Unknown vibration event waveform detection method based on phi-OTDR
By combining the methods of phase demodulation, differential winding and correlation coefficient judgment, the problem that the prior art cannot detect unknown vibration events is solved, and the rapid and accurate detection and recovery of unknown vibration event waveforms are achieved.
Patent Information
- Application Number
- CN202411342088.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2025-02-14
AI Technical Summary
The existing phase demodulation method is only suitable for known types of external disturbance signals and cannot effectively detect waveforms of unknown vibration events.
The unknown vibration event waveform detection method based on φ-OTDR is adopted, and the combination of phase demodulation, differential winding and correlation coefficient judgment is used to automatically judge the type of disturbed signal and select an appropriate demodulation method.
It realizes rapid and accurate detection of unknown vibration event waveforms, improves the system's waveform recovery efficiency and accuracy of unknown vibration events, and ensures signal demodulation performance in complex environments.
Smart Images
Figure CN119437389A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of optical fiber sensing, and in particular to a method for detecting an unknown vibration event waveform based on φ-OTDR. Background Art
[0002] Phase-sensitive optical time-domain reflectometer (φ-OTDR) is a typical distributed sensing technology with broad application prospects due to its high sensitivity, accurate positioning, and wideband response. φ-OTDR realizes quantitative detection of external disturbances by extracting the phase difference of the Rayleigh backscattering signal (RBS) at two different positions on the optical fiber. Since the external disturbance is linearly proportional to the phase difference, the magnitude and frequency of the external disturbance can be restored by phase demodulation. Currently, the commonly used phase demodulation methods include 3×3 coupler demodulation method, orthogonal demodulation method and Hilbert transform method. In these commonly used phase demodulation methods, the inverse tangent operation is used to extract the phase, and the extracted phase is wrapped in the interval [-π, π]. Therefore, in order to obtain the correct phase signal, a phase unwrapping method is required.
[0003] Since the previous phase demodulation methods all perform phase demodulation for known types of external disturbance signals, the strain range of the known types of external disturbance signals can be determined empirically based on information such as the amplitude range and frequency range of the known types of external disturbance signals, and different phase unwrapping methods are used based on the determined strain range of the external disturbance signals, such as using the traditional phase unwrapping method for small dynamic strains and using the differential unwrapping method for large dynamic strains. When using differential unwrapping for large dynamic strains, first perform traditional phase unwrapping on the disturbance signal, and determine whether the recovered phase waveform is consistent with the known type of external disturbance signal: if consistent, the traditional phase unwrapping is effective; otherwise, it is necessary to perform differential unwrapping on the disturbance signal again, and then determine whether the recovered phase waveform is consistent with the known type of external disturbance signal: if consistent, no secondary differential is required; otherwise, secondary differential is required, and so on, performing n differentials until the recovered phase waveform is consistent with the known type of external disturbance signal. It can be seen from this that the existing phase demodulation method needs to determine the strain range based on the known type of external disturbance signal to determine whether it belongs to small or large dynamic strain, and when performing differential dewinding, it is also necessary to judge whether the differential has stopped based on the known type of external disturbance signal. Therefore, it is only applicable to known types of external disturbance signals. For unknown types of external disturbance signals, since their amplitude range and frequency range and other information cannot be known, the existing phase demodulation method is not applicable. Summary of the invention
[0004] The present invention aims to solve the problem that the existing phase demodulation method is only applicable to known types of external disturbance signals and therefore cannot detect the waveform of unknown vibration events, and provides a method for detecting the waveform of unknown vibration events based on φ-OTDR.
[0005] To solve the above problems, the present invention is achieved through the following technical solutions:
[0006] A method for detecting an unknown vibration event waveform based on φ-OTDR comprises the following steps:
[0007] Step 1: Use φ-OTDR to collect the disturbance signal caused by the unknown vibration event to be measured, and perform phase demodulation on the disturbance signal to obtain the initial phase φ;
[0008] Step 2, perform conventional phase unwrapping on the initial phase φ to obtain an initial demodulated phase waveform φ0, perform differential unwrapping and summing on the initial phase φ to obtain a first demodulated phase waveform φ1;
[0009] Step 3: Calculate the initial correlation coefficient C0 between the initial demodulation phase waveform φ0 and the first demodulation phase waveform φ1:
[0010]
[0011] in, represents the covariance of the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1, represents the standard deviation of the initial demodulated phase waveform φ0, represents the standard deviation of the first demodulated phase waveform φ1;
[0012] Step 4: Compare the initial correlation coefficient C0 with the set decision threshold δ:
[0013] If the initial correlation coefficient C0 is greater than or equal to the decision threshold δ, the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1 are the time domain waveforms of the unknown vibration event to be measured;
[0014] Otherwise, set n=1 and go to step 5;
[0015] Step 5: Demodulate the nth phase waveform φ n Perform n+1 times of differential detangling and summing to obtain the n+1th demodulated phase waveform φ n+1 ;
[0016] Step 6: Calculate the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The nth correlation coefficient C n :
[0017]
[0018] in, Represents the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The covariance of Represents the nth demodulated phase waveform φ n The standard deviation of Represents the n+1th demodulated phase waveform φ n+1 The standard deviation of
[0019] Step 7: The nth correlation coefficient C n Compare with the decision threshold δ:
[0020] If the nth correlation coefficient C n If the value is greater than or equal to the set decision threshold δ, then the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 That is, the time domain waveform of the unknown vibration event to be measured;
[0021] Otherwise, set n=n+1 and return to step 5.
[0022] In the above step 1, the value range of the initial phase φ is [-π, π].
[0023] In the above steps 4 and 7, the decision threshold δ is set to 0.999.
[0024] Compared with the prior art, the present invention has the following characteristics:
[0025] 1. On the basis of existing phase demodulation, the correlation coefficient is introduced to judge the similarity of phase waveform. The strain type of the external disturbance signal can be quickly and accurately determined by the similarity of signal waveform, and the traditional unwrapping and differential unwrapping are combined together by using the judgment condition;
[0026] 2. When performing differential dewrapping, there is no need to obtain the amplitude and frequency information of the external disturbance signal in advance, that is, there is no need to compare the recovered phase waveform with the external disturbance signal to determine the number of differentials. Instead, the required number of differentials is directly determined by the correlation coefficient;
[0027] 3. Effectively improve the system's waveform recovery efficiency and accuracy for unknown vibration events, ensuring signal demodulation performance in various complex environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 The schematic diagram is a schematic diagram of an unknown vibration event waveform detection method based on φ-OTDR.
[0029] Figure 2The waveform diagram of Experiment 1, (a) is the waveform of the traditional phase unwrapping demodulation, and (b) is the waveform of the first differential unwrapping summation demodulation.
[0030] Figure 3 The waveform diagrams of Experiment 2 are as follows: (a) is the waveform of the traditional phase unwrapping demodulation, (b) is the waveform of the first differential unwrapping and summing demodulation, and (c) is the waveform of the second differential unwrapping and summing demodulation. DETAILED DESCRIPTION
[0031] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in combination with specific examples and with reference to the accompanying drawings.
[0032] A method for detecting unknown vibration event waveform based on φ-OTDR, such as Figure 1 As shown, it includes the following steps:
[0033] Step 1: Use φ-OTDR to collect the disturbance signal caused by the unknown vibration event to be measured, and perform phase demodulation on the disturbance signal to obtain the initial phase φ, whose value range is [-π, π].
[0034] The existing method is used to demodulate the phase of the disturbance signal, and the initial phase φ(i,j) obtained is:
[0035]
[0036] Among them, φ ij It represents the phase value of the disturbance at the jth fiber sampling point under the i-th pulse, i=1,2,…,I, I is the number of fiber disturbance sampling pulses, which determines the data duration; j=1,2,…,J, J is the number of fiber disturbance sampling points, which is related to the fiber length and sampling frequency. The row vector in the matrix represents the phase change in the fiber length direction (spatial domain), and the column vector represents the phase change in the number of pulses (time domain).
[0037] Step 2: Perform traditional phase unwrapping on the initial phase φ to obtain an initial demodulated phase waveform φ0, and perform a differential unwrapping and summing on the initial phase φ to obtain a first demodulated phase waveform φ1.
[0038] The process of traditional phase unwrapping is as follows:
[0039] ① In order to eliminate the noise accumulated along the fiber length (in space) of the disturbance event, a reference phase φ is selected before the disturbance event and near the disturbance position. 0 (i,1), and let the initial phase φ(i,j) minus the reference phase φ 0 (i, j) obtains the phase φ minus the initial phase effect 1 (i,j); where φ 1 (i,j)=φi,j)-φ0 (i,1).
[0040] ② The phase φ after subtracting the initial phase effect 1 (i,j) is unwrapped in time to obtain the unwrapped phase φ 2 (i,j), where φ 2 (i,j)=φ 1 (i,j)+2kπ, where k is an integer.
[0041] ③ For the unwinding phase φ 2 (i,j) minus the mean Remove the influence of temporal noise to obtain the mean-subtracted phase φ 3 (i,j), where
[0042] The process of a differential decomposition and winding summation is as follows:
[0043] ① The phase φ after subtracting the initial phase effect 1 (i, j) performs a time difference operation on adjacent pulses to obtain a differential phase φ 4 (i,j), where φ 4 (i,j)=Δ 1 φ 1 (i,j), the specific operation process is:
[0044] The phase φ after subtracting the initial phase effect 1 (i, j) Subtract two adjacent pulses to obtain the subtraction phase φ 4′ (i',j), where φ 4′ (i',j)=φ 1 (2:I,j)-φ 1 (1:I-1,j),φ 1 (2:I,j) is φ 1 The I-1×J-dimensional submatrix formed by the 2nd to the Ith row of (i,j), φ 1 (1:I-1,j) is φ 1 The I-1×J-dimensional submatrix formed by the 1st to I-1th rows of (i,j), i'=1,2…,I-1.
[0045] Since one pulse will be reduced when the phase of two adjacent pulses is subtracted, that is, the phase matrix will have one less row, in order to keep the dimension of the phase unchanged, a 1×j-dimensional all-zero phase is vertically spliced to the subtraction phase φ 4′ The first row of (i',j) yields a differential phase φ 4 (i,j).
[0046] ② For the primary differential phase φ 4(i, j) is unwrapped in time to obtain an unwrapped phase φ 5 (i,j), where φ 5 (i,j)=φ 4 (i,j)+2kπ, where k is an integer.
[0047] ③ For the first unwinding phase φ 5 (i, j) performs a summation operation on adjacent pulses in time to obtain a summation phase φ 6 (i,j), where φ 6 (i,j)=∑ 1 φ 5 (i,j), the specific operation process is:
[0048] The first unwinding phase φ 5 (i, j) Two adjacent pulses are added to obtain the added phase φ 6 (i',j), where φ 6′ (i',j)=φ 5 (2:I,j)+φ 5 (1:I-1,j),φ 5 (2:I,j) is φ 5 The I-1×J-dimensional submatrix formed by the 2nd to the Ith row of (i,j), φ 5 (1:I-1,j) is φ 5 The I-1×J-dimensional submatrix formed by the 1st to I-1th rows of (i,j), i'=1,2…,I-1.
[0049] Since one pulse will be reduced when two adjacent pulses of the phase are summed, that is, the phase matrix will have one less row, in order to keep the dimension of the phase unchanged, the phase φ(1,j) at the initial moment of the initial phase φ(i,j) is vertically spliced to the added phase φ 6′ The first row of (i',j) yields a summation phase φ 6 (i,j).
[0050] ④ For the first summation phase φ 6 (i,j) minus the mean Remove the influence of temporal noise to obtain the mean-subtracted phase φ 7 (i,j), where
[0051] Step 3: Calculate the initial correlation coefficient C0 between the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1.
[0052] The initial correlation coefficient C0 is:
[0053]
[0054] in, represents the covariance of the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1, represents the standard deviation of the initial demodulated phase waveform φ0, represents the standard deviation of the first demodulated phase waveform φ1.
[0055] Step 4: Compare the initial correlation coefficient C0 with the set decision threshold δ:
[0056] If the initial correlation coefficient C0 is greater than or equal to the decision threshold δ (C0 ≥ δ), the unknown vibration event to be measured is determined to be a small dynamic strain signal. For small dynamic strain, the absolute phase jump between two adjacent points is less than or equal to π, and the traditional phase demodulation method can accurately restore the phase, so the traditional unwinding is effective, and the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1 are the time domain waveforms of the unknown vibration event to be measured;
[0057] Otherwise, the unknown vibration event to be measured is determined to be a large dynamic strain signal. For large dynamic strain, the absolute phase jump between two adjacent points is large, and the traditional phase demodulation method cannot accurately restore the phase. The traditional unwrapping is invalid, and it is necessary to further use the differential unwrapping method to solve it, that is, set n = 1, and continue to execute step 5.
[0058] Step 5: Demodulate the nth phase waveform φ n Perform n+1 times of differential detangling and summing to obtain the n+1th demodulated phase waveform φ n+1 .
[0059] The process of n-times difference decomposition and winding summation is as follows:
[0060] ① The phase φ after subtracting the initial phase effect 1 (i, j) performs n adjacent pulse differential operations in time to obtain n differential phases φ 8 (i,j), where φ 8 (i,j)=Δ n φ 1 (i,j), the specific operation process is:
[0061] The phase φ after subtracting the initial phase 1 (i, j) performs n cycles of iterations in step ① of the differential dewrapping and summation process to obtain the n-time differential phase φ 8 (i,j).
[0062] ② For n-order differential phase φ 8 (i,j) is unwrapped in time to obtain n unwrapped phases φ 9 (i,j), where φ 9 (i,j)=φ 8(i,j)+2kπ, where k is an integer.
[0063] ③ For the nth unwinding phase φ 9 (i, j) performs n-time summation of adjacent pulses in time to obtain n-time summation phase φ 10 (i,j), where φ 10 (i,j)=∑ n φ 9 (i,j), the specific operation process is:
[0064] For the nth unwrapping phase φ 9 (i, j) performs n iterations of the operation process of step ③ in the differential detangling summation to obtain the summation phase φ 10 (i,j).
[0065] ④ For n times the summed phase φ 10 (i,j) minus the mean Remove the influence of temporal noise to obtain the mean-subtracted phase φ 11 (i,j), where
[0066] Step 6: Calculate the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The nth correlation coefficient C n :
[0067]
[0068] in, Represents the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The covariance of Represents the nth demodulated phase waveform φ n The standard deviation of Represents the n+1th demodulated phase waveform φ n+1 The standard deviation of .
[0069] Step 7: The nth correlation coefficient C n Compare with the set decision threshold δ:
[0070] If the nth correlation coefficient C n is greater than or equal to the decision threshold δ, then the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 That is, the time domain waveform of the unknown vibration event to be measured;
[0071] Otherwise, set n=n+1 and return to step 5.
[0072] The performance of the present invention is described below by experiments:
[0073] Experiment 1: A distributed sensing experiment was conducted on a standard single-mode optical fiber of 800 m. An acousto-optic modulator with a frequency shift of 200 MHz was introduced and driven by an arbitrary waveform generator. The detection pulse repetition period was 10 μs and the pulse width was 100 ns. The disturbance event was excited by a piezoelectric ceramic stretcher (PZT) located at 550 m-600 m of the optical fiber. A disturbance signal was randomly generated by the PZT. The signal was demodulated to obtain two phase waveforms, as shown in Figure 2 As shown, (a) is the waveform of the traditional phase unwrapping demodulation, and (b) is the waveform of the first differential unwrapping sum demodulation. As can be seen from the figure, the correlation coefficient between the waveform of the first differential unwrapping sum demodulation and the waveform of the traditional phase unwrapping demodulation is 0.999, and the two waveforms are consistent. The phase waveform can be correctly restored using the traditional phase unwrapping, so this disturbance signal is a small dynamic strain.
[0074] Experiment 2: A distributed sensing experiment was conducted on a standard single-mode optical fiber of 800m. An acousto-optic modulator with a frequency shift of 200MHz was introduced and driven by an arbitrary waveform generator. The detection pulse repetition period was 15μs and the pulse width was 100ns. The disturbance event was excited by a piezoelectric ceramic stretcher (PZT) located at 550m-600m of the optical fiber. A disturbance signal was randomly generated by the PZT. The signal was demodulated by traditional phase unwrapping, first differential dewrapping and summing, and second differential dewrapping and summing. The waveforms are shown below. Figure 3 As shown, (a) is the waveform of the traditional phase unwrapping demodulation, (b) is the waveform of the first differential unwrapping and demodulation, and (c) is the waveform of the second differential unwrapping and demodulation. As can be seen from the figure, the correlation coefficient between the waveform of the first differential unwrapping and demodulation and the waveform of the traditional phase unwrapping demodulation is -0.465, which does not meet the conditions, and the traditional phase unwrapping fails. This signal has a large dynamic strain and requires differential operation; the correlation coefficient between the waveform of the second differential unwrapping and demodulation and the waveform of the first differential unwrapping and demodulation is 1, which meets the conditions. The two waveforms are consistent, and the phase waveform can be correctly restored using the first differential unwrapping and demodulation.
[0075] The above two experimental results show that the present invention can correctly restore the waveform of an unknown external disturbance event.
[0076] The present invention proposes a waveform detection method for unknown vibration events based on φ-OTDR, which uses correlation coefficients to determine the type of disturbance signal, uses traditional phase unwrapping demodulation for small dynamic strains, and uses differential unwrapping sum demodulation for large dynamic strains, and uses correlation coefficients to determine the required number of differentials when performing differential unwrapping sum demodulation. The present invention can accurately restore the waveform of an external disturbance event without relying on prior knowledge, can reduce data processing time, and improve waveform detection efficiency.
[0077] It should be noted that although the embodiments of the present invention described above are illustrative, they are not intended to limit the present invention, and therefore the present invention is not limited to the above specific embodiments. Without departing from the principles of the present invention, any other embodiments obtained by those skilled in the art under the guidance of the present invention are deemed to be within the protection of the present invention.
Claims
1. A method for detecting unknown vibration event waveform based on φ-OTDR, characterized in that: The steps include: Step 1: Use φ-OTDR to collect the disturbance signal caused by the unknown vibration event to be measured, and perform phase demodulation on the disturbance signal to obtain the initial phase φ; Step 2, perform conventional phase unwrapping on the initial phase φ to obtain an initial demodulated phase waveform φ0, perform differential unwrapping and summing on the initial phase φ to obtain a first demodulated phase waveform φ1; Step 3: Calculate the initial correlation coefficient C0 between the initial demodulation phase waveform φ0 and the first demodulation phase waveform φ1: in, represents the covariance of the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1, represents the standard deviation of the initial demodulated phase waveform φ0, represents the standard deviation of the first demodulated phase waveform φ1; Step 4: Compare the initial correlation coefficient C0 with the set decision threshold δ: If the initial correlation coefficient C0 is greater than or equal to the decision threshold δ, the initial demodulated phase waveform φ0 and the first demodulated phase waveform φ1 are the time domain waveforms of the unknown vibration event to be measured; Otherwise, set n=1 and go to step 5; Step 5: Demodulate the nth phase waveform φ n Perform n+1 times of differential detangling and summing to obtain the n+1th demodulated phase waveform φ n+1 ; Step 6: Calculate the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The nth correlation coefficient C n : in, Represents the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 The covariance of Represents the nth demodulated phase waveform φ n The standard deviation of Represents the n+1th demodulated phase waveform φ n+1 The standard deviation of Step 7: The nth correlation coefficient C n Compare with the decision threshold δ: If the nth correlation coefficient C n If the value is greater than or equal to the set decision threshold δ, then the nth demodulated phase waveform φ n and the n+1th demodulated phase waveform φ n+1 That is, the time domain waveform of the unknown vibration event to be measured; Otherwise, set n=n+1 and return to step 5.
2. The method for detecting an unknown vibration event waveform based on φ-OTDR according to claim 1, characterized in that: In step 1, the value range of the initial phase φ is [-π, π].
3. The method for detecting an unknown vibration event waveform based on φ-OTDR according to claim 1, characterized in that: In steps 4 and 7, the decision threshold δ is set to 0.999.