JP2001033218A - Optical fiber distortion measuring method and recording medium for realizing the method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make measurable distortion with high accuracy by providing a recording medium having recorded a program to be executed by a computer thereon, and determining a peak frequency, a peak value and a half value full width so that the power of Brillouin rear scattered light becomes maximum, taking the incident light pulse width into account. SOLUTION: The measured value of power spectrum of Brillouin scattered light in each measuring points in the longitudinal direction along an optical fiber used as a sensor in a recording medium is analyzed taking an optical frequency as a variable. Subsequently, the power spectrum of Brillouin scattered light taking the incident light power distribution when the incident light pulse time width is short into account is stored including the peak frequency, peak value and half value full width of the power spectrum to be determined in variables, and the measured spectrum is subjected to fitting to the spectrum of scattered light by a least square method to determine the peak frequency, peak value and half value full width of the spectrum to be determined. Thus, operation for determining the respective values on all longitudinal measuring points to be obtained of the optical fiber is repeated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical fiber strain measuring method for measuring strain using an optical fiber as a sensor and a recording medium for realizing the method, and more particularly, to a concrete or steel structure, The present invention relates to an optical fiber strain measurement technique for fixing a sensing optical fiber to a measurement target such as the ground and continuously measuring strain generated in the measurement target.

[0002]

2. Description of the Related Art One of commonly used strain measurement methods
One is a strain gauge. This strain gauge has the advantage that the spatial resolution is as high as a gauge length of about several mm to several cm, and that local strain can be easily measured with high accuracy. However, since this strain measurement is performed at discrete points where gauges are attached, it is necessary to attach many gauges densely for continuous measurement. In addition, it is necessary to supply power to the gauge through a power line that also serves as a signal line, and multipoint measurement has drawbacks in that not only the handling of the power line is complicated, but also that it is affected by electromagnetic noise such as lightning.

On the other hand, a strain measurement method using an optical fiber as a sensor not only enables long distance measurement over several km to several tens km along the optical fiber, but also makes it possible to perform electromagnetic measurement as described above. There is an advantage that it is not affected by noise. Thus, the strain gauge
Optical fiber strain measurement technology requires a wide range of measurement where the strain generation location is unknown, whereas the location where the remarkable strain occurs is known or predictable and the measurement range is narrow. Suitable for the case. Because of these features, the optical fiber strain measurement technology is expected to be applied to large structures such as tunnels and civil structures such as embankments.

[0004] As an example of a conventional optical fiber strain measuring technique, an optical fiber strain measuring apparatus disclosed in Japanese Patent Application Laid-Open No. 10-90121 is shown in FIG.
In the figure, reference numeral 1 denotes a light source that emits continuous light of a single wavelength, and this light source 1 is connected to an incident end of an optical splitter 2 by an optical fiber 11. The optical splitter 2 is configured to emit continuous light emitted from the light source 1 to two emission ends at an intensity ratio of 1: 1. One of the split lights emitted from the optical splitter 2, that is, the split light incident on the optical frequency shifter 3 described later is the signal light, and the other split light, that is, the split light incident on the optical combiner 6 described later. Is a reference light.

Reference numeral 3 denotes an optical frequency shifter, whose input end is connected to one output end of the optical splitter 2 by an optical fiber 12, and the optical frequency of the incident signal light is stepped at a predetermined period. The pulse width is changed so that the pulse width is converted to an optical pulse train having a time width of about 2 μs. Reference numeral 4 denotes a pulser. The input end of the pulser 4 is connected to the output end of the optical frequency shifter 3 via an optical fiber 13, and the pulse width of the incident optical pulse train is 10 ns to 1 μs.
It is configured to convert light pulses into the same order.

Reference numeral 5 denotes an optical directional coupler, which has an input end 5a, an input / output end 5b, and an output end 5c, emits light incident from the input end 5a from the input / output end 5b, and outputs the light. The light incident from 5b is emitted from the emission end 5c. The incident end 5a of the optical directional coupler 5 is connected to the pulser 4 via an optical fiber 14, and an optical pulse is incident. The light emitted from the optical directional coupler 5 is incident on the sensing optical fiber 10. Due to the emitted light, backscattered light generated in the sensing optical fiber 10 is incident on the input / output end 5b.

Reference numeral 6 denotes an optical multiplexer having two incident ends and one output end, and one of the incident ends and the optical splitter 2 described above.
Are connected by an optical fiber 15, and the other input end is connected to the output end 5 c of the optical directional coupler 5 by an optical fiber 16. The optical multiplexer 6 is configured such that the backscattered light emitted from the sensing optical fiber 10 and the reference light emitted from the optical splitter 2 are incident and multiplexed.

Reference numeral 7 denotes a photodetector. The input end of the photodetector 7 is connected to the output end of the optical multiplexer 6 by an optical fiber 17, and performs coherent detection on the incident light. It is configured to convert power to power and output it. Reference numeral 8 denotes a signal processing unit which is configured to output a measured waveform having a characteristic corresponding to a distortion generated in the sensing optical fiber 10 based on the detected optical power.

In such a configuration, continuous light emitted from the light source 1 enters the optical splitter 2 and is split into signal light and reference light. The signal light is incident on the optical frequency shifter 3, and the optical frequency is changed stepwise in time series, and is converted into an optical pulse train having a time width of about 2 μs. The optical pulse train emitted from the optical frequency shifter 3 enters the pulsing device 4 and is converted into an optical pulse having a time width of about 10 ns to 1 μs. The optical frequency shifter 3 and the pulser 4 generate an optical pulse having a desired optical frequency. The light pulse emitted from the pulsing device 4 enters the sensing optical fiber 10 via the optical directional coupler 5.

When an optical pulse is incident on the sensing optical fiber 10, it is subjected to Rayleigh scattering (scattering loss due to non-uniformity of the refractive index of the core) and Brillouin scattering in the sensing optical fiber 10 to generate backscattered light. . This backscattered light is incident on one incident end of the optical multiplexer 6 via the optical directional coupler 5 and the optical fiber 16.

The reference light emitted from the optical splitter 2 enters the optical multiplexer 6 via the optical fiber 15 and is combined with the backscattered light. The multiplexed light emitted from the optical multiplexer 6 enters the photodetector 7 and is coherently detected.
A detection signal corresponding to the received power is output to the signal processing unit 8.

Next, the relationship between the characteristics of Brillouin scattered light and distortion will be described.

The Brillouin scattered light is light that is scattered while causing a periodic change in the refractive index of the material when the light incident on the material propagates through the material and returns to the incident end. Is used, the power P _B (z, ν) of the scattered light is given as follows.

[0014]

(Equation 1)

[0015]

(Equation 2)

[0016]

(Equation 3)

[0017]

(Equation 4)

Here, z is the distance from the light pulse incident end along the optical fiber, ν is the optical frequency related to Brillouin scattered light, c is the speed of light in vacuum, n ₀ is the refractive index of the glass of the optical fiber, P is the product of the pulse time width τ of the incident light and the power, and α _z is the loss factor of the optical fiber. In order to receive the loss of the round trip between the incident end and the scattering position, the equation (1)
Has a coefficient of z of 2. g (ν, ν _B ) is a gain spectrum of Brillouin scattered light, and is given by a Lorentz function represented by Expression (2). [nu _B is g (ν, ν _B) peak frequency at which a peak value h of, w is g (ν, ν _B)
Is the full width at half maximum of p ₁₂ photoelastic coefficient of the glass of the optical fiber, lambda is the wavelength of the light pulses incident, [rho is the density of the glass of the optical fiber, v _A is the speed of sound in the optical fiber. In equation (1) that gives the reception power P _B (z, ν),

[0019]

[Outside 1]

The term represents the effect of optical fiber loss and varies with the scattering position z. However, it is considered that the frequency distribution g (ν, ν _B ) does not depend on z.

Assume that an optical pulse having an optical frequency of f ₀ is incident, thereby generating Brillouin scattered light having a peak frequency of ν _B. This difference s _B (= f ₀ −ν _B ) is called a Brillouin frequency shift and is given by the following equation.

[0022]

(Equation 5)

Where v _A is

[0024]

(Equation 6)

Is given by Where E is Young's modulus, κ
Is the Poisson's ratio. When distortion occurs in the optical fiber, v _A changes according to the relationship of Expression (6), and as a result, s _{B of} Expression (5) also changes. Therefore, if f ₀ is kept constant, ν _B changes according to the change in strain. It has been found that this change in the peak frequency ν _B is proportional to the magnitude of the strain caused by the stress acting on the optical fiber (for example, T. Horiguchi,
T.Kurashima, and M.Tateda, "Tensile strain dependenc
e of Brillouin frequency shift in silica optical f
ibers, "IEEE Photonics Technol. Lett., Vol. 1, No. 5, p
p.107-108, 1989.5). Therefore, in advance, obtains a change Δε strain epsilon, by previously obtained relation between the change .DELTA..nu _B of the peak frequency [nu _B of the power spectrum of the Brillouin scattered light, the epsilon strain from the obtained value of the peak frequency [nu _B be able to. The measurement point of the strain, that is, the distance z from the incident end of the sensing optical fiber 10 is z = Tc / (2n ₀ ) from the time T from when the light pulse enters the optical fiber to when the Brillouin scattered light is detected. Desired.
The spatial resolution Δz is given by Δz = cτ / (2n ₀ ), where τ is the light pulse time incident on the optical fiber.

The power spectrum of Brillouin scattered light is
It is affected not only by strain but also by temperature. Changes in temperature,
The resulting peak frequency is proportional to the change in peak value and inversely proportional to the full width at half maximum

[0027]

[Outside 2]

Therefore, two of them, for example,
By simultaneously solving the relationship between the peak frequency and the peak value and the strain and the temperature, it is possible to remove the influence of the temperature on the strain or to obtain both simultaneously (for example, TRParker, M. Farhadiroushan, VAHandere
k, and AJRogers, "Temperature and strain dependenc
e of the power level and frequency of spontaneous
Brillouin backscattering in optical fibers, "Opt.L
ett., vol. 22, no. 11, pp. 787-789, 1997). As described above, in order to improve the measurement accuracy of the strain, it is necessary to accurately determine the peak frequency ν _B , the peak value h, and the full width at half maximum w that give the power spectrum of the Brillouin scattered light. Various things have been done.

FIG. 10 is a flowchart for explaining a method of calculating the peak frequency ν _B , the peak value h, and the full width at half maximum w.

First, the first step is a step of preparing for analysis at each measurement point along the sensing optical fiber using the optical frequency as a variable. For each measurement point, the optical frequency is used as a variable. And sorting the measured Brillouin scattered light power. As described above, in the conventional strain measuring apparatus, an optical pulse is injected with a fixed optical frequency, the power of the Brillouin scattered light is measured, and then the optical frequency is changed by a predetermined amount to change the optical frequency again. Fixing and measuring the scattered light power are repeated. By this processing, power with the measurement point as a variable is obtained for each optical frequency, and this is sorted with respect to the optical frequency. More specifically, the distance to the i-th measurement point is represented by z _i (i = 1 to I, I is the total number of measurement points),
The j-th optical frequency is ν _j (j = 1 to J, J is all measured frequencies), and the measured power value at that time is

[0031]

[Outside 3]

Then, ν _j is set to a predetermined optical frequency value, and z _i is changed.

[0033]

[Outside 4]

After the measurement up to the distance z _I is completed, ν _j is changed to the next optical frequency and measurement is performed up to ν _J. In this first step, a certain distance z _i is selected and fixed, and for a light frequency v _j

[0035]

[Outside 5]

This is the step of sorting. as a result,
Got

[0037]

[Outside 6]

Have the same distance z _i , are sorted with the optical frequency ν _j as a variable, and are processed independently for each z _i. In the following this

[0039]

[Outside 7]

Will be described.

The second step is the power obtained in the first step.

[0042]

[Outside 8]

On the other hand, the peak frequency ν _B giving the power spectrum of the Brillouin scattered light, the peak value h,
This is the step of calculating the full width at half maximum w. Since the equation (2) is a nonlinear function for the optical frequency ν, an analytical solution cannot be obtained. Therefore, a method of linearizing and solving as follows is proposed (for example, CNPannell, J. Dhliwayo, D.
J.Webb, "How to estimate the accuracy of a Brilloui
n distributed temperature sensor, "Proc. OFS'97, p
p.524-527, 1997). However, since this document deals only with the peak frequency ν _B , a method of calculating the peak value h and the full width at half maximum w will also be described.
In this method, the reciprocal of the equation (2) is calculated, and the evaluation function _Er is obtained.

[0044]

(Equation 7)

Consider the following. Here, the coefficients a ₁ , a ₂ and a ₃ are respectively

[0046]

(Equation 8)

[0047]

(Equation 9)

[0048]

(Equation 10)

Where W _m is a weight for avoiding a change caused by taking the reciprocal, and σ _m is a standard deviation of noise included in the measured value of power. Also, instead of performing the fitting calculation on the entire measured value, a certain threshold is usually set, and the fitting calculation is performed on the vicinity of the peak frequency having a higher value. M is used as a parameter representing a portion to be used. That is, m (= 1 to M) is a part of j (= 1 to J). From equations (8) to (10), the peak frequency is ν _B , the peak value h, and the full width at half maximum w are:

[0050]

[Equation 11]

[0051]

(Equation 12)

[0052]

(Equation 13)

Is obtained. The coefficients a ₁ , a ₂ , and a ₃ are obtained from the condition for minimizing the equation (7) using the least squares method.
The peak frequency ν _B , the peak value h, and the full width at half maximum w are calculated by substituting the results into the equations (11) to (13).

The third step is a step of repeating the second step for all measurement points to be obtained along the optical fiber. That is, while changing the distance z _i

[0055]

[Outside 9]

From

[0057]

[Outside 10]

Then, the second step is repeated until the distance z _I.

[0059]

In the conventional strain measuring method described above, the time width of the incident light pulse is sufficiently long and the power spectrum of the scattered light is given by the Lorentz function expressed by the equation (2). Is assumed. However, it has been pointed out that if the time width of the light pulse is shortened to increase the spatial resolution, the power spectrum of the scattered light differs from the Lorentz function, and the power will be distributed over a wide range.

[0060]

[Outside 11]

Since the power spectrum has not been analyzed so far, an appropriate function to be applied has not been clarified, and the peak frequency ν _B , the peak value h,
There was a problem that the measurement accuracy of the full width at half maximum w deteriorated.
In addition, v _m ⁴ is used as a practical weight W _m , but this has not only a clear basis but also causes an error in analyzing the measurement data.

The present invention has been made in view of such a problem, and an object of the present invention is to provide an optical fiber strain measurement method capable of obtaining the power spectrum of Brillouin scattered light and improving the accuracy of strain measurement. Another object of the present invention is to provide a computer-readable recording medium on which a program for causing a computer to execute the method is recorded.

[0063]

According to the present invention, in order to achieve the above object, an optical pulse is incident on an optical fiber used as a sensor, and this is one of backscattered light generated by the incident optical pulse. Measure the power spectrum of the Brillouin scattered light, determine the peak value of the power spectrum, the peak frequency that gives the peak value, and the full width at half maximum,
In the optical fiber strain measuring method for obtaining a strain distribution in a length direction generated in the optical fiber, the measurement of the power of the Brillouin scattered light using the optical frequency as a variable for each measurement point in the length direction along the optical fiber. A first step of sorting the values, and the power spectrum of the Brillouin scattered light in consideration of the incident light power distribution when the time width of the incident light pulse is short, the peak frequency, peak value and full width at half maximum of the power spectrum to be determined. Is described in the form of variables, the measured power spectrum is subjected to least squares fitting to the power spectrum of the Brillouin scattered light, the peak frequency, peak value and full width at half maximum of the power spectrum to be determined A second step of determining the above, and for all measurement points in the length direction to be determined of the optical fiber There are, is made of a third step of repeatedly performing the second step.

Further, the present invention provides a recording medium on which a program for causing a computer to execute the optical fiber strain measurement method having the above steps is provided, and the power of the Brillouin backscattered light is maximized in consideration of the incident light pulse width. The peak frequency, the peak value, and the full width at half maximum are determined so that distortion can be measured with high accuracy.

[0065]

Embodiments of the present invention will be described below.

FIG. 1 is a flowchart showing one embodiment of the optical fiber strain measuring method of the present invention. As the optical fiber strain measuring device, the conventional strain measuring device shown in FIG. 9 can be used.

Therefore, a method for theoretically obtaining the power spectrum of the measured Brillouin scattered light in consideration of the case where the light pulse width is narrowed, and then applying the measured power to the distribution will be described below. .

As an incident light pulse, an ideal light pulse in which the pulse width is τ in the time domain and the power is concentrated at the optical frequency f ₀ in the frequency domain is considered. The electric field E _P (t) of this light pulse is expressed by E ₀ as a constant.

[0069]

[Equation 14]

The power spectrum P _P (f, f ₀ ) of this light pulse is given by

[0071]

(Equation 15)

Is given by Here, P ₀ is a constant, and the power spectrum of the incident light pulse is a function of the optical frequency f related to the incident light pulse and the optical frequency f ₀ at which the incident light pulse power is concentrated. It is described as P _P (f, f ₀ ) for explicit representation. FIG.
Is a diagram showing an outline of the equation (15). That is, when the time width τ of the optical pulse is sufficiently long, 1 / τ ≒ 0, and the optical pulse whose power is concentrated at f ₀ is incident on the optical fiber. Conversely, as the time width τ of the optical pulse becomes shorter, the peak value becomes smaller and an optical pulse having a spectrum in which power is distributed over a wide range is incident on the optical fiber.

Now, assuming that the Brillouin scattered light generated by the incident light pulse has no phase correlation in the electric field and the principle of superposition of power is satisfied, P _P (f, f ₀ ) expressed by equation (15) The power spectrum of the Brillouin scattered light when is incident is approximately determined. Here, the frequency-dependent component of Brillouin scattered light

[0074]

(Equation 16)

Is considered. When the optical frequency f ₀ of the incident light pulse shifts to the peak frequency ν _B of the Brillouin scattered light, the relationship between the incident light pulse and the power spectrum of the Brillouin scattered light becomes as shown in FIG. That is, the optical frequency of the Brillouin scattered light shifts from the optical frequency f ₀ of the incident optical pulse to an optical frequency ν around the optical frequency ν _B at which the power is maximum. Considering a small optical frequency width df near the optical frequency f on the power spectrum P _P (f, f ₀ ) of the incident light pulse, the power spectrum of the Brillouin scattered light generated thereby has a peak frequency of f−s.
_{In B} , the full width at half maximum becomes a Lorentz function of w. As described above, this Brillouin frequency shift s _B is the difference between f ₀ and ν _B. Incident light pulse power _P P (f, f ₀₎ contained in the df power dH (ν) dν of Brillouin scattered light generated by dt is given by the following equation.

[0076]

[Equation 17]

Therefore, the integration of equation (17)

[0078]

(Equation 18)

The power spectrum H (ν) of the entire Brillouin scattered light is obtained by calculating the following equation. However, in order to simplify the formula, the constant P ₀ was set to 1. By integrating the equation (18), H (ν) is obtained as follows.

[0080]

[Equation 19]

Where α and A are respectively

[0082]

(Equation 20)

[0083]

(Equation 21)

Is as follows. Therefore, power is concentrated in a sufficiently narrow optical frequency range, and the time width τ of the optical pulse is long enough to be regarded as continuous light or continuous light (A = πτ).
w is large) in the above equation

[0085]

(Equation 22)

Therefore, the power spectrum of the Brillouin scattered light generated thereby becomes the original Lorentz function. However, as the time width τ of the incident light pulse becomes shorter, the peak value of the power spectrum of the Brillouin scattered light becomes smaller and the power is distributed over a wider range due to the effect of the second term of the equation (19).

The power peak value H (0) in consideration of the time width τ of the incident light pulse is τh (A−1 + e ^−A ) /
The peak value is τ (A−1 + e ^−A ) since h is the case where A is not considered (in the case of the Lorentz function).
/ A times. FIG. 4 shows the results of theoretically and experimentally determining the full width at half maximum of H (α) when the time width τ of the incident light pulse is changed based on the full width at half maximum of the Lorentz function. In FIG. 4, the black circles in the figure indicate the results of the experiment using the strain measuring device shown in FIG. 9, and the solid lines indicate the results of the theory of the present invention. As can be seen from FIG. 4, it is confirmed that they are well matched.

As described above, the power spectrum H of the Brillouin backscattered light in consideration of the time width of the incident light pulse.
(Ν) is obtained.

Next, the conjugate gradient method used to apply the measured power spectrum to the power spectrum obtained by the method of the present invention will be described.

The conjugate gradient method is one of the typical non-linear optimization methods without restrictions, and is a method of obtaining y that minimizes the function e = q (y) in the form of a successive approximation method (for example, Toshihide Ibaraki, Masao Fukushima, "Optimization Methods," Kyoritsu Shuppan, 19
96). In this conjugate gradient method, when the k-th approximate solution is y _k , a correction of Δy _k is added to the k-th approximate solution to obtain the next (k + 1) -th approximate solution y _{k + 1} .

[0091]

(Equation 23)

Is updated. At the time of this update, Δy _k is _defined as the product of the correction amount μ _k and the correction direction vector d _k

[0093]

(Equation 24)

And μ _k and d _k are determined as follows. Here, a conjugate direction vector is used as d _k . In this case, if e = q (y) is a quadratic function, the conjugate direction vector d _k becomes the gradient direction vector r _k, that is,

[0095]

(Equation 25)

By using

[0097]

(Equation 26)

Can be expressed as follows. Where d ₁ = r ₁
It is. On the other hand, the correction amount mu _k with d _k function

[0099]

[Equation 27]

In consideration of the above, the value is determined as a value that minimizes φ (μ _k ) under the condition of μ _k > 0. Μ determined in this way
_k , d _k, and Δy _k is calculated.
The accuracy of the approximate solution is increased one after another by a series of repeated calculations substituted into 3). At the minimum point of the function e

[0101]

[Equation 28]

Therefore, the calculation is repeatedly performed until the condition of Expression (28) is satisfied, and the minimum point of the function e is actually calculated numerically.

Next, a description will be given of a fitting calculation method when the conjugate gradient method is used.

Now, the m-th measurement frequency is ν _m , and the power measured there is

[0105]

[Outside 12]

And the evaluation function E _r

[0107]

(Equation 29)

Consider the following. Here, M is given in advance because it is the number of optical frequencies used for measurement. Y _k , r _k described above
Are each

[0109]

[Equation 30]

[0110]

(Equation 31)

Since the [0111], actually peak frequency [nu _B is a component of r _k, the peak value h, the gradient direction of the full width at half maximum w vector _{∂E r / ∂ν B, ∂E r} / ∂h, ∂E r Calculating / ∂w is as follows.

[0112]

(Equation 32)

[0113]

[Equation 33]

[0114]

(Equation 34)

However,

[0116]

(Equation 35)

[0117]

[Equation 36]

[0118]

(37)

Is as follows.

Hereinafter, the optical fiber strain measuring method of the present invention will be described more specifically with reference to the flowchart shown in FIG.

The first step (S1) of the present invention is similar to the first step of the conventional method, and the Brillouin scattered light measured using the optical frequency as a variable at each measurement point along the sensing optical fiber is used. Power of

[0122]

[Outside 13]

In this step, the power sorted at each measurement point with respect to the optical frequency is

[0124]

[Outside 14]

This is the step of obtaining

The second step (S2) is to determine the power spectrum of the Brillouin scattered light in consideration of the incident light power distribution when the time width of the incident light pulse is short, to determine the peak frequency ν _B and peak value of the power spectrum to be determined. h and the full width at half maximum w are described as variables, and the measured power spectrum is the measured power of the Brillouin scattered light.

[0127]

[Outside 15]

In this step, the peak frequency ν _B , peak value h, and full width at half maximum w of the power spectrum to be determined are determined by performing fitting by the least square method. In addition,

[0129]

[Outside 16]

As described in the description of the prior art,

[0131]

[Outside 17]

The portion having power equal to or higher than the threshold value. Equation (19) representing the frequency distribution H (ν) of the Brillouin scattered light described in a form including these variables to be determined, that is, the peak frequency ν _B , the peak value h, and the full width at half maximum w.
To (21) and the equation (29) representing the sum of the measurement errors included in the power measured for all the measurement frequencies, by fitting using the least squares method, the peak frequency ν _{B of} the power spectrum to be determined, The peak value h and the full width at half maximum w are obtained.

Next, the measurement method of the present invention will be specifically described by simulation, and the effectiveness of the strain measurement method of the present invention will be described.

As an example of the true values of the peak frequency ν _B , the peak value h, and the full width at half maximum w, the peak frequency ν _{B is set} to 1
0.5 GHz, peak value h is 1, and full width at half maximum is 81.4.
MHz, and the time width τ of the light pulse was 10 ns. Substituting these values into equations (19) to (21) gives 10 MHz
H (α) was calculated for each case. The calculated value is taken as the true value of the power spectrum, and normal noise having a standard deviation of 1/100 of H (0) is added thereto, and the measured value of the Brillouin scattered light power is obtained.

[0135]

[Outside 18]

[0136] Using 1/5 of the peak value as the threshold value, the power measurement value of the optical frequency part larger than that

[0137]

[Outside 19]

The calculation is performed for the peak frequency ν _B ,
The peak value h and the full width at half maximum w were determined. FIG. 5 is a diagram showing an example of a result obtained by applying a measured power spectrum to a power spectrum obtained by the method of the present invention. FIG.
In the figure, the true value of the power spectrum of the scattered light given by the Lorentz function and the true value of the power spectrum of the scattered light with a time interval of the light pulse of 10 ns are represented by a thin line, and the measured values obtained by the above-described procedure are indicated by circles The curve obtained by the method of applying the reciprocal described in the conventional method and the curve obtained by the method of the present invention are indicated by solid lines and broken lines. Peak frequency ν determined by the method of the present invention
_B , the peak value h, and the full width at half maximum w are 10.500, respectively.
0 GHz, 1.0015, 80.33 MHz, and 10.4998 GHz, 0.6539, 1
31.41 MHz. In the conventional method, the peak value h and the full width at half maximum w are not correctly obtained because the time width of the light pulse is not taken into account and the method is simply applied to the Lorentz function. Nevertheless, as shown in FIG.
The peak frequency ν _B is accurately determined because the Lorentz function to which the measured value is applied is also symmetric with respect to the peak frequency ν _B.

The third step is a step of repeating the second step for the data of all the measurement points to be obtained.

[0140]

[Outside 20]

Are obtained, and the second step is repeatedly performed on them.

The obtained frequency distributions of the peak frequency ν _B , the peak value h, and the full width at half maximum w are shown by solid lines in FIGS. For comparison, the distribution obtained by the conventional method is shown by a thin line. For the peak frequency ν _B ,
In each case, the calculation results are distributed around the true value. This is because, as described above, each method is applied to a function symmetrical with respect to the peak frequency ν _B. On the other hand, it can be understood from FIGS. 7 and 8 that the peak value h and the full width at half maximum w are distributed around values different from the true values, and are not correctly obtained.

Here, the distribution of the obtained peak frequency ν _B , peak value h, and full width at half maximum w is quantitatively compared using the median value. Regarding the peak frequency ν _B , both the method of the present invention and the conventional method are 10.50000 GHz, and the result is sufficient accuracy of 1 × 10 ⁻⁶ or less in terms of distortion error. On the other hand, the peak value h is distributed around 0.9988 and the true value 1 in the method of the present invention, while it is distributed around 0.6523 in the conventional method. Also, the full width at half maximum w is distributed around 80.1725 MHz and the true value of 80 MHz in the method of the present invention, whereas it is 130.times. In the conventional method.
Distributed around 5703 MHz. As a result, it can be seen that the measurement accuracy of the method of the present invention is improved by about 300 times in both the peak value h and the full width at half maximum w compared to the conventional method.

The above result is corrected by dividing the result by the conventional method for the peak value h by τ (A-1 + e− ^A ) / A, and correcting the full width at half maximum w by using the relationship in FIG. In this case, the peak value h is calculated to be 1.0283, and the full width at half maximum w is calculated to be 77.5450 MHz. Even if the result calculated by the conventional method is corrected, the method of the present invention can be compared with the conventional method. It is confirmed that the precision is required about 24 times and about 15 times. This is because, although the measured power value is the power spectrum given by the equation (19), it is applied to a different Lorentz function.

Further, by using a computer-readable recording medium on which a program for causing a computer to execute the above-described optical fiber strain measurement method is used, the computer can easily perform the strain measurement process. That is, on the recording medium, the step of analyzing the measured value of the Brillouin scattered light power at each measurement point in the length direction along the optical fiber used as a sensor, using the optical frequency as a variable, and the time width of the incident light pulse is short. The power spectrum of the Brillouin scattered light considering the incident light power distribution is stored in a form that includes the peak frequency, peak value, and full width at half maximum of the power spectrum to be determined as variables, and the measured power spectrum is Brillouin scattered. A step of performing fitting by the least squares method on the power spectrum of light to determine the peak frequency, peak value, and full width at half maximum of the power spectrum to be determined, and all measurement points in the length direction of the optical fiber to be determined. Repeating the step of determining the peak frequency, peak value and full width at half maximum; and The program for executing the optical fiber strain measuring method in a computer having By recording, it is possible to facilitate the calculation of strain measurement.

[0146]

As described above, according to the present invention,
In consideration of the incident light pulse width, the peak frequency, peak value, and full width at half maximum where the power of the Brillouin backscattered light is maximum are determined, so that the optimal fitting calculation can be performed regardless of the pulse width. As a result, the peak frequency, peak value, and full width at half maximum can be determined with high accuracy. This not only makes it possible to measure strain with high accuracy,
It is possible to remove the influence of temperature with high accuracy.

Further, since the steps for realizing the strain measurement method are recorded on a recording medium in a computer-readable manner as a program for causing a computer to execute, it is possible to easily use the recording medium to perform a calculation process for strain measurement. Can be

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating a method of calculating a peak frequency, a peak value, and a full width at half maximum, showing an embodiment of an optical fiber strain measurement method according to the present invention.

FIG. 2 is a diagram for explaining a power spectrum of an incident light pulse.

FIG. 3 is a diagram for explaining a relationship between an incident light pulse and a power spectrum of Brillouin scattered light generated thereby.

FIG. 4 is a comparison diagram of a full width at half maximum obtained in consideration of a time width of an optical pulse and a full width at half maximum without consideration;

FIG. 5 shows the power spectrum obtained by the method of the present invention,
FIG. 9 is a diagram illustrating an example of a result obtained by applying a measured power spectrum.

FIG. 6 is a diagram for explaining a frequency distribution of the obtained peak frequency.

FIG. 7 is a diagram for explaining a frequency distribution of the obtained peak values.

FIG. 8 is a diagram for explaining a frequency distribution of the obtained full width at half maximum;

FIG. 9 is a block diagram illustrating an example of a conventional strain measurement device using an optical fiber.

FIG. 10 is a flowchart for explaining a conventional calculation method for the measured peak frequency, peak value, and full width at half maximum of Brillouin scattered light.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 light source 2 optical splitter 3 optical frequency shifter 4 pulser 5 optical directional coupler 6 optical coupler 7 photodetector 8 signal processing unit 10 sensing optical fiber

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2F065 AA06 AA65 CC00 CC40 DD03 FF12 FF42 GG22 LL00 LL02 LL57 PP22 QQ18 QQ29 QQ44 QQ45 UU05 2G086 CC02 DD05

Claims (2)

[Claims] 1. A light pulse is incident on an optical fiber used as a sensor, a power spectrum of Brillouin scattered light which is one of backscattered light generated by the incident light pulse is measured, and a peak value of the power spectrum is measured. In the optical fiber strain measuring method for determining the peak frequency giving the peak value and the full width at half maximum and determining the strain distribution in the length direction generated in the optical fiber, the length in the length direction along the optical fiber For each measurement point,
A first step of sorting the measured value of the power of Brillouin scattered light with the optical frequency as a variable, and the power spectrum of the Brillouin scattered light considering the incident light power distribution when the time width of the incident light pulse is short,
The peak frequency and peak value of the power spectrum to be determined and the peak value and the full width at half maximum are described in a form including variables, and the measured power spectrum is subjected to fitting by the least square method with respect to the power spectrum of the Brillouin scattered light,
A second step of determining the peak frequency, peak value, and full width at half maximum of the power spectrum to be determined; and repeating the second step for all measurement points in the length direction of the optical fiber to be determined. 3. A method for measuring optical fiber strain, comprising: a third step; 2. A step of analyzing the measured value of the power of Brillouin scattered light at each measurement point of the optical fiber using the optical frequency as a variable, and determining the power spectrum of the Brillouin scattered light in consideration of the time width of the incident light pulse. The peak frequency, peak value, and full width at half maximum of the power spectrum to be recorded are recorded in the form of variables, and the measured power spectrum is subjected to least squares fitting to the power spectrum of the Brillouin scattered light. Determining the peak frequency, peak value and full width at half maximum of the power spectrum to be performed, and for all of the measurement points, repeating the step of determining the peak frequency, peak value and full width at half maximum, A program for causing a computer to execute the optical fiber strain measurement method. A computer-readable recording medium that has been recorded.