ie, Therefore, the Brillouin frequency shift amount cannot be accurately measured.

  Specifically, if the Brillouin spectrum obtained by scanning the frequency ν is V (t, ν), the Brillouin spectrum is

It is represented by Here, γ _R is a constant, and “ ^{t, ν} *” represents a two-dimensional convolution (convolution) with respect to t, ν. G (t, ν) in equation (1) is a Lorentz spectrum, and when the Brillouin frequency shift at the position z of the optical fiber is ν _b (z),

It is represented by Here, v _g represents the speed of light in the optical fiber.

  In addition, Ψ (t, ν) in equation (1) is a point spread function,

It is represented by Here, f (t) is a function representing a pulse shape, h (t) is an impulse response of a low-pass filter in the receiving unit, and Fτ [f (τ) h (t−τ)] is , Τ represents the Fourier transform.

  The Lorentz spectrum G (t, ν) is a spectrum determined only by the characteristics of the optical fiber without depending on the observation, and the point spread function Ψ (t, ν) is a function determined depending on the measurement method.

  Here, the ideal Brillouin spectrum to be obtained is that the Lorentz spectrum G (t, ν) itself (see FIG. 4B) is observed, but in actual observation, it is expressed by Equation (1). In this way, a Lorentz spectrum G (t, ν) blurred by the convolution with the point spread function Ψ (t, ν) (see FIG. 15B) is observed.

  Therefore, in order to improve the spatial resolution, it is desirable that the point spread function Ψ (t, ν) has a width as narrow as possible in both time and frequency.

  However, by integrating the point spread function Ψ (t, ν), the spread functions in the time direction and the frequency direction are respectively obtained.

And each center μ _T , μ _F and radius Δ _T , Δ _F are defined as

The following equation (8) is established.

This represents an uncertainty relationship in which the spread of the point spread function Ψ (t, ν) in the time direction and the frequency direction cannot be reduced simultaneously.

  Therefore, an object is to provide a distributed optical fiber sensor having high spatial resolution and high frequency resolution.

  Accordingly, in order to solve the above-described problems, the present invention is a distributed optical fiber sensor using an optical fiber as a sensor, and generates first pulsed light having a pulse width based on the spatial resolution of the sensor. A predetermined phase difference between the first pulse light and the second pulse light, and a second pulse light generation unit that generates a second pulse light having a pulse width based on the reciprocal of the frequency resolution of the sensor. An inspection light generator that synthesizes the first pulse light and the second pulse light to generate inspection light, and emits the inspection light toward a specific end of the optical fiber, and the inspection light includes A spectrum detector for detecting a Brillouin spectrum based on light related to a Brillouin scattering phenomenon emitted from a specific end of an incident optical fiber, and a Briller due to strain or / and temperature generated in the optical fiber. Comprising a shift amount measuring unit for measuring the frequency shift amount. The inspection light generation unit generates a plurality of inspection lights having different phase differences from each other, and the spectrum detection unit applies the first pulse light and the second light in each inspection light to the light related to the Brillouin scattering phenomenon. A detector that respectively obtains Brillouin spectra corresponding to the plurality of inspection lights by applying a filter corresponding to a phase difference with the pulsed light, and a combining unit that synthesizes the Brillouin spectra obtained by the detector, The shift amount measurement unit measures the Brillouin frequency shift amount based on a combined spectrum that is a Brillouin spectrum combined by the combining unit.

  According to this distributed optical fiber sensor, a Brillouin spectrum is obtained by measurement using a plurality of different inspection lights, and the Brillouin spectrum having a high spatial resolution and a high frequency resolution is obtained by combining the plurality of Brillouin spectra. A spectrum (synthetic spectrum) can be obtained.

  Specifically, according to the distributed optical fiber sensor described above, the first Brillouin spectrum obtained by combining a plurality of inspection lights having different phase differences between the first pulsed light and the second pulsed light can be A combined spectrum having a spatial resolution corresponding to the pulse width of the pulsed light and a frequency resolution corresponding to the reciprocal of the pulse width of the second pulsed light is obtained. Therefore, a short pulse light having a small pulse width corresponding to the target spatial resolution is used as the first pulse light, and a long pulse light having a large pulse width corresponding to the reciprocal of the target frequency resolution is used as the second pulse light. Thus, a synthesized spectrum having a high spatial resolution and a high frequency resolution can be obtained, and as a result, the Brillouin frequency shift amount can be accurately measured.

  Specifically, the combining unit of the distributed optical fiber sensor described above extracts each Brillouin from the Brillouin spectrum obtained by the detector so that a component having a small spread in the time direction and a small spread in the frequency direction is extracted. By taking the weighted sum of the spectra, a combined spectrum having a spatial resolution corresponding to the pulse width of the first pulse light and a frequency resolution corresponding to the reciprocal of the pulse width of the second pulse light is generated.

  In the distributed optical fiber sensor according to the present invention, the detector may obtain a Brillouin spectrum by scanning a center frequency of the filter by signal processing. Thus, measurement time can be shortened by calculating | requiring a Brillouin spectrum by signal processing. Specifically, in the past, the Brillouin spectrum was obtained for each frequency while scanning the frequency of the reference light used when detecting the light related to the Brillouin scattering phenomenon. However, the reference light frequency was fixed and received in a wide band. The measurement time can be shortened by obtaining the Brillouin spectrum within the bandwidth by signal processing for scanning the center frequency of the filter.

  The spectrum detection unit includes a separation unit that separates the light related to the Brillouin scattering phenomenon into a horizontal component and a horizontal component, and the detection unit includes the light related to the Brillouin scattering phenomenon separated by the separation unit. It is preferable to obtain a Brillouin spectrum by applying a filter corresponding to the phase difference to a vertical component and a horizontal component of each and then taking the sum of squares thereof.

  According to this configuration, it is possible to suppress the influence of the polarization of light related to the Brillouin scattering phenomenon in the optical fiber and obtain the Brillouin spectrum with high accuracy. That is, using the fact that the absolute value of the polarization vector is stored in the optical fiber, the two components of the polarization vector, that is, the vertical component and the horizontal component of the light related to the Brillouin scattering phenomenon are individually measured and squared. By taking the sum, the influence of polarized light in the optical fiber can be suppressed.

  In order to solve the above problems, the present invention is a distributed optical fiber sensor that uses an optical fiber as a sensor, and has a pulse width longer than a time required for light to reciprocate a distance corresponding to the spatial resolution of the sensor. A first pulsed light generating unit that generates first pulsed light based on pulsed light; a second pulsed light generating unit that generates second pulsed light having a pulse width based on the reciprocal of the frequency resolution of the sensor; and the first pulsed light An inspection light is generated by combining the first pulse light and the second pulse light by providing a predetermined phase difference between the light and the second pulse light, and the inspection light is transmitted to a specific end of the optical fiber. An inspection light generation unit that emits light toward a part, a spectrum detection unit that detects a Brillouin spectrum based on light related to a Brillouin scattering phenomenon emitted from a specific end of the optical fiber on which the inspection light is incident, A shift amount measurement unit that measures a Brillouin frequency shift amount due to strain or / and temperature generated in the optical fiber, wherein the first pulsed light generation unit converts the long pulsed light into a plurality of widths based on the spatial resolution. The divided long pulse light is modulated using a predetermined code sequence to generate the first pulse light, and the inspection light generator generates a plurality of inspection lights having different phase differences from each other. The spectrum detector applies a filter corresponding to the phase difference between the first pulse light and the second pulse light in each inspection light to the light related to the Brillouin scattering phenomenon, and generates the first pulse light. A detection unit that obtains Brillouin spectra corresponding to the plurality of inspection lights by performing pulse compression after performing demodulation corresponding to the modulation in the unit, and obtains the detection by the detection unit A synthesis unit that synthesizes the Brillouin spectra, and the shift amount measurement unit measures the Brillouin frequency shift amount based on a synthesis spectrum that is a Brillouin spectrum synthesized by the synthesis unit. Features.

  Even with this distributed optical fiber sensor, a Brillouin spectrum is obtained by measurement using a plurality of different inspection lights, and by combining these Brillouin spectra, a combination with high spatial resolution and high frequency resolution is obtained. A spectrum can be obtained. In the distribution unit of the distributed optical fiber sensor, a Brillouin spectrum (synthetic spectrum) having a spatial resolution corresponding to the cell width of the first pulse light and a frequency resolution corresponding to the reciprocal of the pulse width of the second pulse light is obtained. It is done.

  Further, according to the distributed optical fiber sensor, the first pulse light is measured by using the long pulse light that has been modulated, and the first pulse light is demodulated at the time of reception to perform the pulse compression. Since the energy contained in the first pulse light is increased as compared with a method using short pulse light without performing pulse compression as pulse light, the SN ratio can be increased.

  In order to solve the above problems, the present invention is a distributed optical fiber sensor using an optical fiber as a sensor, wherein the first pulse light having a pulse width based on the spatial resolution of the sensor and the reciprocal of the spatial resolution of the sensor. A plurality of combined pulse light generators for generating a combined pulse light by combining the first pulse light and the second pulse light by providing a predetermined phase difference between the second pulse light based on The combined pulse light generated by the generation unit is merged into inspection light, and the inspection light is emitted toward a specific end of the optical fiber, and the specific end of the optical fiber on which the inspection light is incident A separation unit that separates light related to the Brillouin scattering phenomenon emitted from the unit into separated light corresponding to each combined pulse light before joining at the joining unit, and filtering the separated separated light The Brillouin frequency shift amount is calculated based on a plurality of detection units for obtaining Liluan spectra, a synthesis unit for synthesizing the Brillouin spectra obtained by each detection unit, and a synthesis spectrum which is a Brillouin spectrum synthesized by the synthesis unit. And a shift amount measuring unit for measuring. The predetermined phase difference is different for each of the combined pulse light generation units, and each detection unit has a phase difference between the first pulse light and the second pulse light constituting the combined pulse light corresponding to the separated light. A suitable filter is used.

  Even with this distributed optical fiber sensor, a Brillouin spectrum is obtained by measurement using a plurality of different inspection lights, and by combining these Brillouin spectra, a combination with high spatial resolution and high frequency resolution is obtained. A spectrum can be obtained. In addition, a plurality of combined pulse lights having different phase differences can be simultaneously generated by a plurality of combined pulse light generation units, and a Brillouin spectrum corresponding to each combined pulse light can be simultaneously obtained by a plurality of detection units. Therefore, the measurement time can be shortened.

  In addition, the phase difference between the first pulse light and the second pulse light of the composite pulse light generated for each composite pulse light generation unit can be fixed, and thereby a plurality of composites having different phase differences at the same site. The configuration of the means for forming the phase difference can be simplified compared to the case where all the pulsed light is generated.

  As described above, according to the present invention, it is possible to provide a distributed optical fiber sensor having high spatial resolution and high frequency resolution.

It is a functional block diagram which shows the structure of the distributed optical fiber sensor which concerns on 1st Embodiment. It is a figure for demonstrating the short pulse light generation part (or long pulse light generation part) of the said distributed optical fiber sensor. 3A shows the shape function of the short pulse light that constitutes the inspection pulse light, FIG. 3B shows the shape function of the long pulse light that constitutes the inspection pulse light, and FIG. FIG. 3D shows the impulse response of the element corresponding to the short pulse light constituting the low-pass filter, and FIG. 3D shows the impulse response of the element corresponding to the long pulse light constituting the low-pass filter. 4A shows a point spread function in the distributed optical fiber sensor, and FIG. 4B shows a combined spectrum obtained in the distributed optical fiber sensor. It is a figure for demonstrating the optical heterodyne detection part of the said distributed optical fiber sensor. It is a flowchart for demonstrating the measurement operation | movement of the distortion or temperature by the said distributed optical fiber sensor. It is a functional block diagram which shows the structure of the distributed optical fiber sensor which concerns on 2nd Embodiment. It is a flowchart for demonstrating the measurement operation | movement of the distortion or temperature by the said distributed optical fiber sensor. 9A shows the shape function of the first pulsed light, FIG. 9B shows the shape function of the second pulsed light, and FIG. 9C shows the low frequency band corresponding to the first pulsed light. The impulse response of the filter element is shown. FIG. 9D shows the impulse response of the low-pass filter element corresponding to the second pulse light. It is a functional block diagram which shows the structure of the distributed optical fiber sensor which concerns on other embodiment. It is a figure for demonstrating the optical heterodyne detection part and heterodyne detection part in 2 step | paragraph heterodyne reception. It is a figure for demonstrating the detection part in the distributed optical fiber sensor which concerns on other embodiment. It is a figure for demonstrating the distributed optical fiber sensor provided with the some test | inspection light emission part. It is a figure for demonstrating the distributed optical fiber sensor provided with the some test | inspection light emission part. FIG. 15A shows a point spread function in a conventional distributed optical fiber sensor, and FIG. 15B shows a Brillouin spectrum obtained in the conventional distributed optical fiber sensor.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

<First embodiment>
The distributed optical fiber sensor of the present embodiment uses an optical fiber (detection optical fiber) as a sensor, and measures the strain or / and temperature at each position in the longitudinal direction with high accuracy. As shown in FIG. 1, the distributed optical fiber sensor includes a detection optical fiber 12, an inspection pulse light emission unit 20, an optical circulator 14, a spectrum detection unit 30, a shift amount measurement unit 40, and a control. The processing unit 50 and the output unit 16 are provided.

  The detection optical fiber 12 is an optical fiber for a sensor that detects strain or temperature. The detection optical fiber 12 has a first end (specific end) 12a and a second end 12b that is the end opposite to the first end 12a. The first end portion 12a of the detection optical fiber 12 causes the inspection pulse light (inspection light) f to enter the detection optical fiber 12, and emits light related to the Brillouin scattering phenomenon caused by the inspection pulse light to the outside. . Here, when measuring strain or / and temperature generated in a measurement object such as a structure such as a pipe, an oilfield pipe, a bridge, a tunnel, a dam, a building, or the ground, the detection optical fiber 12 is an adhesive. It is fixed to the measurement object by a fixing member or the like.

  The inspection pulse light emitting unit 20 generates inspection pulse light f and emits the inspection pulse light f toward the first end 12 a of the detection optical fiber 12. Specifically, the inspection pulse light emitting unit 20 includes a first light source 21 and an optical pulse generation circuit 22.

  The first light source 21 can emit light having a predetermined frequency (continuous light), and can change the oscillation wavelength (oscillation frequency) by changing the temperature and drive current under the control of the control processing unit 50. In the present embodiment, a laser light source is used as the first light source 21, and for example, 1.55 μm coherent light is output from the first light source 21.

The optical pulse generation circuit 22 generates short pulse light (first pulse light) f ₁ and long pulse light (second pulse light) f ₂ from the continuous light emitted from the first light source 21 and synthesizes them. Is a circuit for generating inspection pulse light f. The optical pulse generation circuit 22 includes an optical coupler 23, a short pulse light generation unit 24, a long pulse light generation unit 25, a phase shifter 26, and an optical combiner 27.

  The optical coupler 23 branches the light from the first light source into two, and supplies the branched light to the short pulse light generation unit 24 and the long pulse light generation unit 25, respectively.

The short pulse light generation unit 24 generates the short pulse light f ₁ from the continuous light supplied from the first light source 21. Specifically, as shown in FIG. 2, the short pulse light generation unit 24 includes an intensity modulator 241 and a modulation signal input unit 242 that inputs a modulation signal to the intensity modulator 241. In the present embodiment, an LN intensity modulator is used as the intensity modulator 241, but the present invention is not limited to this. Further, in the present embodiment, the modulation signal Bias-T is used as the input unit 242, this Bias-T, a pulse signal as an electric signal corresponding to the short pulse light f ₁ to a DC voltage and the target is inputted The

The short pulse light generation unit 24 configured as described above generates short pulse light f ₁ having a pulse width based on the spatial resolution of the distributed optical fiber sensor 10 of the present embodiment. Specifically, for example, if the spatial resolution of the distributed optical fiber sensor 10 of the present embodiment is 10 cm, the light is in the detection optical fiber 12 (section of 10 cm in the longitudinal direction of the detection optical fiber 12). ), A short pulsed light f ₁ having a pulse width of 1 ns corresponding to the time of reciprocating is generated.

The long pulse light generator 25 generates the long pulse light f ₂ from the continuous light supplied from the first light source 21. The long pulse light generation unit 25 includes an intensity modulator 251 and a modulation signal input unit 252 as in the case of the short pulse light generation unit 24 (see FIG. 2). In this embodiment, an LN intensity modulator is used as the intensity modulator 251, but the present invention is not limited to this. In the present embodiment, Bias-T is used as the modulation signal input unit 252, and a pulse signal as an electric signal corresponding to the DC voltage and the target long pulse light f ₂ is input to Bias-T. The

The long pulse light generation unit 25 configured as described above generates long pulse light f ₂ having a pulse width based on the reciprocal of the frequency resolution of the distributed optical fiber sensor 10 of the present embodiment. Specifically, for example, if the frequency resolution (frequency width) of the distributed optical fiber sensor 10 of the present embodiment is 10 MHz, the long pulse light f ₂ having a pulse width of 100 ns, which is the inverse of the frequency resolution, is generated. In the present embodiment, the frequency width is the frequency range of the point spread function Ψ (t, ν) that dulls the Lorentz spectrum G (t, ν) when the Brillouin spectrum is expressed by the following equation (9). It means width.

The phase shifter 26 provides a predetermined phase difference between the short pulse light f ₁ and the long pulse light f ₂ . The phase shifter 26 is provided between the and the synthesizer 27 long pulse light generator 25, based on an instruction signal from the control processing unit 50, the long pulse light f ₂ the phase of the short pulse light f ₁ On the other hand, it is shifted by θ _j . That is, the phase shifter 26 phase-modulates the phase of the long pulse light f ₂ within a predetermined range with respect to the short pulse light f ₁ . In the phase shifter 26 of the present embodiment, θ _j , j = 1 to 4.

Here, how to obtain the plurality of check pulse light f to be incident on the detection optical fiber 12, i.e., the phase difference between the short pulse light f ₁ constituting the test pulse light f with long pulse light f _2, and the check pulse light The matched filter (low-pass filter in this embodiment) h corresponding to f will be described below.

As the elements constituting the inspection pulse light f, the following two pulse lights of a short pulse light f ₁ (t) and a long pulse light f ₂ (t) are considered.

Here, r is the amplitude ratio between the short pulse light f ₁ and the long pulse light f ₂ .

The pulse width D ₁ of the short pulse light f ₁ is set based on the target spatial resolution in the distributed optical fiber sensor 10. The pulse width D ₂ of the long pulse light f ₂ is set to be sufficiently longer than the lifetime of the phonon generated when the inspection pulse light f is injected into the detection optical fiber 12. A low-pass filter (a filter in the square detection unit 351 of the signal processing unit 35) h corresponding to the inspection pulse light composed of the short pulse light f ₁ and the long pulse light f ₂ is also composed of two elements h ₁ and h _2. It is assumed that the elements h ₁ and h ₂ are matched filters of the elements f ₁ and f ₂ of the inspection pulse light f. That is, the low-pass filters h ₁ and h ₂ are as follows.

Using the above elements, the inspection pulse light f and the low-pass filters h ₁ and h ₂ are expressed as follows.

Here, θ and φ are parameters representing the phase.

3A to 3D show examples of the shape functions of the elements f ₁ and f ₂ of the inspection pulse light f and the impulse response of the elements of the low-pass filter h. 3A shows the shape function of the first element (short pulse light) f ₁ constituting the inspection pulse light f, and FIG. 3B shows the second element (long pulse) constituting the inspection pulse light f. shows the shape function of the light) f _2, FIG. 3 (C) shows a first element (impulse response elements) h ₁ corresponding to the short pulse light constituting the low-pass filter h, FIG. 3 (D) is shows (the elements corresponding to the long pulse light) impulse response h ₂ second elements constituting a low-pass filter h.

  The point spread function corresponding to the above equations (14) and (15) is

It becomes. Here, R [...] in Equation (16) represents the real part.

  This point spread function expresses the Brillouin spectrum V (t, ν) as

Ψ (t, ν). Fτ {...} In this equation (16) represents a Fourier transform with respect to τ. G (t, ν) is a Lorentz spectrum, and a Brillouin frequency shift at a position z (distance z from the first end portion 12a) in the longitudinal direction of the detection optical fiber 12 is set to ν _B (z). When,

It is expressed.

  Moreover, in Formula (16),

It was.

This function F _kl (t, ν) is _derived from the difference in pulse width between f ₁ and f ₂ .
F ₁₁ : Small spread in time direction, large spread in frequency direction F ₁₂ : Large spread in time direction, large spread in frequency direction F ₂₁ : Large spread in time direction, in frequency direction F ₂₂ has the characteristic that the spread in the time direction is large and the spread in the frequency direction is small. Therefore, among the components of Ψ (t, ν), only F ₁₁ F ₂₂ ^* has an ideal characteristic that it is localized in both the time direction and the frequency direction, and other components are at least in the time direction and the frequency direction. It spreads in one direction. Therefore, for sensing, only F ₁₁ F ₂₂ ^* is a desirable component to be extracted, and the others are unnecessary components.

Therefore, in order to obtain a synthesis method for extracting only the element R (F ₁₁ F ₂₂ ^* ) of the point spread function Ψ (t, ν), the following phase difference (short pulse light f constituting the inspection pulse light f) is obtained. Consider a pair of a phase difference θ between ₁ and the long pulse light f ₂ and a matching (low-pass) filter h ₁ , h ₂ phase difference φ) corresponding thereto.

That is,

  The point spread function Ψ (t, ν) for these is

It becomes. By combining these as follows (that is, taking a weighted sum), only desired components can be extracted.

In the present embodiment, measurement is performed using four phase difference pairs (see Expression (20)), that is, four inspection pulse lights f having different phase differences between the short pulse light f ₁ and the long pulse light f _2. Thus, a desired component is extracted to obtain a composite spectrum V _S (t, ν), but is not limited thereto.

That is, by performing measurement with the following pair of phase differences (θ, φ) that generalizes the above, from a plurality of Brillouin spectra V (t, ν) corresponding to the inspection pulse light f of each phase difference θ, A combined spectrum V _S (t, ν) obtained by extracting components localized in both the time direction and the frequency direction can be obtained.

  First, p types of phase difference pairs (θ, φ) are prepared as follows.

  P types of pairs of inspection pulse light f and low-pass filter (matching filter) h corresponding thereto

And These equations (28) and (29) and the synthesis coefficients c _j , j = 1, 2,.

To be satisfied. Assuming that the point spread function for each pair of the inspection pulse light f and the low-pass filter h is Ψ _j (t, ν), the point spread function by synthesis is

It becomes. In this embodiment, p = 4,

It is.

4 (A) and 4 (B) show the point spread function Ψ (t, ν) and the synthesized spectrum (Brillouin spectrum after synthesis) V _S (t, ν) obtained by the present embodiment. An example is shown. 4A shows the point spread function Ψ (t, ν), and FIG. 4B shows the combined spectrum V _S (t, ν). The combined spectrum V _S (t, ν) is drawn with noise removed. The pulse width of the short pulse light f ₁ is 1 ns, and the pulse width of the long pulse light f ₂ is 40 ns. As can be seen from FIG. 4A, the point spread function Ψ (t, ν) has a small spread in both the time direction and the frequency direction. Further, as can be seen from FIG. 4B, the line width of the combined spectrum V _S (t, ν) becomes as small as the line width of the Lorentz spectrum G (t, ν).

  15A and 15B show the point spread function ψ (t, ν) in the conventional BOTDR (BOTDR using the inspection pulse light f consisting of a single pulse) and the measured Brillouin spectrum. An example with V (t, v) is shown. FIG. 15A shows the point spread function ψ (t, ν), and FIG. 15B shows the Brillouin spectrum V (t, ν). The Brillouin spectrum V (t, v) is drawn with noise removed. In this example, the pulse width of f is 1 ns. As shown in FIG. 15A, since the point spread function ψ (t, ν) has the uncertainty relationship shown in the equation (8), if the spread in the time direction is reduced, the spread in the frequency direction is increased. Become. Therefore, in the observation, as shown in FIG. 15B, the Lorentz spectrum G (t, ν) blurred by the convolution with the point spread function ψ (t, ν) is observed.

The optical combiner 27 combines the short pulse light f ₁ generated by the short pulse light generation unit 24 and the long pulse light f ₂ generated by the long pulse light generation unit 25 and phase-shifted by the phase shifter 26. The inspection pulse light f is generated, and the generated inspection pulse light (synthetic pulse light) f is output (emitted) toward the detection optical fiber 12.

In the inspection pulse light emitting section 20 configured as described above, a plurality of (four in this embodiment) inspection pulse lights f having different phase differences θ between the short pulse light f ₁ and the long pulse light f _2. Is generated.

  The optical circulator 14 is an irreversible optical component in which incident light and outgoing light have a cyclic relationship with their terminal numbers. That is, the light incident on the first terminal 14a is emitted from the second terminal 14b and not emitted from the third terminal 14c. The light incident on the second terminal 14b is emitted from the third terminal 14c. At the same time, the light that is not emitted from the first terminal 14a but is incident on the third terminal 14c is emitted from the first terminal 14a and is not emitted from the second terminal 14b. The first terminal 14a of the optical circulator 14 is connected to the inspection pulse light emitting unit 20, the second terminal 14b is connected to the first end 12a of the detection optical fiber 12, and the third terminal 14c is spectral detection. Connected to the unit 30.

  The spectrum detection unit 30 calculates the Brillouin spectrum V (t, v) based on the light (detection light) related to the Brillouin scattering phenomenon emitted from the first end 12a of the detection optical fiber 12 on which the inspection pulse light f is incident. The detection and detection unit 31 and the spectrum synthesis unit 36 are included. The spectrum detector 30 of the present embodiment has a Brillouin generated in the detection optical fiber 12 when the detection pulse light f is emitted from the inspection pulse light emitting unit 20 to the detection optical fiber 12 via the optical circulator 14. Light related to the scattering phenomenon is input from the detection optical fiber 12 via the optical circulator 14, and the Brillouin spectrum is detected based on this light.

The detection unit 31 responds to light related to the Brillouin scattering phenomenon (hereinafter also simply referred to as “detection light”) with the short pulse light f ₁ and the long pulse in each inspection pulse light f generated by the inspection pulse light emitting unit 20. A Brillouin spectrum V (t, ν) corresponding to each inspection pulse light f is obtained by applying a low-pass (matched) filter h corresponding to the phase difference θ with respect to the light f ₂ . The detection unit 31 includes a second light source 32, an optical heterodyne detection unit 33, an AD conversion unit 34, and a square detection unit 351 of the signal processing unit 35.

  The second light source 32 is a light source that emits reference light when the detection light is subjected to optical heterodyne detection. Like the first light source 21, the second light source 32 changes the temperature and the drive current by the control by the control processing unit 50. (Oscillation frequency) can be changed. In the present embodiment, a laser light source is used as the second light source 32.

  The optical heterodyne detection unit 33 converts the detection light into a baseband signal by detecting the detection light using the reference light from the second light source 32. As shown in FIG. 5, the optical heterodyne detection unit 33 according to the present embodiment combines an optical coupler 331 that combines the detection light and the reference light from the second light source 32 and then divides them into two, and outputs an optical coupler 331. Photodiodes PD1 and PD2 into which the branched lights from the coupler 331 are respectively input, and an amplifier 332 that amplifies each output from the photodiodes PD1 and PD2.

  The optical heterodyne detection unit 33 outputs the baseband signal as a two-channel signal of I and Q. The baseband signal of this embodiment is a signal in a band of about -1 GHz to 1 GHz.

  The AD converter 34 converts an analog signal into a digital signal. That is, the AD conversion unit 34 digitally samples the I and Q two-channel signals from the optical heterodyne detection unit 33 and outputs them to the signal processing unit 35 as digital signals.

The signal processing unit 35 includes a square detection unit 351 and a synthesis unit 352, and the square detection unit 351 constitutes a part of the detection unit 31 as described above. The square detection unit 351 applies a filter h corresponding to the phase difference θ between the first pulse light f ₁ and the second pulse light f ₂ to the digital signal input from the AD conversion unit 34 to perform square detection. Do. Thereby, the Brillouin spectrum V (t, v) corresponding to each inspection pulse light f generated by the inspection pulse light emitting unit 20 is detected.

On the other hand, the synthesizing unit 352 of the signal processing unit 35 constitutes the spectrum synthesizing unit 36 of the spectrum detecting unit 30, and synthesizes the Brillouin spectra V (t, ν) obtained by the detecting unit 31 with each other to produce a synthesized spectrum V _S. (T, v) (see FIG. 4B) is generated.

The shift amount measuring unit 40 measures the Brillouin frequency shift amount based on the combined spectrum V _S (t, ν) combined by the spectrum combining unit 36. Specifically, the Brillouin scattering phenomenon occurs at each position in the longitudinal direction of the detection optical fiber 12, and the scattered light returns to the first end 12a with a delay time proportional to the distance from the first end 12a. In the scattered light, a frequency shift proportional to the strain of the detection optical fiber 12 at the scattered position or / and the amount of change in temperature occurs (see FIG. 4B). Therefore, the shift amount measuring unit 40 measures the frequency shift amount from the combined spectrum V _S (t, ν) obtained by the combining unit 352, and detects the distortion of each position in the longitudinal direction of the detection optical fiber 12 or / And detect the temperature.

  The control processing unit 50 is a part for controlling each component of the distributed optical fiber sensor 10, and includes, for example, a microprocessor, a working memory, and a memory for storing necessary data.

The output unit 16 receives the measurement result of the shift amount measurement unit 40 and outputs (displays) the temperature or / and strain of the detection optical fiber 12 at each position in the longitudinal direction of the detection optical fiber 12. The form output unit 16 displays the temperature or / and strain at each position in the longitudinal direction of the detection optical fiber 12 on the screen, but may be configured to output (display) by printing or the like. The output unit 16 is configured to display not only the temperature and / or strain at each position of the detection optical fiber 12 but also the Brillouin frequency shift amount and the combined spectrum V _S (t, ν) at each position. Also good.

  Next, an operation for measuring strain or temperature in the distributed optical fiber sensor 10 as described above will be described. FIG. 6 is a flowchart for explaining a strain or temperature measurement operation by the distributed optical fiber sensor 10.

In FIG. 6, S1 is a loop performed by changing the phase difference θ between the short pulse light f ₁ constituting the test pulse light f with long pulse light f ₂ for each phase difference, S2 is the S1 This is a function for selecting the phase difference θ (that is, the amount of phase shift in the phase shifter 26). In this embodiment, each phase shift amount is

Given in.

  S3 is a loop in which the frequency ν of the reference light used in the optical heterodyne detection unit 33 is shifted, and the Brillouin spectrum V (t, ν) is measured for each shifted frequency. S4 is the frequency in S3. This is a shift amount designation function. For example, when measuring the entire 2 GHz band in increments of 5 MHz, k = 401. That is, the loop of S3 is performed 401 times.

S5 executes the contents of the functional block diagram of FIG. Specifically, the first light source 21 outputs coherent light having a wavelength of 1.55 μm, for example. This coherent light is branched into two by the optical coupler 23 in the optical pulse generation circuit 22, one is input to the short pulse light generation unit 24, and the other is input to the long pulse light generation unit 25. For example, when the target spatial resolution of the distributed optical fiber sensor 10 is 10 cm, the short pulse light generation unit 24 has a short pulse width of 1 ns corresponding to the round trip time of the section in the detection optical fiber 12. Pulse light (short pulse light) f ₁ is generated. On the other hand, for example, when the target frequency resolution of the distributed optical fiber sensor 10 is 10 MHz, the long pulse light generation unit 25 generates pulse light (long pulse light) f ₂ having a long pulse width of 100 ns, which is the inverse of the target frequency resolution. Generate. The phase shifter 26 shifts the phase of the long pulse light f ₂ by θ _{j with} respect to the short pulse light f ₁ . Then, the optical combiner 27 combines the short pulse light f ₁ and the phase-shifted long pulse light f ₂ , and the combined composite pulse light inspection pulse light f is passed through the optical circulator 14 to the first end portion. Injection into the detection optical fiber 12 from 12a.

In the detection optical fiber 12, the Brillouin scattering phenomenon occurs at all positions in the longitudinal direction, and the light related to the Brillouin scattering phenomenon has the first end with a delay time proportional to the distance from the first end 12a. Injected from the part 12a. The light related to the Brillouin scattering phenomenon includes a Brillouin frequency proportional to the strain of the detection optical fiber 12 and / or the amount of change in temperature at the position z where the Brillouin scattering occurs (that is, the distance z from the first end 12a). A shift ν _b (z) occurs.

  The light related to the Brillouin scattering phenomenon is input to the spectrum detection unit 30 through the optical circulator 14 as detection light. In the spectrum detection unit 30, the optical heterodyne detection unit 33 performs optical heterodyne detection using the coherent light output from the second light source 32 as reference light. By this optical heterodyne detection, detection light is converted into a baseband signal (electric signal). This baseband signal is a baseband signal in a band of about -1 GHz to 1 GHz. The AD converter 34 samples and digitally converts the I and Q two-channel signals output from the optical heterodyne detector 33 and outputs the digital signals to the square detector 351 of the signal processor 35 as digital signals.

  The square detection unit 351 performs square detection by applying a low-frequency (matched) filter h to the signal input from the AD conversion unit 34. At this time, the impulse response of the low-pass filter h is

Each element on the right side of Expression (34) is an element of inspection pulse light f (short pulse light f ₁ and long pulse light f ₂ ),

That relationship. Since the low-pass filter h is used in digital processing in the signal processing unit 35, it is discretized.

And A digital signal obtained by sampling the output signal of optical heterodyne detection

The processing of the low-pass filter of the square detection unit 351 is as follows:

It is expressed. At this time, the output from the square detector 351 is:

It is expressed.

In S6, the Brillouin spectrum V (t, v) for each j is synthesized to obtain a synthesized spectrum V _S (t, v). Specifically, the synthesizer 352 synthesizes the Brillouin spectrum V (t, ν) corresponding to each j input from the square wave detector 351 by the following equation (40) to obtain a synthesized spectrum V _S (t, ν) is obtained, and this synthesized spectrum V _S (t, ν) is output to the shift amount measuring unit 40.

  Where the coefficient is

In S7, the Brillouin frequency shift amount is measured based on the combined spectrum V _S (t, ν). Specifically, the shift amount measurement unit 40 has a parabola or an appropriate value for V _S (t _n , ν _n ) every time t _n is fixed with respect to the composite spectrum V _S (t, ν) input from the synthesis unit 352. A precise curve position is obtained, and a precise peak position is obtained, which is used as an estimate of the Brillouin frequency shift (see FIG. 4B). The shift amount measuring unit 40 measures the Brillouin frequency shift amount from the obtained estimated value, and outputs this to the output unit 16.

  In S8, the output unit 16 outputs (displays, etc.) the measurement result in the shift amount measuring unit 40.

As described above, according to the distributed optical fiber sensor 10 of the first embodiment, the Brillouin spectrum V (t, v) is obtained by measurement using a plurality of different inspection pulse lights f, and the plurality of Brillouin spectra are obtained. By synthesizing the spectrum V (t, v), a Brillouin spectrum (synthesized spectrum) V _S (t, v) having a high spatial resolution and a high frequency resolution can be obtained.

Specifically, according to the distributed optical fiber sensor 10 of the first embodiment, each obtained by a plurality of inspection pulse lights f having different phase differences θ _j between the short pulse light f ₁ and the long pulse light f _2. By synthesizing the Brillouin spectrum V (t, v), a synthesized spectrum V _S (having spatial resolution corresponding to the pulse width of the short pulse light f ₁ and frequency resolution corresponding to the reciprocal of the pulse width of the long pulse light f ₂ is obtained. t, v) is obtained. Therefore, pulse light having a small pulse width corresponding to the target spatial resolution is used as the short pulse light f ₁ , and pulse light having a sufficiently large pulse width corresponding to the reciprocal of the target frequency resolution is used as the long pulse light f _2. By using it, it is possible to obtain a synthesized spectrum V _S (t, ν) having a high spatial resolution and a high frequency resolution, and as a result, the Brillouin frequency shift amount can be accurately measured.

<Second Embodiment>
Next, the configuration of the distributed optical fiber sensor 10a according to the second embodiment of the present invention will be described with reference to FIGS. 7 and 8. The same reference numerals are used for the same configurations as in the first embodiment. Detailed description will be omitted, and only different configurations will be described in detail.

  As shown in FIG. 7, the distributed optical fiber sensor 10a includes a detection optical fiber 12, an inspection pulse light emitting unit 20, an optical circulator 14, a spectrum detecting unit 30a, a shift amount measuring unit 40, A control processing unit 50 and an output unit 16 are provided.

  The spectrum detection unit 30a includes a detection unit 31a and a spectrum synthesis unit 36 (synthesis unit 352). The detection unit 31a includes an optical heterodyne detection unit 33, a local oscillator 37, a heterodyne detection unit 38, and an AD conversion unit. 34a and a signal processing unit 35a.

  Unlike the first embodiment, the optical heterodyne detection unit 33 uses, as reference light, coherent light output from the first light source 21 and branched by the optical coupler 28 before being input to the optical pulse generation circuit 22. Perform detection. Thereby, the optical heterodyne detection unit 33 converts the input detection light into a microwave signal having a frequency of about 9 to 11 GHz and outputs the microwave signal.

  The local oscillator 37 outputs a reference signal used for heterodyne detection in the heterodyne detection unit 38. The local oscillator 37 is controlled by the control processing unit 50 and outputs a sine wave signal having a fixed oscillation frequency as a reference signal.

  The heterodyne detection unit 38 converts the microwave signal into a baseband signal by heterodyne detection of the microwave signal from the optical heterodyne detection unit 33 using the reference signal from the local oscillator 37. The heterodyne detection unit 38 of the present embodiment generates a baseband signal by causing a reference signal to interfere with a microwave signal using a mixer circuit.

  The AD conversion unit 34 a digitally samples the baseband signal from the heterodyne detection unit 38 and outputs the digitally sampled signal to the signal processing unit 35.

The signal processing unit 35 includes a square detection unit 351a and a synthesis unit 352. The square detection unit 351a performs square detection by applying a bandpass filter having ν _k as a center frequency to the digital signal digitally sampled from the baseband signal by the AD conversion unit 34a. Here, K frequencies ν _k , k = 1, 2,. For example, in the square wave detection unit of the present embodiment, K = 401 is obtained when measuring the entire 2 GHz band in increments of 5 MHz.

  As described above, the square detection unit 351a (signal processing unit 35) of the present embodiment receives the output signal from the heterodyne detection unit 38 in which the frequency of the reference signal is fixed in a wide band, and the Brillouin spectrum V ( t, v) are obtained. That is, instead of scanning the frequency of the reference light (reference signal) for heterodyne detection, the frequency of the reference light is fixed and received in a wide band and then digitally sampled at a high speed, and signal processing is performed in the square detection unit 351a. The Brillouin spectrum V (t, ν) is obtained. Thereby, compared with the case where the frequency of the reference light is scanned at the time of heterodyne detection (the first embodiment or the like), the measurement time can be shortened to 1 / K.

  Specifically, the spectrum detection unit 30a of the present embodiment is different from the spectrum detection unit 30 (see FIG. 1) of the first embodiment as follows.

The spectrum detection unit 30 according to the first embodiment scans the frequency of the reference light (reference signal) and detects the intensity of the spectrum at each frequency when detecting the detection light in order to obtain the Brillouin spectrum V (t, v). Is measured. Specifically, the spectrum detection unit 30 according to the first embodiment scans the frequency of the reference light and calculates the Brillouin spectrum V (t, ν) for the discrete frequencies ν _k , k = 1, 2,. measure. For example, if the measurement band is -1 GHz to 1 GHz and the interval between discrete frequencies is 5 MHz, K = 401 frequencies are measured, that is, optical heterodyne detection and square detection are performed for K = 401 frequencies. Each done.

  On the other hand, in the spectrum detection unit 30a of this embodiment, the frequency of the reference signal (reference light) at the time of detection of the detection light is fixed, and the AD conversion unit 34a digitally samples the baseband signal at high speed. The square detection unit 351a obtains each Brillouin spectrum V (t, v) in the band by signal processing.

  Specifically, the sampling period of the baseband signal is Δt, and the sampling time is

far. Here, t ₀ represents the round trip time of the inspection pulse light f from the first end portion 12a of the detection optical fiber 12 to the start position of the section to be observed in the optical fiber 12, and N represents the digital signal Represents length. The total time T = NΔt is the sum of the time T _i in which the inspection pulse light f reciprocates in the section to be observed in the detection optical fiber 12 and the time width (pulse width) T _p of the inspection pulse light f. Also, the sample value of the baseband signal that has undergone two-stage heterodyne (heterodyne in the optical heterodyne detection unit 33 and heterodyne in the heterodyne detection unit 38) is obtained.

far.

If the support for the low pass filter h (t) is at [−T _p , 0], the output after passing through the low pass filter h (t) is

And in discrete time,

Can be approximated by Furthermore, since this output is absolute, the first phase term is not necessary,

It becomes. Then, the square detection unit 351a square-detects this output as shown in the following equation (47) to obtain the Brillouin spectrum V _n (ν) and outputs it.

The output Y _n (ν) at discrete time shown in Expression (46) can be applied with a recursive algorithm when the impulse response of the low-pass filter h is a rectangle or a combination of rectangles. Specifically, first, check pulse light f is a rectangular pulse of duration T _p, a low-pass filter h corresponding thereto

Then, the output after passing through this low-pass filter h is

And this can be rewritten as a recursive algorithm

It becomes. Here, N _f = T _f / Δt is the number of signals in the detection optical fiber 12.

As in the distributed optical fiber sensor 10a, obtains a plurality of Brillouin spectrum V corresponding to the phase difference θ between the short pulse light f ₁ constituting the test pulse light f with long pulse light f ₂ (t, ν), When sensing by combining these, the signal with the parameter θ is

And low-pass filter h

Then, the output after passing through the low-pass filter h is

It becomes. When this equation (56) is rewritten into a recursive algorithm,

It becomes.

  Next, the strain or temperature measurement operation in the distributed optical fiber sensor 10a as described above will be described. FIG. 8 is a flowchart for explaining a strain or temperature measurement operation by the distributed optical fiber sensor 10a. In this embodiment, unlike the first embodiment, the frequency of the reference signal (reference light) used in the heterodyne is not shifted (the frequency of the reference signal is fixed). Therefore, the flowchart of FIG. The loop corresponding to S3 in (see FIG. 6) is not performed.

  This will be specifically described below.

In FIG. 8, S1 is a loop performed by changing the phase difference θ between the short pulse light f ₁ constituting the test pulse light f with long pulse light f ₂ for each phase difference, S2 is the S1 This is a function for selecting the phase difference θ (that is, the amount of phase shift in the phase shifter 26).

S5a executes the contents of the functional block diagram of FIG. Specifically, the coherent light output from the first light source 21 is branched by the optical coupler 28, one is input to the optical pulse generation circuit 22, and the other is used as reference light in the optical heterodyne detection unit 33. The short pulse light generation unit 24 and the long pulse light generation unit 25 generate the short pulse light f ₁ and the long pulse light f ₂ having a predetermined pulse width from the coherent light input to the optical pulse generation circuit 22. The delaying theta _j the phase of the phase shifter 26 is long pulse light f ₂ with respect to the short pulse light f ₁ (by phase-shift), the long pulse light combiner 27 is the phase shift between pulse light f ₁ The light f ₂ is combined to generate inspection pulse light f that is composite pulse light. The inspection pulse light f is injected into the detection optical fiber 12.

  The optical heterodyne detection unit 33 uses the coherent light from the first light source 21 as reference light and heterodyne-detects the detection light emitted from the first end 12a of the detection optical fiber 12. The optical heterodyne detection unit 33 converts the detection light into a microwave signal having a frequency of about 9 to 11 GHz. Then, the heterodyne detection unit 38 converts the microwave signal output from the optical heterodyne detection unit 33 into a baseband signal using the reference signal from the local oscillator 37. This baseband signal is a baseband signal in a band of about -1 GHz to 1 GHz. The AD conversion unit 34a digitally samples the baseband signal and outputs it to the signal processing unit 35a.

The square detection unit 351a of the signal processing unit 35a performs square detection by passing the baseband signal from the AD conversion unit 34a through a band filter having ν _k as a center frequency.

  Specifically, the impulse response of the low-pass filter h in the square detection unit 351a is

It is expressed in the form of Each element on the right side of the equation (59) is an element of the inspection pulse light f.

The relationship is as follows. Incidentally, signals, passing a band filter having a center frequency [nu _k is equivalent to applying a low-pass filter h from shifting the frequency of the signal downward by [nu _k.

  The low-pass filter h is discretized because it is used in digital processing in the square detection unit 351a.

And A digital signal obtained by sampling the output signal of the heterodyne detection unit 38 by the AD conversion unit 34a

The processing of the band filter in the square detection unit 351a is

It is represented by The square detection unit 351a square-detects the filtered signal as shown in the following equation (64) and outputs the result.

In S 6, the Brillouin spectrum V (t, ν) for each j is synthesized to obtain a synthesized spectrum V _S (t, ν), and this synthesized spectrum V _S (t, ν) is output to the shift amount measuring unit 40. . In S 7, the Brillouin frequency shift amount is measured based on the combined spectrum V _S (t, ν), and the measurement result is output to the output unit 16. In S8, the output unit 16 outputs (displays, etc.) the measurement result in the shift amount measuring unit 40.

  As described above, according to the distributed optical fiber sensor 10a of the second embodiment, the frequency of the reference light (reference signal) for heterodyne detection is fixed and received in a wide band, and then digital sampling is performed at high speed, and signal processing is performed. By obtaining the Brillouin spectrum V (t, ν) in the band, the measurement time can be shortened. Specifically, in the past, the Brillouin spectrum V (t, v) was obtained for each frequency while scanning the frequency of the reference light used when detecting light related to the Brillouin scattering phenomenon, but the frequency of the reference light is fixed. The center frequency of the filter is scanned after reception in a wide band, that is, digital sampling is performed at a high speed by the AD converter 34a, and then the Brillouin spectrum V (t, By obtaining ν), the measurement time can be shortened.

  The distributed optical fiber sensor of the present invention is not limited to the first embodiment and the second embodiment, and various changes can be made without departing from the scope of the present invention. .

In the first embodiment and the second embodiment, the composite pulse obtained by combining the short pulse light f ₁ having the pulse width based on the spatial resolution and the long pulse light f ₂ having the pulse width based on the reciprocal of the frequency resolution as the inspection pulse light f. Although light f is used, it is not limited to this. For example, the inspection pulse light f is generated by combining the pulse light f _1A having a large pulse width modulated using the code sequence and the long pulse light f ₂ having a pulse width based on the reciprocal of the frequency resolution. May be. By using such inspection pulse light f, a high S / N ratio can be achieved.

  Specifically, it is as follows.

Instead of the short pulse light f ₁ in the first embodiment and the second embodiment, a pulse light having a large pulse width is modulated using a code sequence, and demodulated or converted at the time of reception, An effect equivalent to that of a pulse light having a substantially small pulse width and having a strong signal intensity is obtained. As a method of performing modulation using this code, there are a pulse compression method using correlation (pulse compression method), a method based on Hadamard matrix without using correlation (matrix inversion method), and the like. In either method, a high S / N ratio can be obtained.

Specifically, in the pulse compression method, the inspection pulse light f is generated by the first pulse light f _1A and the second pulse light f ₂ . Second pulse light f _2, as in the first embodiment and the second embodiment, a long pulse light having a pulse width based on the inverse of the frequency decomposition.

The first pulsed light f _1A is a sufficiently large pulsed light with a pulse width D ₁ , and the inside of the pulse is divided into M cells. The width d of each cell is equal, d = D ₁ / M, and this d is based on the spatial resolution of the distributed optical fiber sensor, as is the pulse width of the short pulse light f ₁ of the first and second embodiments. Width. Each cell

It is represented by When the pseudo-random number sequence is a sequence rm, m = 1, 2,..., M that takes a value of ± 1, the shape function of the first pulsed light f _1A is

It is represented by

  At this time, the pseudo-random number has its autocorrelation function

It is ideal to satisfy. Here, δk, 0 represents the Kronecker delta. No single sequence having this property has been found, but it can be realized by the sum of autocorrelation functions of two sequences as in the Golay code sequence.

  Other configurations are the same as those in the first embodiment and the second embodiment.

FIG. 9 corresponds to the shape function of the elements of the inspection pulse light f (first pulse light f _1A and second pulse light f ₂ ) when performing pulse compression, and the elements of the low-pass filter h (first pulse light). An example of an impulse response of an element and an element corresponding to the second pulse light is shown. In FIG. 9, (A) shows the shape function of the first pulsed light f _1A , (B) shows the shape function of the second pulsed light, and (C) shows the low frequency band corresponding to the first pulsed light. The impulse response of the filter element is shown, and (D) shows the impulse response of the low-pass filter element corresponding to the second pulse light. In FIGS. 9A and 9C, + and − represent +1 and −1, respectively.

Also in the case of performing pulse compression, a synthetic method (like the first and second embodiments, each Brillouin spectrum V (t, v) corresponding to a plurality of inspection pulse lights f is synthesized to produce a synthesized spectrum V _S ( The point spread function obtained by the method for obtaining t, v) is

It is expressed. Of the formula (68), F ₂₂ (t, ν) is less spread in the frequency direction by increasing the pulse width of the second pulsed light f ₂ . For this reason, F ₂₂ (t, ν) can be approximated by δ (ν) by removing constants in the frequency direction. Here, δ (ν) is a Dirac δ function. Therefore, the characteristic of F ₁₁ in the vicinity of ν = 0 is important, but according to the expression (67) of the property of pseudorandom numbers,

It becomes. If both sides are integrated by t,

Therefore, if D1 is fixed and d → 0,

Note that

From the above, Ψ _s (t, ν) behaves like a δ function in a two-dimensional space of time and frequency. At this time, since the coefficient of the δ function is proportional to M, the intensity of the observed Brillouin spectrum is M times that in the case where pulse compression is not performed (when M = 1). Is regarded as the amplitude of this signal, the intensity of the signal is to be increased to M ^2. On the other hand, since the signal time is M times, the noise intensity (dispersion) is M times. As a result, the S / N ratio increases M ² / M = M times.

  Here, the specific configuration of the pseudo random number in the pulse compression method will be described in more detail. First, a method using the above Golay code will be described.

When the sum of the autocorrelation of each pair of length M code strings A _m , B _m , k = 0, 1,..., M−1 becomes 0 for all non-zero shifts,

Is called a complementary series. The Golay code sequence is a binary complementary sequence having a value of ± 1.

  In order to perform modulation using the Golay code, two types of sequences (A sequence and B sequence) of the following formula (73) and formula (74) are prepared for the first pulse light.

  In addition, the impulse response of the corresponding low-pass filter

And

  A synthetic method is applied to each of the A and B sequences of the Golay code, and then the spectrum is summed. The point spread function at this time is

And the value of F ₁₁ ^A (t, ν) + F ₁₁ ^B (t, ν) at ν = 0 depends on the nature of the Golay code.

It becomes. Therefore, the same result as that obtained when an ideal pseudorandom number is used can be obtained.

  Next, a method using a unipolar Golay code will be described.

  A code that takes two values ± 1 is called a bipolar type, and a code that takes two values 0 and 1 is called a unipolar type. A bipolar type code is realized by two-phase phase modulation of 0 and π, while a unipolar type code can be realized in the form of amplitude modulation or light intensity modulation. The Golay code sequence is a bipolar type, but can be transformed into a monopolar type.

  The following four types of code sequences are prepared with {Am} and {Bm} as Golay code sequences.

At this time, A _m ⁺ −A _m ⁻ = A _m , B _m ⁺ −B _m ⁻ = B _m ,

Is established. Therefore, measurement is performed using four sequences of A _m ⁺ , A _m ⁻ , B _m ⁺ , and B _m ⁻ , and correlation is made with A _m , −A _m , B _m , and −B _m , respectively. When combined, the same δ function as the Golay code is obtained.

  Next, in the same manner as in the pulse compression method, the specific configuration of the matrix in the matrix inversion method that provides a high S / N ratio will be described in detail. First, a method using a Hadamard matrix will be described.

  The Hadamard matrix is a square matrix whose elements have binary values of ± 1 and whose rows are orthogonal to each other. That is, the Mth order Hadamard matrix is

Meet. Here, I _M is an M-th unit matrix. As a result, the inverse matrix becomes

Given in. For example,

Becomes a Hadamard matrix.

  Using this Hadamard matrix, the following M types of sequences are prepared for the first pulsed light.

That is, the mth first pulsed light corresponds to the mth row of the Hadamard matrix. The low-pass filter corresponding to the first pulse light is

And This is common for each m.

For each m, _V S the Brillouin spectrum obtained when performing the measurement of the BOTDR by synthetic _{methods, m (t, ν),} m = 1,2, ..., M Distant, follows by multiplying the ^{H T} M Brillouin spectra as shown in Equation (89) are obtained.

  Add these together while shifting the time.

And Such Brillouin spectrum thus obtained is likewise SN ratio in the case of performing pulse compression increases to ^twice M. Note that M ² is equal to the number of elements of the matrix, and has the same effect as an S / N ratio increased M times by pulse compression using M cells.

In order to confirm the effect of increasing the S / N ratio in the method using the above matrix, the behavior of F ₁₁ (t, 0) of the element of the point spread function may be examined as in the case of pulse compression. First, for each first pulse light,

It becomes.

To this, when the action of ^{H T,}

It becomes. Therefore, what is superimposed at different times is

Thus, the shape is the same as in the case of pulse compression.

  The method using the Hadamard matrix can be extended to other similar matrices, and the same effect can be obtained. For example, there is a method of using a matrix obtained by removing the first row and the first column of the Hadamard matrix, or an S matrix obtained by converting it into a unipolar type. In particular, a method using an S matrix converted to a unipolar type is called a simplex method and is well known as a method for increasing the SN ratio.

  The distributed optical fiber sensor 10 of the first embodiment measures the Brillouin spectrum by one-stage heterodyne reception that performs optical heterodyne detection on the detection light from the detection optical fiber 12 and AD-converts the output to square detection. However, the present invention is not limited to this, and similarly to the distributed optical fiber sensor 10a of the second embodiment, the detection light is subjected to optical heterodyne detection, and then heterodyne detection is performed. The Brillouin spectrum may be measured by two-stage heterodyne reception for detection.

  For example, specifically, as shown in FIG. 10, the detection unit 31b of the spectrum detection unit 30b includes an optical heterodyne detection unit 33, a local oscillator 37, a heterodyne detection unit (mixer circuit in FIG. 10) 38, AD A conversion unit 34 and a signal processing unit 35 are included. Further, unlike the second embodiment, the spectrum detector 30b is a frequency shift unit between the local oscillator 37 and the heterodyne detector 38 in order to detect the Brillouin spectrum at each frequency by scanning the frequency of the reference signal. 39. The frequency shift unit 39 is a circuit for modulating the frequency of the reference signal (standard signal) output from the local oscillator 37. A phase locked loop (PLL) is used for the frequency shift unit 39 of this embodiment (see FIG. 11). The frequency shift unit (phase synchronization circuit) 39 includes a phase comparator 391, a low pass filter (LPF) 392, an amplifier 393, and a voltage controlled oscillator 394. The phase comparator 391 is a circuit that converts the phase difference between two input signals (in this embodiment, a signal from the local oscillator 37 and an output signal from the voltage controlled oscillator 394) into a voltage and outputs the voltage. The low-pass filter 392 is a feedback loop filter, and in a circuit including feedback, a short-cycle signal fluctuation is amplified, and unnecessary oscillation may occur, so an unnecessary short-cycle fluctuation is cut off to avoid this. To do. The voltage controlled oscillator 394 is a circuit that controls the output frequency according to the input voltage. The voltage controlled oscillator 394 according to the present embodiment receives the signal that has passed through the low-pass filter 392 and the frequency shift voltage obtained by amplifying the frequency shift voltage based on the instruction from the control processing unit 50 according to the frequency shift voltage. A signal having a different frequency is oscillated as a reference signal in the heterodyne detection unit 38.

In this distributed optical fiber sensor 10b, the optical heterodyne detection unit 33 performs heterodyne detection on the detection light from the detection optical fiber 12 using the coherent light branched from the first light source 21 as a reference light, and has a frequency of 9 to It converts into a microwave signal of about 11 GHz. Then, the heterodyne detection unit 38 performs heterodyne detection on the microwave signal from the optical heterodyne detection unit 33 and converts it to a baseband signal having a frequency of about −1 GHz to 1 GHz. What is used as a reference signal for heterodyne detection in the heterodyne detection unit 38 is a sine wave signal generated by the local oscillator 37 and shifted in frequency by ν _k by the frequency shift unit 39. The I and Q two-channel signals output from the heterodyne detection unit 38 are sampled by the AD conversion unit 34 and input to the signal processing unit 35 (specifically, the square detection unit 351) as a digital signal. The output from the heterodyne detection unit 38 may be a 1-channel signal in which the band of the baseband signal is 0 GHz to 2 GHz, instead of the 2 channels of I and Q. In this case, the sampling rate in the AD converter 34 is doubled.

The square detection unit 351 of the signal processing unit 35 performs square detection by applying a low-pass filter h to the signal from the AD conversion unit 34. At this time, the impulse response of the low-pass filter h is in the form of Expression (34) as in the first embodiment, and each element on the right side of Expression (34) is referred to as the element of the inspection pulse light f and Expression (35). In relation, the processing of the low-pass filter h is expressed by Expression (38). The output from the square detection unit 351 at this time is expressed by Expression (39). The synthesizer 352 synthesizes the Brillouin spectrum V (t, ν) corresponding to each j input from the square wave detector 351 according to the equation (40) to obtain a synthesized spectrum V _S (t, ν). The combined spectrum V _S (t, ν) is output to the shift amount measuring unit 40.

  The distributed optical fiber sensor detects the detection light by dividing it into two components in order to suppress the bias of the Brillouin spectrum value due to the fluctuation of the polarization plane generated in the optical fiber, and by taking the square sum thereof, You may comprise so that the influence of polarization may be suppressed.

  Specifically, as the light travels through the optical fiber, the polarization state of this light changes (that is, the polarization plane fluctuates), and as the optical fiber becomes longer, its control and prediction becomes difficult. It was randomized using a polarization scrambler and multiple pulses. However, with this method, since it is necessary to use a very large number of pulses, if the measurement time is long or randomization is not sufficient, the measured Brillouin spectrum value is biased and sufficient accuracy cannot be secured. There was a case.

  In BOTDR, the absolute value of the polarization vector is preserved. Therefore, if the two components (S wave and P wave) of the vector are individually measured and the sum of the squares is taken, the influence of the change in the polarization state is eliminated. And can be measured with high accuracy.

For example, specifically, as shown in FIG. 12, the detector 31c includes a polarization separator (PSB) 311 that separates the detected light into S and P waves, and an optical coupler that separates the reference light into two. 312 and a first optical heterodyne detection unit 33a, a first heterodyne detection unit 38a, and a first AD conversion unit 34a for detecting a Brillouin spectrum from the separated S wave. A second optical heterodyne detector 33b, a second heterodyne detector 38b, and a second AD converter 34b for detecting the Brillouin spectrum from the P wave are provided. In this detection unit 31c, the S wave and the P wave are subjected to optical heterodyne detection and heterodyne detection, respectively, and are digitally sampled into two I and Q channels by the AD conversion units 34a and 34b, respectively, and input to the signal processing unit 35. Is done. The square wave detector 351c is a bandpass filter having ν _k as a center frequency for K frequencies ν _k , k = 1, 2,..., K for each component of polarized light (S wave and P wave). And square detection. The square detection unit 351c calculates the sum of the square detection output for each polarization component, and outputs the sum as a Brillouin spectrum V (t, v).

A plurality of inspection pulse lights f having different phase differences θ between the short pulse light f ₁ (or the pulse light f _1A, which is internally divided into M cells) and the long pulse light f ₂ constituting the inspection pulse light. The method of generating is not limited. In the first embodiment and the second embodiment, the plurality of inspection pulse lights f are generated in the common optical pulse generation circuit 22. However, the present invention is not limited to this. For example, as shown in FIG. a plurality of optical pulse generating circuit 22 d, 22e, 22f, ... are provided, the light pulse generation circuit 22 d, 22e, 22f, test pulses ... each to the phase difference between the short pulse light _{f 1} and long pulse light _{f 2} theta is different The light f may be generated. In this case, as shown in FIG. 14, detectors 31d, 31e, 31f,... Corresponding to the respective optical pulse generators 22d, 22e, 22f,... Are provided, and the detectors 31d, 31e, 31f,. The Brillouin spectrum V (t, ν) detected by is synthesized in the common synthesis unit 352.

Specifically, wavelength division multiplexing (WDM) is used. Specifically, a plurality of (four in FIG. 13) inspection pulse light emitting sections 20d, 20e, 20f, 20g,..., A first optical multiplexer / demultiplexer (merging section) 60, an optical circulator 14, and a detection Corresponding to the optical fiber 12, the second optical multiplexer / demultiplexer (separator) 61, and the inspection pulse light emitting portions 20d, 20e, 20f, 20g,... (That is, the short pulse light f in each inspection pulse light f) comprising ₁ to correspond to the phase difference θ between the long pulse light _{f 2)} a plurality of detection portions 31d, 31e, 31f, 31g, ... and, a combining unit 352, a shift amount measuring unit 40, an output unit 16, the .

  Each of the inspection pulse light emitting units 20d, 20e, 20f, 20g,... Includes a first light source 21 and optical pulse generation circuits 22d, 22e, 22g, 22f,. Each of the optical pulse circuits 22d, 22e, 22g, 22f,... Includes a short pulse light generation unit 24, a long pulse light generation unit 25, phase shift modulation members 26d, 26e, 26f, 26g,. Each is provided.

Each phase modulation member 26d, 26e, 26f, 26g, ... is the different from the first embodiment and the phase shifter of the second embodiment, the phase shift amount of the long pulse light _{f 2} with respect to the short-pulse light _{f 1} is fixed The phase shift amount is different for each of the phase shift modulation members 26d, 26e, 26f, 26g,. For example, each phase shift modulation member 26 d, 26e, 26f, 26 g, as ..., thickness glass of according to the phase difference θ between the short pulse light _{f 1} and long pulse light _{f 2} as a target is used, respectively.

  The first optical multiplexer / demultiplexer 61 is composed of, for example, an optical signal multiplexing circuit or the like, and superimposes (synthesizes) the inspection pulse lights f emitted from the inspection pulse light emission units 20d, 20e, 20f, 20g,. ) The second optical multiplexer / demultiplexer 61 is output to the detection optical fiber 12 and is composed of, for example, an optical signal separation circuit or the like, and is emitted from each of the inspection pulse light emitting sections 20d, 20e, 20f, 20g,. The detection light from the detection optical fiber 12 is separated so as to correspond to the pulsed light f.

  Each of the detection units 31d, 31e, 31f, 31g,... Includes a second light source 32, an optical heterodyne detection unit 33, an AD conversion unit 34, and a square detection unit 351, respectively. The Brillouin spectrum V (t, v) is obtained from the detected light based on the detected light.

The synthesizer 352 synthesizes the Brillouin spectrum V (t, ν) obtained by the detectors 31d, 31e, 31f, 31g,... To generate a synthesized spectrum V _S (t, ν). In FIG. 14, a phase synthesis filter is used as the synthesis unit 352.

Also with this distributed optical fiber sensor, a plurality of Brillouin spectra V (t, ν) are synthesized, so that a synthesized spectrum V _S (t, ν) having a high spatial resolution and a high frequency resolution can be obtained. . In addition, a plurality of inspection pulse lights f having different phase differences θ can be simultaneously generated in the plurality of inspection pulse light emitting sections 20d, 20e, 20f, 20g,... And the detection light from the detection optical fiber 12 , The Brillouin spectrum V (t, ν) corresponding to each inspection pulse light f can be obtained simultaneously in the plurality of detection units 31d, 31e, 31f, 31g,..., And the measurement time can be shortened.

Further, the phase difference θ between the first pulse light f ₁ (or f _1A ) and the second pulse light f ₂ of the inspection pulse light f generated for each of the detection pulse light emitting units 20d, 20e, 20f, 20g,. As a result, it is possible to simplify the configuration of the portion that forms the phase difference θ as compared to the case where all of the plurality of inspection pulse lights f having different phase differences θ are generated by the same circuit or the like. .

  In the first embodiment and the second embodiment, if the fact that the vibration at a certain point in the longitudinal direction of the optical fiber for detection 12 appears as a phenomenon in the frequency axis direction of the Brillouin spectrum at that point is utilized, It becomes possible to detect vibration at each position in the longitudinal direction of the detection optical fiber 12.

For example, specifically, one of the steepest frequencies of the shoulder inclination when the synthesized spectrum V _S (t, ν) is graphed (see FIG. 4B) is specified, and the spectrum intensity at this frequency is determined. By measuring the vibration, the vibration of the detection optical fiber 12 at the position where the combined spectrum V _S (t, ν) is obtained is measured. The sensitivity of this method is proportional to the slope of the spectrum, but the detected combined spectrum V _S (t, ν) is high in the distributed optical fiber sensors 10, 10a, etc. of the first and second embodiments. Since the spatial resolution is maintained and the line width can be kept small, a distributed vibration sensor with high spatial resolution and high sensitivity can be configured.

10 distributed optical fiber sensor 12 optical fiber for detection (optical fiber)
12a First end (specific end)
DESCRIPTION OF SYMBOLS 20 Inspection pulse light emission part 22 Optical pulse generation circuit 24 Short pulse light generation part 25 Long pulse light generation part 26 Phase shifter 27 Optical synthesizer 30 Spectrum detection part 31 Detection part 36 Spectrum synthesis part 40 Shift amount measurement part f Inspection pulse light (Inspection light)
f ₁ short pulse light (first pulse light)
f ₂ long pulse light (second pulse light)
h Low-pass (matching) filter (filter)
V Brillouin spectrum V _S composite spectrum θ Phase difference between short pulse light and long pulse light constituting inspection pulse light

Claims (6)

A distributed optical fiber sensor using an optical fiber as a sensor,
A first pulsed light generator that generates first pulsed light having a pulse width based on the spatial resolution of the sensor;
A second pulsed light generator that generates second pulsed light having a pulse width based on the reciprocal of the frequency resolution of the sensor;
A predetermined phase difference is provided between the first pulsed light and the second pulsed light, the first pulsed light and the second pulsed light are combined to generate inspection light, and the inspection light is transmitted to the optical fiber. An inspection light generator that emits toward a specific end of
A spectrum detector for detecting a Brillouin spectrum based on light related to a Brillouin scattering phenomenon emitted from a specific end of the optical fiber on which the inspection light is incident;
A shift amount measuring unit for measuring a Brillouin frequency shift amount due to strain or / and temperature generated in the optical fiber,
The inspection light generation unit generates a plurality of inspection lights having different phase differences from each other,
The spectrum detector corresponds to the plurality of inspection lights by applying a filter corresponding to the phase difference between the first pulse light and the second pulse light in each inspection light to the light related to the Brillouin scattering phenomenon. A detection unit for obtaining the Brillouin spectrum, and a synthesis unit for synthesizing the Brillouin spectra obtained by the detection unit,
The distributed optical fiber sensor, wherein the shift amount measuring unit measures the Brillouin frequency shift amount based on a combined spectrum that is a Brillouin spectrum combined by the combining unit.   The synthesizing unit takes a weighted sum of the Brillouin spectra so that components having a small spread in the time direction and a small spread in the frequency direction are extracted from each Brillouin spectrum obtained by the detection unit. The distributed optical fiber sensor according to claim 1.   The distributed optical fiber sensor according to claim 1 or 2, wherein the detection unit obtains a Brillouin spectrum by scanning a center frequency of the filter by signal processing. The spectrum detection unit includes a separation unit that separates the light related to the Brillouin scattering phenomenon into a horizontal component and a horizontal component,
The detection unit applies a filter corresponding to the phase difference to the vertical component and the horizontal component of the light related to the Brillouin scattering phenomenon separated by the separation unit, and then calculates the Brillouin spectrum by taking the sum of squares thereof. The distributed optical fiber sensor according to any one of claims 1 to 3, wherein: A distributed optical fiber sensor using an optical fiber as a sensor,
A first pulsed light generating unit that generates first pulsed light based on long pulsed light having a pulse width greater than the time required for light to reciprocate at a distance corresponding to the spatial resolution of the sensor;
A second pulsed light generator that generates second pulsed light having a pulse width based on the reciprocal of the frequency resolution of the sensor;
A predetermined phase difference is provided between the first pulsed light and the second pulsed light, the first pulsed light and the second pulsed light are combined to generate inspection light, and the inspection light is transmitted to the optical fiber. An inspection light generator that emits toward a specific end of
A spectrum detector for detecting a Brillouin spectrum based on light related to a Brillouin scattering phenomenon emitted from a specific end of the optical fiber on which the inspection light is incident;
A shift amount measuring unit for measuring a Brillouin frequency shift amount due to strain or / and temperature generated in the optical fiber,
The first pulsed light generation unit divides the long pulsed light into a plurality of cells having a width based on the spatial resolution, modulates the divided long pulsed light using a predetermined code sequence, and then modulates the first pulsed light. Produce light,
The inspection light generation unit generates a plurality of inspection lights having different phase differences from each other,
The spectrum detector applies a filter corresponding to the phase difference between the first pulse light and the second pulse light in each inspection light to the light related to the Brillouin scattering phenomenon, and modulates in the first pulse light generation unit And detecting a Brillouin spectrum corresponding to each of the plurality of inspection lights by performing pulse compression after demodulation, and a combining unit for synthesizing the Brillouin spectra obtained by the detection unit. And
The distributed optical fiber sensor, wherein the shift amount measuring unit measures the Brillouin frequency shift amount based on a combined spectrum that is a Brillouin spectrum combined by the combining unit. A distributed optical fiber sensor using an optical fiber as a sensor,
A predetermined phase difference is provided between the first pulse light having a pulse width based on the spatial resolution of the sensor and the second pulse light based on the reciprocal of the spatial resolution of the sensor, and the first pulse light and the second pulse light A plurality of combined pulse light generators for generating combined pulsed light,
The combined pulse light generated by each combined pulse light generation unit is combined into inspection light, and the inspection unit emits the inspection light toward a specific end of the optical fiber;
A separation unit that separates light related to the Brillouin scattering phenomenon emitted from a specific end of the optical fiber into which the inspection light is incident into separated light corresponding to each combined pulsed light before joining at the joining unit;
A plurality of detectors each for obtaining a Brillouin spectrum by applying a filter to each separated separated light;
A synthesis unit for synthesizing the Brillouin spectra obtained by each detection unit;
The shift amount measuring unit that measures the Brillouin frequency shift amount based on the combined spectrum that is the Brillouin spectrum combined by the combining unit, and
The predetermined phase difference is different for each of the combined pulse light generation units,
Each detection unit uses a filter according to the phase difference between the first pulsed light and the second pulsed light constituting the combined pulsed light corresponding to the separated light, and a distributed optical fiber sensor.