WO2025017774A1 - Signal processing method for phase otdr - Google Patents

Abstract

The present invention is a device comprising a signal processing device that calculates an oscillation waveform using the phase of scattered light from a frequency-multiplexed optical pulse being reflected or scattered by optical fiber. The signal processing device includes a functional unit that: selects a reference optical frequency from a plurality of frequency-multiplexed optical frequencies; rotates the phase of a scattered light vector of another optical frequency among the scattered light vectors of the plurality of optical frequencies, using the phase of the reference optical frequency as a reference; and, thereafter, averages the scattered light vectors of the plurality of optical frequencies. To select the reference optical frequency, the functional unit temporally tracks and dynamically selects an optical frequency having a greater scattered light intensity.

Description

Signal processing method in phase OTDR

The present invention relates to a phase OTDR (Optical Time Domain Reflectometer) that measures the phase of scattered light signals from each point in a measured optical fiber.

As a means of measuring the physical vibrations applied to an optical fiber in a distributed manner along the length of the optical fiber, a method known as Distributed Acoustic Sensing (DAS) is known in which a pulsed test light is injected into the optical fiber under test and backscattered light due to Rayleigh scattering is detected (Non-Patent Document 1).

There is a method called DAS-P (DAS-phase) that measures the phase of the scattered light signal from each point of the optical fiber under test and measures the change in phase over time. With DAS-P, the phase changes linearly with respect to the change in the optical path length of the optical fiber due to vibration, and the rate of change can be considered to be approximately the same at each point along the optical fiber's length, making it possible to quantitatively measure vibration and faithfully reproduce the vibration waveform applied to the optical fiber under test (for example, Non-Patent Document 2).

In DAS-P, which uses a single-frequency pulse, interference between scattered lights within the pulse width creates points where the scattered light intensity is low, resulting in a deterioration of sensitivity. One method of preventing this deterioration in sensitivity is to perform frequency multiplexing and average the signals at different frequencies (Patent Document 1). Hereinafter, the method disclosed in Patent Document 1 will be referred to as Prior Art 1. Prior Art 1 makes use of the fact that the points where the scattered light intensity is low change with different frequencies.

In the prior art 1, as an efficient averaging method, specifically, averaging is performed in the following scattered light vector state.
Step S01: Using measured data at multiple times, calculate the difference in phase offset value between the scattered light vector of a frequency component selected as an internal reference of the multiplexed frequency components (hereinafter referred to as the "reference light frequency") and the scattered light vectors of the other frequency components. Using the calculated difference in phase offset value, calculate the phase rotation angle for correcting the phase offset of each light frequency.
Step S02: The scattered light vector of each frequency component at each time is rotated by the phase rotation angle, and then the scattered light vectors after rotation of different optical frequencies are averaged to calculate an optical frequency average vector, and the angular change of the optical frequency average vector is calculated.
Step S03: Using the angle change, the phase difference between two points separated by the gauge length is calculated, and a phase unwrapping process is performed to calculate a vibration waveform.

In particular, the details of step S01 are as follows: where t is the sampling interval of vibration, M points of measurement data from measurement time 0 to (M-1)t (M is a natural number) are used, the scattered light vector of the reference optical frequency (let it be f ₁ ) at each time mt (m is an integer from 0 to (M-1)) is r ₁ (mt, z) (dependent on the distance z from the input end since it depends on the optical fiber position), and the scattered light vector of each other optical frequency f _i (i is an integer from 2 to N, N represents the number of frequency multiplexing) is r _i (mt, z). In prior art 1, the reference optical frequency can be set arbitrarily.

Step S01-1: Calculate the angle θ ₁ (mt,z) of the scattered light vector r ₁ (mt,z) of the reference light frequency at each time and each optical fiber point. If the scattered light vector r ₁ (mt,z) is considered to be a complex vector on the complex plane, that is, a complex number, the angle θ ₁ (mt,z) can be calculated by arg[r ₁ (mt,z)], where arg is an operator that gives the argument of a complex number.

Step S01-2: Rotate the scattered light vector r _i (mt,z) of each optical frequency at each time and each optical fiber point by an angle -θ ₁ (mt,z) to obtain r _{i _rot} (mz,t). If the scattered light vector r _i (mt,z) is a complex vector on a complex plane, the rotated vector r _{i _rot} (mz,t) can be calculated as exp[-θ ₁ (mt,z)] × r _i (mt,z).

Step S01-3: After rotating the scattered light vector of each optical frequency at each optical fiber point, the vector r _{i _ rot} (mt, z) is averaged over M points, where m is the measurement data used, from 0 to (M-1), to calculate the time-averaged vector r _{i _ avet} (z). In actual calculations, vector summation may be used instead of the vector average over M points. Regardless of which method is used, the same results will be obtained in subsequent processing.

Step S01-4: Calculate the phase rotation angle α _i (z) when rotating the scattered light vector of each optical frequency at each optical fiber point in step S02 as −arg[r _{i _ avet} (z)] using the angle of the time-averaged vector. Since the value of +arg[r _{i _ avet} (z)] is the difference in phase offset value, it is possible to correct the difference in phase offset by rotating by −arg[r _{i _ avet} (z)].

The phase rotation angle α _i (z) given in step S01-4 is a value in which noise has been efficiently reduced due to the time averaging performed by the steps up to step S01-3. By using this phase rotation angle α _i (z) to calculate the optical frequency average vector in step S02, the signal-to-noise ratio of the final vibration waveform is improved and waveform distortion can be suppressed.

In step S01 of conventional technology 1, the phase rotation angle of each optical frequency is calculated using M points of measurement data from measurement time 0 to (M-1)t (M is a natural number), and in step S02, the calculated phase rotation angle is used to perform frequency averaging on M points of measurement data from measurement time 0 to (M-1)t (M is a natural number) and on measurement data at other times.

The optimal value of the phase rotation angle calculated by the method of step S01 in prior art 1 varies over time due to changes over time in optical properties such as the laser oscillation frequency, temperature changes in the measured optical fiber itself, and application of large dynamic strain to the measured optical fiber. Therefore, in long-term distributed vibration measurements, it is necessary to continuously update the optimal value of the phase rotation angle and perform the processes from step S02 onwards. However, prior art 1 does not disclose a method for performing the processes from step S02 onwards while continuously updating the optimal value of the phase rotation angle.

A solution to this problem has been proposed in Prior Application 1 (PCT/JP2022/34477). In the method described in Prior Application 1, the processes from step S02 onwards are carried out while continuously updating the optimum value of the phase rotation angle. Hereinafter, the method described in Prior Application 1 will be referred to as Prior Application Invention 1. By using Prior Application Invention 1, it is possible to deal with changes in the optimum value of the phase rotation angle.

However, in Prior Invention 1, the phase rotation angle is calculated and updated for each block with a time length of Mt (M is a natural number corresponding to the number of data points), so the value of the phase rotation angle changes discontinuously each time it moves to the next block. Therefore, a phenomenon is observed in which the final vibration waveform also changes discontinuously between the measurement data points at which the next block is switched to. For example, a phenomenon occurs in which the vibration waveform changes discontinuously between time (MK-1)t, which is the last measurement data point of block K, and time MKt, which is the first measurement data point of block K+1.

As a countermeasure to this, in the method (prior application invention 2) described in prior application 2 (PCT/JP2023/4382), in order to remove discontinuous changes in the phase rotation angle used to average the scattered light vectors of different frequency components, a time window of window length Mt (M is a natural number corresponding to the number of data points) used in signal processing is set in advance, and the phase rotation angle is calculated using the measurement data contained in the time window while moving the time window, and this phase rotation angle is used to average the scattered light vectors of the frequency components at the center of the time window or at a time that is a preset number of measurement data points behind. Calculating the phase rotation angle while moving the time window is a moving average in a broad sense, and it is possible to obtain a smoothly changing phase rotation angle that does not include the discontinuous changes. By using a smoothly changing phase rotation angle, discontinuous changes in the vibration waveform can be removed.

In Prior Art 1, Prior Invention 1, and Prior Invention 2, when averaging scattered light vectors of different optical frequencies, an arbitrary optical frequency is used as a reference, and the scattered light vectors of the different optical frequencies are rotated, and then the scattered light vectors of the different optical frequencies are averaged. In particular, Prior Invention 1 and Prior Invention 2 focus on the fact that the difference between the phase offset of the scattered light vector of an optical frequency different from the reference optical frequency and the phase offset of the scattered light vector of the reference optical frequency changes over time when large vibrations, temperature changes, etc. occur, and by updating the phase rotation angle in accordance with these changes, highly sensitive phase calculations are possible throughout the entire measurement time.

In other words, the vibration waveform finally obtained by updating the phase rotation angle in Prior Application Invention 1 and Prior Application Invention 2 has reduced phase noise and reduced effects of noise-induced phase connection errors compared to a vibration waveform obtained using a signal of only one of the selected reference optical frequency or any other optical frequency, even in the event of large vibrations or temperature changes.

However, if the effects of noise are removed, the final vibration waveform will follow the vibration waveform of the reference optical frequency, and therefore vibration waveform distortion not caused by noise will remain the vibration waveform distortion of the reference optical frequency, and cannot be effectively suppressed by Prior Application Invention 1 and Prior Application Invention 2. Vibration waveform distortion not caused by noise mainly includes vibration waveform distortion caused by interference between scattered light from Rayleigh scatterers randomly distributed in the optical fiber, and is difficult to remove by improving the device configuration, etc. Therefore, Prior Application Invention 1 and Prior Application Invention 2 have the problem of being unable to faithfully measure vibrations occurring in the optical fiber.

As a countermeasure to this, in the method (prior invention 3) described in prior application 3 (PCT/JP2023/11104), each of the multiplexed optical frequencies is selected as a reference optical frequency, an oscillatory waveform is calculated using the method of prior invention 1 or prior invention 2, and the waveforms obtained by the calculation for the number of multiplexed frequencies are averaged to obtain a final oscillatory waveform. The oscillatory waveform distortion of the oscillatory waveform obtained when each of the multiplexed optical frequencies is selected as a reference optical frequency corresponds to the waveform distortion when the optical frequency selected as the reference optical frequency is used alone, so the oscillatory waveform distortion can be reduced by switching and averaging the reference optical frequencies.

However, unlike Prior Art 1, the methods of Prior Application Invention 1, Prior Application Invention 2, and Prior Application Invention 3 update the phase rotation angle used in step S02 and onward, and are therefore affected by temporal fluctuations in the phase rotation angle due to measurement noise. In Prior Application Invention 1 and Prior Application Invention 2, a reference optical frequency is selected at each point on the optical fiber, but if the optical intensity of the selected reference optical frequency is low, the estimation accuracy of the phase rotation angle deteriorates. Prior Application Invention 1 and Prior Application Invention 2 also devise ways to improve the estimation accuracy of the phase rotation angle by performing time averaging within the window length Mt, but it is not possible to completely eliminate the deterioration of the estimation accuracy of the phase rotation angle. Therefore, assuming that the optical intensities of optical frequencies other than the reference optical frequency are the same, the smaller the optical intensity of the selected reference optical frequency, the greater the noise-induced fluctuations included in the time change of the phase rotation angle.

When a scattered light vector of an optical frequency other than the reference optical frequency is rotated using a phase rotation angle that includes large fluctuations, the orientation of the scattered light vector after rotation also fluctuates significantly with respect to the orientation of an ideal scattered light vector that is not affected by noise. The scattered light vector with a fluctuating orientation is added to the scattered light vector of a reference optical frequency with low optical intensity (i.e., short vector length) to calculate the optical frequency average vector, so the orientation of the optical frequency average vector also fluctuates significantly. Therefore, the influence of this fluctuation is reflected in the vibration waveform calculated based on the angle of the optical frequency average vector, resulting in vibration waveform distortion. In the prior invention 3, the vibration waveform when each of the multiplexed optical frequencies is selected as the reference optical frequency is calculated and averaged, so the vibration waveform when a reference optical frequency with low optical intensity is selected is also included in the average and is similarly affected by the temporal fluctuations of the phase rotation angle.

Patent Publication No. 2020-169904 (NTT, registered)

Ali. Masoudi, T. P. Newson, “Contributed Review: Distributed optical fiber dyn amic strain sensing.” Review of Scientific Instruments, vol. 87, p. 011501 (2016) Kenichi Nishiguchi, Cheolhyun Lee, Kuzik-Arter, Mitsunori Yokoyama, and Kinzo Masuda, "Prototype of distributed acoustic wave sensor using optical fiber and its signal processing," IEICE Technical Report, 115 (202), pp. 29-34 (2015) Yoshifumi Wakisaka, Daisuke Iida, Hiroyuki Oshida , and Nazuki Honda, “Fading Suppression of Φ-OTDR” With the New Signal Processing Methodology of Comp lex Vectors Across Time and Frequency Domains,” J. Lightwave Technology. 39, 4279-4293 (2021)

The present invention aims to reduce vibration waveform distortion caused by noise-induced fluctuations in the phase rotation angle used to average frequency-multiplexed scattered light signals.

The phase OTDR system of the present invention comprises a measurement device that measures scattered light signals resulting from reflection or scattering of frequency-multiplexed optical pulses in an optical fiber, and a signal processing device that calculates a vibration waveform in the optical fiber using the scattered light signals measured by the measurement device. The signal processing device of the present invention executes a signal processing method for generating a final vibration waveform from the scattered light signals of the multiple optical frequencies, in which vibration waveform distortion due to fluctuations in the phase rotation angle is reduced.

The signal processing device of the present invention comprises a functional unit which selects a reference optical frequency from a plurality of frequency-multiplexed optical frequencies, rotates scattered light vectors of other optical frequencies among scattered light vectors obtained from measurement data of the plurality of optical frequencies using the reference optical frequency as a reference, and then averages the scattered light vectors of the plurality of optical frequencies;
In selecting the reference light frequency, the functional unit dynamically selects the reference light frequency by tracking over time the light frequency at which the intensity of the scattered light is greater.

The functional unit also averages the scattered light vectors of the multiple optical frequencies at different times using the scattered light vector of the selected reference optical frequency, and performs signal processing based on the change between the optical frequency average vectors at different times obtained by the averaging. As a result, the signal processing device of the present invention includes the functional unit and obtains a vibration waveform based on the angle change of the optical frequency average vector obtained by the averaging.

The signal processing device of the present invention provides a selection rule whereby, when selecting the reference optical frequency, optical frequencies with low scattered light intensity and short vector length are excluded from the selection of the reference optical frequency, and only optical frequencies with high scattered light intensity and long vector length are included in the selection of the reference optical frequency. Since the scattered light intensity varies over time, the selection of whether to exclude or include a frequency in the selection of the reference optical frequency is dynamically updated as necessary along the time axis at each point on the fiber.

Specifically, when the present invention is applied to a method of selecting one reference optical frequency at each point on a fiber, the optical frequency with the greatest measured scattered light intensity is dynamically selected as the reference optical frequency. For example, this corresponds to the case where the present invention is applied to Prior Application Invention 1 or Prior Application Invention 2. When the present invention is applied to a method of selecting and processing multiple reference optical frequencies at each point on a fiber, a selection rule is established such that a threshold is set based on the magnitude of the measurement instrument noise, and only optical frequencies equal to or greater than this threshold are included in the dynamic selection of the reference optical frequency. For example, this corresponds to the case where the present invention is applied to Prior Application Invention 3.

For example, as in Prior Application Invention 1 and Prior Application Invention 2, the functional unit may average the scattered light vectors of the multiple optical frequencies with the scattered light vectors included in a predetermined time window. In this case, the reference optical frequency may be selected with respect to the scattered light vectors included in the time window.

For example, as in Prior Invention 3, the functional unit may average the scattered light vectors of the multiple optical frequencies using two or more optical frequencies among the multiple optical frequencies included in a predetermined time window as a reference. In this case, the reference optical frequency may be selected from the scattered light vectors included in the time window.

In the present invention, when selecting the reference optical frequency, an optical frequency with a greater scattered light intensity is selected, so that optical frequencies with a smaller scattered light intensity can be excluded from the selection of the reference optical frequency. As a result, the present invention can reduce vibration waveform distortion caused by noise-induced fluctuations in the phase rotation angle used to average the frequency-multiplexed scattered light signal.

1 is a configuration example of a measurement system. 3 is an explanatory diagram of an optical pulse output from an optical modulator. FIG. 2 shows an example of the configuration of a signal processing device. 4 shows an example of processing by the signal processing device. 4 shows an example of processing by the signal processing device. 4 shows an example of processing by the signal processing device. 4 shows an example of processing by the signal processing device. 4 shows an example of processing by the signal processing device. 4 shows an example of processing by the signal processing device.

Below, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present disclosure is not limited to the embodiment shown below. These implementation examples are merely illustrative, and the present disclosure can be implemented in various forms with various modifications and improvements based on the knowledge of those skilled in the art. Note that components with the same reference numerals in this specification and drawings are mutually identical.

(System Configuration)
1 shows an example of the configuration of a measurement system according to the present invention. The measurement system according to the present invention comprises a measurement device 31 and a signal processing device 17. The measurement device 31 is a phase OTDR that measures the phase of scattered light from each point of the optical fiber 6 to be measured. The phase OTDR of this embodiment uses frequency multiplexing to measure phases at a plurality of different optical frequencies. In this embodiment, the phase OTDR outputs an in-phase component and a quadrature component, which are used by the signal processing device 17 to calculate a scattered light vector signal representing the phase, and the scattered light vector signal is used to calculate a vibration waveform. Specifically, the measurement system comprises the following configuration.

A continuous light with a single wavelength and an optical frequency of _f0 is emitted from a CW light source 1, and is branched into a reference light and a probe light by a coupler 2. The probe light is shaped by an optical modulator 3 into an optical pulse 4 shown in FIG. 2. The optical pulse 4 is configured by arranging pulses with an optical frequency of f _i =f ₀ +Δf _i (i is an integer) and a pulse width set to a value P corresponding to the spatial resolution of the measurement in the longitudinal direction of the optical fiber, for i=1, 2, ..., N (N is an integer representing frequency multiplexing). f _i is selected so that the intensity of the scattered light at each time and each point is sufficiently separated to be considered uncorrelated between different i. The pulse width P corresponds to the spatial resolution.

There is no specific specification for the type of optical modulator 3 as long as it can generate an optical pulse 4, and there may be more than one. For example, an SSB (Single Side Band amplitude modulation) modulator or a frequency-tunable AO (Acousto-Optics) modulator may be used, and intensity modulation may be performed using a semiconductor optical amplifier (SOA) or the like to increase the extinction ratio in pulsing.

The pulsed probe light 4 is input to the optical fiber 6 under test via the circulator 5. Light scattered or reflected at each point along the length of the optical fiber 6 under test returns to the circulator 5 as backscattered light and is input to one input terminal of the optical 90-degree hybrid 7. The reference light branched by the coupler 2 is input to the other input terminal of the optical 90-degree hybrid 7.

The internal configuration of the optical 90-degree hybrid may be any as long as it has the function of an optical 90-degree hybrid. Two of the four outputs of the optical 90-degree hybrid are detected by a balance detector 13 to obtain an electrical signal 15 that is an analog in-phase component I ^analog . The remaining two outputs of the optical 90-degree hybrid are detected by a balance detector 14 to obtain an electrical signal 16 that is an analog quadrature component Q ^analog . The electrical signals 15 and 16 are sent to a signal processing device 17 that includes an AD conversion function element 17a and an AD conversion function element 17b that can sample the frequency band of the signal without aliasing.

FIG. 3 shows an example of the configuration of the signal processing device 17. The signal processing device 17 is a device that calculates a vibration waveform using measurement data of scattered light that is generated when a frequency-multiplexed optical pulse is reflected or scattered by an optical fiber. Specifically, the signal processing device 17 includes AD conversion function elements 17a and 17b, and signal processing units 17c, 17d, 17e, 17f, and 17g.

In the signal processing device 17, the signal processing unit 17c separates the signals I ^digital of the in-phase component and Q ^digital of the quadrature component output from the AD conversion function element 17a and the AD conversion function element 17b into signals I _i and Q _i due to scattered light due to pulses of each optical frequency f _i (i=1, 2, ..., N) constituting the optical pulse 4. There are no specific restrictions on the method of signal processing for separation, but for example, it is possible to pass I ^digital and Q ^digital through a digital bandpass filter with a center frequency of Δf _i and a passband of 2/P and then compensate for the phase delay. If the filter characteristics of the digital bandpass filter are set to specifications that allow streaming processing of measurement data, taking into account the memory size of the computer used, the measurement data will be streamed up to the signal processing unit 17c.

Signal processing units 17d, 17e, 17f, and 17g continue to perform phase calculations. Signal processing unit 17d is a functional unit that executes step S01, signal processing unit 17e is a functional unit that executes step S02, and signal processing units 17f and 17g are functional units that execute step S03.
Specifically, the signal processing unit 17d is a functional unit that selects a reference optical frequency from among the multiple optical frequencies that are frequency-multiplexed, and calculates the phase rotation angles of the scattered light vectors of other optical frequencies based on the reference optical frequency among the scattered light vectors obtained from the measurement data of the multiple optical frequencies.
The signal processing unit 17e is a functional unit that, after rotating the phase rotation angle, calculates an optical frequency average vector by averaging the scattered light vectors of the multiple optical frequencies, and calculates the angular change of the optical frequency average vector.
The signal processing units 17f and 17g are functional units that use the angle change to calculate the phase difference between two points separated by the gauge length, perform phase unwrapping processing, and calculate a vibration waveform.

In the present disclosure, in order to exclude optical frequencies with low scattered light intensity from the selection of the reference optical frequency, optical frequencies with higher scattered light intensity are tracked over time and dynamically selected in the selection of the reference optical frequency. By using the present invention, it is possible to reduce vibration waveform distortion caused by the influence of angular fluctuations in the phase rotation angle resulting from measurement instrument noise, by excluding optical frequencies with low scattered light intensity from the selection of the reference optical frequency while maintaining the characteristic obtained by updating the phase rotation angle that large vibrations, etc. can be measured with high sensitivity, and it becomes possible to measure the vibration waveform more accurately.

Hereinafter, the pulse incidence interval, i.e., the vibration sampling rate, is defined as t, and the signal processing device 17 acquires measurement data at time kt using an integer k. Also, the distance from the incidence end of the measured optical fiber 6 at which the probe light 4 is incident is defined as z. That is, signals _Ii (kt,z) and _Qi (kt,z) are input to the signal processing unit 17d.

(First embodiment)
In this embodiment, an example of applying the present invention to the prior invention 1 will be described. In the prior invention 1, the processes from step S02 onwards are performed while continuously updating the optimal value of the phase rotation angle. An embodiment example of the case where the present invention is applied to the prior invention 1 will be described.

In the prior invention 1, the signal processing unit 17d calculates the phase rotation angle α _i (z) of each optical frequency f _i from the scattered light vector r _i (kt, z) in the same manner as in the prior art 1. However, unlike the prior art 1, the signal processing unit 17d in the prior invention 1 divides the measurement data into blocks of the number of time points, and calculates and updates the phase rotation angle in block units of the measurement time Mt.

That is, if k=0 is the first point of the measurement data and the calculation of the phase rotation angle is started from the first point, for example, the measurement data of the scattered light vector r ₁ (kt,z) where k is from 0 to M−1 is the first block, and where k is from M to 2M−1 is the second block. K is used as a number symbol to distinguish the blocks. For example, in the above example, k from 0 to M−1 is the block where K=1, and k from M to 2M−1 is the block where K=2. The Kth block is the measurement data where k is from M(K−1) to MK−1. Since the phase rotation angle is calculated for each block, the value changes according to the block K, and the phase rotation angle is expressed as α _i (z,K). The calculation method of the phase rotation angle α _i (z,K) for each block is the same as step S01 in the conventional technique 1 in the present invention, but the present invention adds a process of selecting a reference light frequency.

The main points regarding the signal processing of the measurement data of the Kth block are explained below. A flowchart of the signal processing unit 17d is shown in Figure 4. When the measurement starts, the measurement data obtained by the signal processing unit 17c is sent to the signal processing unit 17d.

Step S101-1: Calculate scattered light vector r _i (kt,z) from signal I _i (kt,z) and signal Q _i (kt,z) at time kt. For example, with the imaginary unit j, it can be calculated as follows.
r _i (kt, z)=I _i (kt, z)+j・Q _i (kt, z) (111)
The calculated scattered light vector r _i (kt,z) data is sequentially streamed to the signal processing unit 17e.

Step S101-2: Data of the scattered light vector r _i in one time block is stored, and the sum of the vector length |r _i (kt,z)| of the scattered light vector r _i (kt,z) in the block is calculated. That is, in the case of the Kth block, the sum of |r _i (kt,z)| is calculated for k from M(K-1) to MK-1. The optical frequency with the largest calculated sum is selected as the reference optical frequency. Here, the reference optical frequency is set to f _ref . The reference optical frequency f _ref depends on the distance z from the input end and the time block K. When it is necessary to take these dependencies into consideration, it will be described below as f _ref (z,K), f _ref (z), or f _ref (K).

Step S101-3: For data in the same block, calculate the angle θ _ref (kt, z) of the scattered light vector r _ref (kt, z) of the reference light frequency at each optical fiber point at time kt. For example, the following is calculated.
θ _ref (kt, z)=arg[r _ref (kt, z)] (113)

Step S101-4: The scattered light vector r _i (kt,z) of each optical frequency at each optical fiber point at time kt in the time block is rotated by an angle −θ _ref (kt,z) to obtain r _{i _rot} (kt,z).
r _{i_rot} (kt, z)=exp[-jθ _ref (kt, z)]×r _i (kt, z)
(114)
Here, j is the imaginary unit.

Step S101-5: Calculate a sum vector within the time block of the calculated r _{i _rot} (kt,z). For example, the following is calculated.
r _{i_avet} (z, K)=Σr _{i_rot} (kt, z) (115)
Step S101-6: Using the obtained r _{i — avet} (z, K), calculate the phase rotation angle α _i (z, K) of each optical frequency using the measurement data of the Kth block.
α _i (z, K)=-arg[r _{i_avet} (z, K)] (116)

When step S101-6 is completed, α _i (z, K) is passed to the signal processing unit 17e.
Signal processing unit 17e rotates and averages the scattered light vector r _i of each optical frequency f _i using the phase rotation angle α _i calculated by signal processing unit 17d to calculate an optical frequency average vector. The phase rotation angle α _i used when calculating the frequency average of the scattered light vectors belonging to a certain block K is the phase rotation angle α _i calculated by signal processing unit 17d in the previous block K-1. That is, detailed processing by signal processing unit 17e for the Kth block is as follows:

Step S102-1: The signal processing unit 17e rotates the scattered light vector r _i (kt, z) of each optical frequency at time kt by the phase rotation angle α _i (z, K-1) calculated in the K-1th block, and calculates the rotated vector R _i (kt, z). The calculation method can be exemplified as follows.
R _i (kt, z) = exp (j・α _i (z, K-1))・r _i (kt, z)
(121)

Step S102-2: The signal processing unit 17e calculates a vector obtained by averaging the frequency of the rotated vector R _i (kt,z). Note that since the final result is the same whether it is vector averaging or vector synthesis (simple vector sum), in the actual calculation procedure, the vectors are synthesized to obtain the optical frequency average vector R _avef (kt,z). The calculation method is as follows.

Step S102-3: Calculate the angle θ _avef (kt,z) of the optical frequency average vector R _avef (kt,z) and transfer it to the downstream signal processing unit 17f. The calculation method is as follows.
θ _avef (kt, z)=arg[R _avef (kt, z)] (123)

In the first block where K=1, since there is no data from the previous blocks, the phase rotation angle is not yet obtained, and the procedure of the signal processing unit 17e cannot be performed. Therefore, the measurement time of the first Mt is treated as a preliminary measurement, and no processing after the signal processing unit 17e is performed. Alternatively, it is possible to store only r _i (kt, z) for the measurement time of the first Mt separately in a computer, and calculate the vibration waveform using prior art 1.

In the signal processing unit 17f in the latter stage, the phase difference between two points separated by the gauge length is calculated using the change in the angle θ _avef (kt, z) of the obtained optical frequency average vector, and a vibration waveform is calculated by performing phase unwrapping processing, etc. A specific calculation method that can be used is the same as the calculation method of the vibration waveform from the phase in a general phase OTDR.

However, if the reference optical frequencies used to calculate the angle θ _avef (kt,z) are different between adjacent blocks, the calculated value of the vibration change from the last time of the younger block to the first time of the next block will include a difference in the phase offset of the reference optical frequency, and the correct value cannot be reflected. For example, the optical frequency serving as the reference for the phase rotation angle used by signal processing unit 17e of block K is f _ref (K-1) selected in step S101-2 using the data of block K-1, and the optical frequency serving as the reference for the phase rotation angle used by signal processing unit 17e of block K+1 is f _ref (K) selected in step S101-2 using the data of block K. In other words, the phase offset of the angle θ _avef (kt, z) of block K is based on the reference optical frequency f _ref (K-1), and the phase offset of the angle θ _avef (kt, z) of block K+1 is based on the reference optical frequency f _ref (K). Therefore, when the reference optical frequencies f _ref (K-1) and f _ref (K) are different, the change over time in the angle θ _avef from MK-1 to MK at time kt includes the effect of the difference in phase offset between the different optical frequencies.

In order to avoid this effect, a method is conceivable in which the time length of each block is set to (M+1)t, adjacent blocks share a data point, and the data point is corrected. For example, block K is set to include (M+1)t pieces of data with k ranging from M(K-1) to MK, and block K+1 is set to include (M+1)t pieces of data with k ranging from MK to (M+1)K. By setting such block divisions, the data point of kt=MKt is included in both the Kth block and the K+1th block. Therefore, if the signal processing unit 17d and the signal processing unit 17e are performed in the same manner as in the signal processing unit 17e except for setting the block length, two types of data are obtained: the last data of block K and the first data of block K+1, with the angle θ _avef (MKt,z) at time kt=MKt. The former is set to θ _avef (MKt,z,K) and the latter is set to θ _avef (MKt,z,K+1) to distinguish them. θ _avef (MKt,z,K+1) - θ _avef (MKt,z,K) is a value that reflects the difference in offset of the reference optical frequency. Therefore, if all angles θ _avef (kt,z) included in the K+1th block are corrected to θ _avef (kt,z) - [θ _avef (MKt,z,K+1) - θ _avef (MKt,z,K)], the influence of the difference in phase offset due to the difference in the reference optical frequency can be reduced.

In this embodiment, the processing of the K+1th block has been described, but by carrying out similar processing starting from the smallest block, the above-mentioned effects can be reduced at all times. Incidentally, this method of setting the time length of each block to (M+1)t, sharing a data point with adjacent blocks, and performing correction at that data point also has the effect of compensating for the shortcoming of Prior Invention 1, namely, phase discontinuity caused by discontinuity in the phase rotation angle, even if the reference optical frequency is the same in adjacent blocks.

In this embodiment, the signal processing unit 17e transfers the value of the angle θ _avef (kt, z) itself to the signal processing unit 17f. However, the signal processing unit 17e may transfer the time difference of the angle expressed by the following equation, and the time difference of the angle may be processed by the signal processing unit 17f.
θ _avef ((k+1)t,z)−θ _avef (kt,z)

In that case, in order to reduce the influence of the difference in phase offset between different optical frequencies, the signal processing unit 17f calculates the angle change of the optical frequency average vector due to the change in time from (MK-1)t to MKt using the following equation.
θ _avef (MKt, z, K) - θ _avef ((MK-1)t, z, K)

In addition, while the procedure of Prior Invention 1 allows for streaming processing of data, the addition of procedure S101-2 according to the present invention creates a new need to hold all scattered light data within one block. If it is desired to maintain the streaming processing characteristics of Prior Invention 1, it is possible to devise a method such as limiting the data points used to add the vector lengths in procedure S101-2 to the first few points of each block that can be held by the memory of the computer.

The feature of the present invention is step S101-2. In the prior invention 1, the signal processing unit 17d divides the measurement data into blocks of M time points, and calculates and updates the phase rotation angle in units of blocks of measurement time Mt, thereby dealing with temporal changes in the optimal value due to changes in optical characteristics such as the oscillation frequency of the laser, changes in temperature of the measured optical fiber itself, and application of large dynamic strain to the measured optical fiber. By making the measurement time Mt per block small relative to the time scale of changes in optical characteristics such as the oscillation frequency of the laser, changes in temperature of the measured optical fiber itself, and application of large dynamic strain to the measured optical fiber, it is possible to update the phase rotation angle with sufficient precision, and there is no problem in using the phase rotation angle calculated by the signal processing unit 17d in block K-1 as the phase rotation angle used when the signal processing unit 17e calculates the frequency average of the scattered light vectors belonging to block K.

According to the prior invention 1, if M is set to about 100, sufficient accuracy of the phase rotation angle can be guaranteed, and if the pulse transmission period t is 1 ms, Mt will be 100 ms or less, which is sufficient to handle cases where the time scale of the optimal phase rotation angle change is 1 s or more. The order of this setting value for Mt is about the same as the order of the time width at which the magnitude relationship of the scattered light intensity of each optical frequency can be considered constant within each block. For this reason, by applying the present invention in step S101-2, selecting a reference optical frequency from the K-1th block, calculating the phase rotation angle, and using it to process the frequency average of the Kth block, it is possible to reduce vibration waveform distortion caused by the influence of angular fluctuations in the phase rotation angle derived from the measurement device noise.

However, even if the length of the time block length used to select the reference optical frequency and the length of the time block length used to update the optimal phase rotation angle used for frequency averaging are not the same, it is possible to implement the present invention in the same manner as this embodiment, as long as the spirit of the present invention is not changed.

Second Embodiment
In this embodiment, an example of applying the present invention to Prior Invention 2 will be described. In Prior Invention 2, a time window of window length Mt (M is a natural number corresponding to the number of data points) used in signal processing is set in advance, and then the time window is moved while calculating a phase rotation angle using the measurement data included in the time window, and the phase rotation angle is used to average the scattered light vectors of the frequency components at the center of the time window or at a time located a preset number of measurement data points behind. The window length Mt may be set in the same manner as the block length in the first embodiment.

The method of applying the present invention to Prior Application Invention 2 is generally the same as the method of applying the present invention to Prior Application Invention 1. However, Prior Application Invention 2 continuously updates the phase rotation angle at each time, and by applying the present invention, the reference optical frequency may also be changed at any time. Therefore, it is necessary to continuously correct the difference in phase offset due to the difference in reference optical frequency, which is different from the case of applying to Prior Application Invention 1.

In the signal processing unit 17d, similarly to the prior invention 2, the scattered light vector r _i (kt,z) is calculated and continuously streamed to the signal processing unit 17e. In addition, the scattered light vector from time (k-W)t to (k+W)t is used to calculate the phase rotation angle α _i (kt,z) of each optical frequency f _i used for frequency averaging. In other words, the window length Mt satisfies Mt=(2W+1)t. The calculated phase rotation angle α _i (kt,z) is continuously streamed to the signal processing unit 17e.

In the present invention, a process for dynamically and continuously updating the selection of the reference optical frequency is added. Also, in order to continuously correct the difference in phase offset caused by dynamically updating the reference optical frequency, the phase rotation angle α _i (kt, z) is used not only at time kt but also for the frequency average at time (k+1)t in the signal processor 17e, and the time change of the phase is transmitted to the signal processor 17f.

First, a specific example of the calculation procedure of the signal processor 17d will be described below. It is assumed that measurement starts from k = 0. A flowchart of the signal processor 17d is shown in FIG.
Step S201-1: Calculate scattered light vector r _i (kt,z) from signal I _i (kt,z) and signal Q _i (kt,z) at time kt. For example, with the imaginary unit j, it can be calculated as follows.
r _i (kt, z)=I _i (kt, z)+j・Q _i (kt, z) (211)
The calculated scattered light vector data is sequentially streamed to the signal processing unit 17e.

Step S201-2: The scattered light vector lengths |r _i (kt, z)| from time (k-2W)t to time kt are summed up, and the optical frequency with the largest sum is set as the reference optical frequency f _ref . The reference optical frequency f _ref depends on the distance z from the incident end and the time kt. When it is necessary to take these dependencies into consideration, it will be written as f _ref (kt, z), f _ref (z), or f _ref (kt) below.

Step S201-3: Calculate the angle θ _ref (kt,z) of the scattered light vector r _ref (kt,z) of the reference light frequency f _ref (kt,z) at each optical fiber point at time kt. For example, it can be calculated as follows.
θ _ref (kt, z)=arg[r _ref (kt, z)] (213)

Step S201-4: The scattered light vector r _i (kt,z) of each optical frequency at each optical fiber point at time kt is rotated by an angle −θ _ref (kt,z) to obtain r _{i _rot} (kt,z).
r _{i_rot} (kt, z)=exp[-jθ _ref (kt, z)]×r _i (kt, z)
(214)

ただし、ｋｔ＝０からｋｔ＝２Ｗｔまで測定することで窓長（２Ｗ＋１）ｔ以上のデータが得られるようになるため、ｒ_{ｉ＿ａｖｅｔ}（ｋｔ，ｚ）のｋｔがＷｔ未満の場合には計算ができない。そのため、時刻ｋｔがＷｔ未満の測定データは予備測定にするか、ｋ＝Ｗの時間平均ベクトルにそろえることができる。後者の場合、ｋ＜Ｗについて、次式とする。
ｒ_{ｉ＿ａｖｅｔ}（ｋｔ，ｚ）＝ｒ_{ｉ＿ａｖｅｔ}（Ｗｔ，ｚ）　　　（２１５－２） Step S201-5: r _{i _rot} (kt, z) from time (k-2W)t to time kt is averaged to calculate the time average vector r _{i _avet} ((k-W)t, z). The argument time of the time average vector is the center time of the time window. In addition, since the final result is the same whether averaging or simple addition is performed, the time average vector is calculated as an addition as follows.

However, since data with a window length of (2W+1)t or more can be obtained by measuring from kt=0 to kt=2Wt, calculation is not possible if kt of _{ri_avet} (kt,z) is less than Wt. Therefore, measurement data with time kt less than Wt can be used as a preliminary measurement, or can be aligned to a time average vector of k=W. In the latter case, for k<W, the following equation is used.
r _{i_avet} (kt, z)=r _{i_avet} (Wt, z) (215-2)

Step S201-6: Calculate the phase rotation angle α _i ((k−W)t, z) using the angle of the calculated time-averaged vector r _{i — avet} ((k−W)t, z).
α _i ((k-W)t, z)=-arg[r _{i_avet} ((k-W)t, z)]
(216)
The calculated phase rotation angles α _i ((k−W)t, z) are sequentially streamed to the signal processing unit 17 e.

In the signal processing unit 17e, the scattered light vector _ri of each optical frequency _fi is rotated and averaged using the phase rotation angle _αi calculated in the signal processing unit 17d to calculate the optical frequency average vector. In the prior invention 2, the angle of the optical frequency average vector is transmitted to the signal processing unit 17f as is, but in the present invention, as a measure against the dynamic replacement of the reference optical frequency, the change in angle (phase) at adjacent times over time is calculated and then transmitted to the signal processing unit 17f. A specific example of the calculation procedure is described below. The calculation procedure is shown in FIG. 6.

Step S202-1: The signal processing unit 17e rotates the scattered light vectors r _i (kt,z) and r _i ((k+1)t,z) streamed in step S201-1 by the phase rotation angle α _i (kt,z) streamed in step S201-6, and calculates the rotated vectors R _i (kt,z) and R _i ((k+1)t,z). The calculation method can be exemplified as follows.
R _i (kt, z)=exp(j・α _i (kt, z))・r _i (kt, z)
(221-1)
R _i ((k+1)t, z)=exp(j・α _i (kt, z))・r _i ((k+1)t, z)
(221-2)

Note that a feature of this method is that the phase rotation angle calculated using the same reference optical frequency is used to calculate the optical frequency average vectors R _i ((k+1)t,z) and R _i (kt,z). However, it is not always necessary to use the same phase rotation angle. Therefore, a phase rotation angle for rotating the scattered light vector r _i ((k+1)t,z) may be calculated separately using the measurement data from time (k-W+1)t to time (k+W+1)t based on the reference optical frequency f _ref (kt,z), and used in place of α _i (kt,z) in (221-2). In this case, the method of calculating the phase _{rotation angle} for rotating r _i ((k+1)t,z) using the measurement data from time (k-W+1)t to time (k+W+1)t based on f ref (kt,z) is to keep the reference optical frequency at f _ref (kt,z) in steps S201-2 to S201-6, and replace the measurement data to be calculated with the measurement data from time (k-W+1)t to time (k+W+1)t, instead of the measurement data from time (k-W)t to time (k+W)t.

Step S202-2: The signal processing unit 17e calculates a vector obtained by frequency averaging the rotated vector R _i (kt,z). Note that since the final result is the same whether it is vector averaging or vector synthesis (simple vector sum), in the actual calculation procedure, vector synthesis is performed to obtain the optical frequency average vector R _avef (kt,z). The calculation method can be exemplified as follows.

Step S202-3: The signal processing unit 17e calculates the angle θ _avef (kt,z) of the optical frequency average vector R _avef (kt,z) and the angle θ _avef ((k+1)t,z) of the optical frequency average vector R _avef ((k+1)t,z), and passes the change θ _avef ((k+1)t,z) - θ _avef (kt,z) to the downstream signal processing unit 17f. The angle θ _avef can be calculated as follows:
θ _avef (kt, z)=arg[R _avef (kt, z)] (223-1)
θ _avef ((k+1)t, z)=arg[R _avef ((k+1)t, z)]
(223-2)

Furthermore, signal processing unit 17f uses the change in phase over time θ _avef ((k+1)t,z)-θ _avef (kt,z) obtained by signal processing unit 17e to calculate the spatial difference in phase between two points separated by the gauge length, and performs phase unwrapping processing, etc. to calculate a vibration waveform. As a specific calculation method, a calculation method similar to that of a general DAS-P can be used.

In this embodiment, the length of the time window used to select the reference optical frequency is the same as the length of the time window used to update the optimal phase rotation angle used for frequency averaging. However, even if they are not the same, the present invention can be implemented in the same manner as this embodiment, as long as the gist of the present invention is not changed.

As in the prior invention 2, the phase rotation angle α _i ((k-W)t,z) can be used for frequency averaging at times (k-W+Delay)t and (k-W+Delay+1)t after time (k-W)t, depending on the memory requirements of the calculator, etc. (the variable Delay is an integer reflecting the amount of delay). In other words, a phase rotation angle calculated using measurement data within a time window having a window length of 2W+1 can be used for frequency averaging at a time after the center time of the time window. The greater the time difference amount Delay, the greater the difference between the optimal phase rotation angle α _i ((k-W+Delay)t,z) at time (k-W+Delay)t and the phase rotation angle α _i ((k-W)t,z) to be used, so it is desirable to keep it to a minimum value determined by memory capacity, etc.

Third Embodiment
In this embodiment, an example will be described in which the present invention is applied to the prior invention 3. The roles of the signal processing units 17d to 17g are as follows.

(Signal processing unit 17d)
A specific example of the calculation procedure is described below. It is assumed that the measurement starts from k = 0. The flowchart of the signal processing unit 17d is also shown in FIG.
Step S301-1: Calculate the scattered light vector r _i (kt,z) from the signal I _i (kt,z) and the signal Q _i (kt,z).
r _i (kt, z)=I _i (kt, z)+j・Q _i (kt, z) (311)
The calculated scattered light vector r _i (kt,z) data is continuously streamed to a signal processor 17e.

Step S301-2: Calculate an index L _i (kt,z) which is the sum of vector lengths |r _i (kt,z)| of scattered light vectors from time (k-2W)t to kt. Here, the window length Mt satisfies Mt=(2W+1)t. Let X be a set of optical frequencies f _i for which L _i (kt,z) is greater than the threshold L _T. It should be noted that since X depends on the position of the time window, sets X calculated using time windows centered on different times are not necessarily the same. In the following steps, an element of X is selected as the reference optical frequency, which is a feature of the present invention. The set value of the threshold L _T depends on the required specifications of the measurement, but for example, the magnitude of the noise of the measuring device can be set as the threshold L _T.

Step S301-3: For the scattered light vectors from time (k-2W)t to kt, calculate the angle θ _i (kt,z) of the scattered light vector r _i (kt,z) of each light frequency f _i included in set X. For example, it can be calculated as follows.
θ _i (kt, z)=arg[r _i (kt, z)] (313)

Step S301-4: For the scattered light vectors from time (k-2W)t to kt, rotate the scattered light vector r _i (kt,z) of each optical frequency f _i by a phase rotation angle -θ _x (kt,z) when each optical frequency f _x is the reference optical frequency, to calculate r _{i _rot,x} (kt,z). For example, it can be calculated as follows.
r _{i_rot, x} (kt, z) = exp [-jθ _x (kt, z)] × r _i (kt, z)
(314)
In the above formula, for the combination of i and x (i, x), only optical frequencies belonging to set X are calculated for x, and all possible cases are calculated for i.

Step S301-5: For data from (k-2W)t to _kt , calculate the time-averaged vector r _{i _ avet,x} ((k-W)t,z) by adding r _{i _ rot,x} when the reference optical frequency is f x. The calculation formula can be exemplified as follows.

However, since data with a window length of (2W+1)t or more can be obtained by measuring from kt=0 to kt=2Wt, calculation is not possible if kt of _{ri_avet,x} (kt,z) is less than Wt. Therefore, measurement data with a time kt less than Wt can be used as a preliminary measurement, or can be aligned to a time average vector of k=W. In the latter case, for k<W, the following equation is used.
r _{i_avet, x} (kt, z) = r _{i_avet, x} (Wt, z) (315-2)
In the above formula, for the combination of i and x (i, x), only optical frequencies belonging to set X are calculated for x, and all possible cases are calculated for i.

Step S301-6: Calculate the phase rotation angle α _{i, x ((k−W)t,z) of each optical frequency f i when each optical frequency f x is the reference optical frequency from the time-averaged vector r i _avet, x (} (k−W)t,z), and stream it to the signal processing unit 17 e. The calculation formula can be exemplified as follows.
α _i,x ((kW)t,z)=-arg[r _{i_avet,x} ((kW)t,z)]
(316)
In the above formula, for the combination of i and x (i, x), only optical frequencies belonging to set X are calculated for x, and all possible cases are calculated for i.

FIG. 8 shows a flowchart of the process executed by the signal processors 17e, 17f, and 17g.
(Signal processing unit 17e)
The signal processing unit 17e rotates and averages the scattered light vector r _i of each optical frequency f _i using the phase rotation angle α _i,x calculated by the signal processing unit 17d to calculate an optical frequency average vector. The calculation is performed for the reference optical frequency f _x selected according to the procedure of the present invention. Specifically, the procedure includes the following steps.
Step S302-1: The scattered light vectors r _i (kt,z) and r _i ((k+1)t,z) streamed in step S301-1 are rotated by the phase rotation angle α _i,x (kt,z) streamed in step S301-6 to calculate the rotated vector R _i,x (kt,z) when the reference light frequency is f _x . The calculation method can be exemplified as follows.
R _i,x (kt, z)=exp(j・α _i,x (kt, z))・r _i (kt, z)
(321-1)
R _i,x ((k+1)t,z)=exp(j・α _i,x (kt,z))・_ri ((k+1)t,z)
(321-2)

Note that, although it is a feature that the phase rotation angle calculated using the same reference optical frequency f _x is used for the calculation of R _i,x ((k+1)t,z) and R _i,x (kt,z), it is not necessarily necessary to use the same phase rotation angle, so a phase rotation angle for rotating r _{i ,x} ((k+1)t,z) may be calculated separately using the measurement data from time (k-W+1)t to time (k+W+1)t based on f _x and used in place of α _i,x (kt,z) in equation (321-2). In that case, the method for calculating the phase rotation angle for rotating r _i,x ((k+1)t,z) using the measurement data from time (k-W+1)t to time (k+W+1)t based on f x may be the same as the method described in Prior Application 3.

Step S302-2: Calculate a vector obtained by averaging the frequency of the rotated vector R _i,x (kt,z). Note that since the final result is the same whether it is vector averaging or vector synthesis (simple vector sum), in the actual calculation procedure, vector synthesis is performed to obtain the optical frequency average vector R _avef,x (kt,z). The calculation method can be exemplified as follows.

Step S302-3: Calculate the angle θ avef _{,x (kt,z) of the optical frequency average vector R avef, x (kt,z) and the angle θ avef,x} ((k+1)t,z) of the _optical frequency average vector R _{avef,x (} (k+1)t,z) when the reference optical frequency is f x, and transfer the change over time θ _avef,x ((k+1)t,z) - θ _avef,x (kt,z) to the signal processing unit 17f. The following is an example of a method for calculating the angle θ _avef .
θ _avef,x (kt,z)=arg[R _avef,x (kt,z)]
(323-1)
θ _avef,x ((k+1)t,z)=arg[R _avef,x ((k+1)t,z)]
(323-2)
Each optical frequency belonging to the set X is calculated as a reference optical frequency _fx .

(Signal processing unit 17f)
Step S303: The signal processing unit 17f calculates the difference in phase between _{two points} separated by a gauge length using the time change θ _{avef ,x} ((k+1)t,z) - θ _avef,x (kt,z) of the angle θ avef,x of the reference optical frequency f x. If the magnitude exceeds π, an integer multiple of 2π is added to correct the magnitude to be equal to or less than π. This operation corresponds to phase unwrapping. The finally obtained value is the time change ψ _x ((k+1)t,z) - ψ _x (kt,z) of the vibration waveform ψ _x (kt,z) when the reference optical frequency is f _x . Note that the specific calculation method can be the same as the calculation method of the vibration waveform from the phase in a general phase OTDR. Calculation is performed with each optical frequency belonging to the set X as the reference optical frequency f _x . The change in the vibration waveform ψ _x (kt,z) over time, ψ _x ((k+1)t,z) - ψ _x (kt,z), is streamed to the signal processing unit 17g.

(Signal processing unit 17g)
Step S304: The signal processing unit 17g averages the change over time ψ _x ((k+1)t,z) - ψ _x (kt,z) for x to calculate the change over time ψ ((k+1)t,z) - ψ (kt,z) of the final vibration waveform ψ (kt,z). The change over time is cumulatively added to obtain the final vibration waveform ψ (kt,z). The offset value at the start time of the cumulative addition can be set in the same way as the offset setting for vibration waveforms in a general phase OTDR.

In the present invention, by selecting the optical frequency that plays the role of the reference optical frequency in step S301-2 based on the scattered light intensity, it is possible to reduce vibration waveform distortion due to the influence of angle fluctuation caused by noise of the phase rotation angle α _i . Also, in the case of this embodiment, the combination of optical frequencies selected as the reference optical frequency can change dynamically, so that when the reference optical frequency is different, the difference in phase offset may have an effect. To address this problem, in this embodiment, the signal processing unit 17f and the signal processing unit 17g perform calculations based on the change in angle θ _avef,x and the vibration waveform ψ _x over time. In order to perform such calculations based on the change over time, in this embodiment, both the angle θ _{avef,x (} kt,z) and the angle θ _avef,x ((k+1)t,z) are calculated using the phase rotation angle α i,x (kt,z). This means that the angle θ _avef,x at a certain time kt is calculated using two phase rotation angles, α _i,x (kt,z) and α _i,x ((k−1)t,z), and this two-time calculation is a characteristic feature.

(Fourth embodiment)
In this embodiment, a calculation method in which the signal processing unit 17d calculates the phase rotation angles α i,x for all possible (i,x) based on the reference optical frequency f _x in step S302-1 of the third embodiment will be described with reference to FIG. ₉ .

Step S401-1: Calculate scattered light vector r _i (kt,z) from signal I _i (kt,z) and signal Q _i (kt,z). For example, with the imaginary unit j, it can be calculated as follows.
r _i (kt, z)=I _i (kt, z)+j・Q _i (kt, z) (411)
The calculated scattered light vector data is sequentially streamed to the signal processing unit 17e.

Step S401-2: Calculate the angle θ _i (kt, z) of the scattered light vector r _i (kt, z) of each optical frequency f _i at each optical fiber point at time kt. For example, it can be calculated as follows.
θ _i (kt, z)=arg[r _i (kt, z)] (412)

Step S401-3: Rotate the scattered light vector r _i (kt, z) of each optical frequency f _i at each optical fiber point at time kt by a phase rotation angle −θ _x (kt, z) when each optical frequency f _x is set as a reference optical frequency to calculate r _{i _rot,x} (kt, z). For example, it can be calculated as follows.
r _{i_rot, x} (kt, z) = exp [-jθ _x (kt, z)] × r _i (kt, z)
(413)
For the combination (i, x), all possible cases are calculated.

Step S401-4: Using the time-average vector r _{i _rot,x} (kt,z) when the reference optical frequency is f _x , calculated using data from time (k-2W-1)t to time (k-1)t, r i _{_rot,} x ((k-2W-1)t,z) at time (k-2W-1)t, and r _{i _rot,x} (kt,z) at time kt, calculate the time-average vector r _{i _avet,x} ((k-W)t,z) when the reference optical frequency is f _x . The calculation formula can be exemplified as follows.
r _{i_avet,x} ((k-W)t,z)=r _{i_avet,x} ((k-W-1)t,z)
-r _{i_rot,x} ((k-2W-1)t,z)+r _{i_rot,x} (kt,z)
(414)

However, since data for a time window of length (2W+1)t or more can be obtained by measuring from k=0 to k=2W, when k=2W, the time average vector is obtained as the sum of r _{i_rot,x} (kt,z) up to k≦2W. The following is an example of the calculation formula.

For data of k<2W, a preliminary measurement is performed, or the time-averaged vector is calculated using the conventional technique 1. When the conventional technique 1 is used, the time-averaged vectors for all times of k<2W can be aligned with the time-averaged vector for k=2W. In other words, the following equation is established.
r _{i_avet, x} (kt, z) = r _{i_avet, x} (Wt, z) (416)
For the combination (i, x), all possible cases are calculated.

Steps S401-5 and S401-6: From the time-averaged vector r _{i_avet,x} ((k−W)t,z), calculate the phase rotation angle α _i,x ((k−W)t,z) of each optical frequency f _i when each optical frequency f _x is the reference optical frequency for all possible (i, x), and stream the result to the signal processing unit 17 e.

Other Embodiments
In the above embodiment, the vibration waveforms that have been subjected to the phase unwrapping process are averaged in the processing by the signal processing units 17f and 17g. However, the order of the specific calculation steps can be changed within the scope that does not impair the essence of the present invention, for example, by carrying out averaging before carrying out the phase unwrapping process and then carrying out the phase unwrapping process.

In this embodiment, the length of the time window used to select the reference optical frequency is the same as the length of the time window used to update the optimal phase rotation angle used for frequency averaging. However, even if they are not the same, the present invention can be implemented in the same manner as this embodiment, as long as the gist of the present invention is not changed.

In addition, this embodiment describes a specific implementation procedure for applying the present invention when prior application invention 3 is implemented based on prior application invention 2. However, prior application invention 3 is not limited to being based on prior application invention 2, and can be widely applied to cases where prior application invention 1 or other methods of averaging frequency multiplexed signals in a vector state are used as a base, and the present invention can be applied to each of these calculation methods.

The signal processing device of the present disclosure can also be realized by a computer and a program, and the program can be recorded on a recording medium or provided via a network. The program of the present disclosure is a program for realizing a computer as each functional unit of the signal processing device of the present disclosure, and is a program for causing a computer to execute each step of the method executed by the signal processing device of the present disclosure.

1: CW light source 2: Coupler 3: Optical modulator 4: Optical pulse 5: Circulator 6: Optical fiber to be measured 7: 90 degree optical hybrid 13, 14: Balance detector 15, 16: Electrical signal 17: Signal processing device 17a, 17b: AD conversion function element 17c, 17d, 17e, 17f, 17g: Signal processing unit 31: Measuring device

Claims (4)

a signal processing device that calculates a vibration waveform using measurement data of scattered light that is generated when the frequency-multiplexed optical pulse is reflected or scattered by an optical fiber;
the signal processing device comprises a functional unit which selects a reference optical frequency from among a plurality of frequency-multiplexed optical frequencies, rotates scattered light vectors of other optical frequencies among scattered light vectors obtained from measurement data of the plurality of optical frequencies using the reference optical frequency as a reference, and then averages the scattered light vectors of the plurality of optical frequencies;
The functional unit dynamically selects a reference light frequency by tracking a light frequency at which the intensity of the scattered light is greater over time in selecting the reference light frequency.
Device. The functional unit includes:
using the scattered light vector of the selected reference light frequency, averaging the scattered light vectors of the plurality of light frequencies at different times;
signal processing based on the change between the optical frequency average vectors at different times obtained by the averaging;
2. The apparatus of claim 1. The functional unit averages the scattered light vectors of the plurality of optical frequencies with scattered light vectors included in a predetermined time window,
the selection of the reference light frequency is performed on a scattered light vector included within the time window;
2. The apparatus of claim 1. the functional unit averages scattered light vectors of the plurality of optical frequencies using two or more optical frequencies among the plurality of optical frequencies included in a predetermined time window as a reference; and
the selection of the reference light frequency is performed on a scattered light vector included within the time window;
2. The apparatus of claim 1.

Images