Heterodyne coherent detection phi-OTDR frequency shift demodulation method based on multi-frequency decomposition

Abstract

The invention relates to the technical field of distributed optical fiber vibration sensing, in particular to a heterodyne coherent detection phi-OTDR frequency shift demodulation method based on multi-frequency decomposition, which comprises the following steps: decomposing a conventional single pulse detection signal into an equivalent sweep frequency pulse detection signal through a multi-frequency decomposition method based on short-time Fourier transform (STFT); the signal presents a Rayleigh scattering pattern in distance and frequency dimensions; and performing cross-correlation on the Rayleigh scattering pattern of the equivalent sweep frequency pulse detection signal along the frequency axis, and demodulating the frequency shift compensation amount required by the dynamic strain. Restoring the dynamic dependent variable of the vibration signal according to the relationship between the frequency shift and the strain; and deriving a quantitative relationship among the sampling rate, the signal-to-noise ratio of the detection signal, the interception width of the sliding window and the demodulation strain noise power. According to the invention, the problems of complex structure and high cost of a stepping frequency sweeping or chirped pulse modulation system are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed optical fiber vibration sensing, and particularly relates to a heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition. Background Art

[0002] In recent years, the phase-sensitive optical time domain reflectometer (Φ-OTDR) has been widely applied to fields such as oil and gas pipeline monitoring, national defense security monitoring, power cable monitoring, and structural health monitoring of large-scale infrastructure due to its characteristics of large scattered power, high sensitivity, and fast response speed. With the demand for quantitative monitoring of disturbances in practical applications, the dynamic strain demodulation of Ф-OTDR has become a current research hotspot.

[0003] In the past decade, researchers have proposed various methods for quantifying dynamic strain from two perspectives: phase demodulation and frequency shift demodulation. First is phase demodulation, that is, the disturbance information is restored by demodulating the scattered light phase that has a linear response to the disturbance. Its characteristics are high demodulation accuracy and simple demodulation process. However, phase demodulation has the disadvantage of being affected by coherent fading. For this reason, researchers have proposed an intrinsically anti-coherent fading frequency shift demodulation method.

[0004] In 2009, Y. Koyamada et al. from Ibaraki University, Japan, proposed a coherent optical time domain reflectometer (COTDR) based on Rayleigh scattering pattern demodulation. This scheme first constructs a Rayleigh scattering pattern by obtaining the Rayleigh scattering intensity signals of a series of detection pulses with different frequencies. Since the change in the optical path of the detection light caused by strain can be linearly compensated by the frequency shift of the detection light, the intensity of the Rayleigh scattering signal under disturbance is translated on the frequency axis. The frequency shift amount can be obtained through cross-correlation calculation to achieve quantitative demodulation of the disturbance. In 2016, J. Pastor-Graells et al. from the University of Alcalá, Spain, proposed chirped pulse Φ-OTDR. This scheme uses broadband chirped pulses as the detection light, converts the frequency shift phenomenon equivalent to the change in the optical path caused by the disturbance into the delay of the intensity waveform in the time domain, and thus the magnitude of the frequency shift amount can be obtained by estimating the time delay of the local scattered signal intensity waveform, and then the magnitude of the disturbance can be calculated. In addition, in 2021, H Qian proposed a conversion algorithm that converts a conventional narrow pulse detection signal into an equivalent chirped pulse detection signal through chirp factor convolution operation. Therefore, the previous frequency shift demodulation methods applicable to chirped pulse type Φ-OTDR (CP-ΦOTDR) can be applied to traditional coherent detection Φ-OTDR, breaking through the limitation of frequency shift demodulation on the detection pulse and reducing the system cost of demodulation.

[0005] The noise power is an important indicator for evaluating the sensing performance of the Φ-OTDR system, which directly determines the sensing length of the system, the strain floor noise, and the quality of the demodulated disturbance signal. The evaluation of the noise power has always been a key point in the research of Φ-OTDR. The Cramér-Rao lower bound (CRLB) is an important scheme for evaluating the minimum noise power that all these unbiased estimation methods can achieve, and it has been widely used in fields such as radar and sonar. In 2019, Luís Costa et al. calculated the theoretical lower limit of the noise of chirped-pulse Φ-OTDR, reaching the performance limit of a given signal parameter set. The directly detected COTDR can also be regarded as a "time delay (frequency shift)" estimation problem in the frequency domain. In October 2021, the Thévenaz team in Switzerland utilized this characteristic to obtain the Cramér-Rao lower bound for the frequency shift estimation of directly detected COTDR. In 2022, J Jiang et al. from the University of Electronic Science and Technology of China calculated the Cramér-Rao lower bound of coherent detection COTDR. Although the frequency shift demodulation method is intrinsically anti-fading, it requires stepped frequency sweeping or chirped pulse modulation of the detection pulse, resulting in a complex system structure and high cost, making it difficult to apply in complex situations. Therefore, it is of great significance to study the equivalent conventional pulse conversion method and analyze the influence of parameters on the performance. Summary of the Invention

[0006] The purpose of the present invention is to provide a heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition, which solves the problems of complex system structure and high cost of stepped frequency sweeping or chirped pulse modulation systems. Through equivalent conversion, the dynamic disturbance signal is restored using frequency shift demodulation in a conventional single pulse. Since the equivalent frequency-swept pulse detection signal is generated based on the conversion of a conventional single pulse, there is no need for a time-consuming frequency-sweeping process. In addition, because the demodulated strain noise determines the demodulation performance, the relationship between the demodulated strain noise, the truncation width, the signal-to-noise ratio, and the sampling rate is explained through theoretical derivation. This technology not only reduces the cost of demodulation but also provides theoretical guidance for the optimization of such systems in the future.

[0007] To achieve the above technical objectives and reach the above technical effects, the present invention is realized through the following technical solutions:

[0008] A heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition, comprising the following steps:

[0009] S1: Decompose the conventional single-pulse detection signal into an equivalent frequency-swept pulse detection signal through a multi-frequency decomposition method based on the short-time Fourier transform (STFT), and this signal presents a Rayleigh scattering pattern in the distance and frequency dimensions.

[0010] S2: Perform cross - correlation on the Rayleigh scattering pattern of the equivalent swept - frequency pulse detection signal along the frequency axis to demodulate the frequency - shift compensation amount required for dynamic strain. According to the relationship between frequency shift and strain, restore the dynamic strain amount of the vibration signal.

[0011] S3: Derive the quantitative relationship between the sampling rate, the signal - to - noise ratio of the detection signal, the intercept width of the sliding window, and the demodulated strain noise power.

[0012] Furthermore, the Rayleigh scattering pattern in step S1 includes:

[0013] The complex form of the output electrical signal of the heterodyne coherent detection Φ - OTDR can be expressed as

[0014]

[0015] where is the convolution operation, m(t) is the pulse modulation signal, and f(t) is the impulse response of the heterodyne coherent detection Φ - OTDR system.

[0016] Perform STFT transformation on v(t) using a sliding window with a window function g(t w ), and the obtained Rayleigh scattering pattern is expressed as

[0017]

[0018] where

[0019]

[0020] In the formula, m e (f w ,t) is the equivalent modulation pulse, f w is the frequency variable in STFT, M(f w ), G(f w ) are the Fourier transforms of m(t w ), g(t w ) respectively.

[0021] Furthermore, the equivalent swept - frequency pulse detection signal in step S1 includes:

[0022] Comparing v(t) and V e (f w ,t), it can be seen that the response signal v(t) is generated by convolving the system response f(t) with the conventional modulation pulse m(t). After STFT transformation, it can be equivalently considered as the response signal V e (f w ,t) generated by convolving the system response f(t) with the upper - equivalent modulation pulse m e (f w ,t). By changing V e (fw , the frequency variable f in (, t) w , equivalent swept-frequency signals of different frequencies can be obtained.

[0023] Furthermore, the step S2 includes:

[0024] In frequency shift demodulation, when the entire effective scattering region of the pulse is uniformly disturbed, the relative change in the optical path difference between scattering points caused by the disturbance strain can be linearly compensated by the frequency shift of the incident light. Among them, the relationship between the frequency shift amount and the disturbance is:

[0025]

[0026] In the formula, Δv is the frequency shift amount, v 0 is the central frequency, and Δε is the strain amount.

[0027] Through the relationship between the frequency shift amount and the disturbance, perform a cross-correlation operation on the intensity curve of the backward Rayleigh scattered light, estimate the position of the correlation peak to obtain the frequency shift amount, and through the above conversion, restore the dynamic strain amount of the vibration signal.

[0028] Furthermore, the quantitative relationship between the sampling rate, the signal-to-noise ratio of the detection signal, the width of the sliding window intercept, and the demodulated strain noise power in the step S3 is:

[0029]

[0030] where SNR is the signal-to-noise ratio in the time domain, T e is the width of the short-time Fourier transform intercept window, and Fs is the sampling rate. The demodulated strain noise power is inversely proportional to SNR, T e 3 .

[0031] On the other hand, the present invention proposes a system based on the above frequency shift demodulation method, including: a narrow linewidth laser, an acousto-optic modulator, an erbium-doped fiber amplifier, a circulator, a measurement fiber, an optical coupler, a balanced photodetector, and a data acquisition card.

[0032] Furthermore, the continuous-wave light source generated by the narrow-linewidth laser is divided into two paths, the upper path and the lower path, by a coupler. 90% of the optical power is used to generate pulsed sound, and 10% of the optical power serves as the local oscillator light. The acousto-optic modulator modulates the continuous-wave light source into a pulsed optical signal. The acousto-optic modulator not only modulates the continuous light into pulsed light but also introduces a fixed frequency shift. The pulsed optical signal is amplified after entering the erbium-doped fiber amplifier, and then the signal enters the measurement fiber through a circulator. When the optical signal is transmitted in the fiber, backward Rayleigh scattering signals are continuously generated due to Rayleigh scattering. The backward Rayleigh scattering signals returning along the original path are received by a balanced photodetector after beating with the local oscillator light, and the optical signal is converted into an electrical signal. The electrical signal is digitally processed by a data acquisition card to obtain a digital signal.

[0033] Furthermore, in the Φ-OTDR system using a narrow-linewidth laser, the optical pulse has good stability and coherence; the erbium-doped fiber amplifier increases the input fiber power of the modulated pulsed signal and improves the sensing distance of the signal in the fiber; introducing local oscillator light beating can amplify the Rayleigh scattering light and effectively suppress the influence of the broadband noise generated by the erbium-doped fiber amplifier on the signal, further improving the signal-to-noise ratio of the system.

[0034] The beneficial effects of the present invention:

[0035] Through the multi-frequency decomposition method based on the short-time Fourier transform (STFT), the present invention converts the single-pulse detection signal into an equivalent swept-frequency pulse detection signal, and frequency shift demodulation can be realized in the traditional coherent detection Φ-OTDR system. Based on the present invention, the frequency shift demodulation method applicable to the swept-frequency pulse type (COTDR) can be applied to the traditional coherent detection Φ-OTDR system, breaking through the limitation of frequency demodulation on the detection pulse and reducing the demodulation cost.

[0036] The present invention adopts the traditional coherent detection Φ-OTDR system, and the system structure is simple and easy to implement. Only by using the single-pulse detection signal combined with the conversion algorithm can frequency shift demodulation be realized, without the time-consuming stepped frequency sweeping and complex chirped pulse modulation. Due to the advantages of simple system structure and easy implementation, the system has high economy and practicability in different application scenarios.

[0037] The present invention inherits the advantages of frequency demodulation against coherent fading and independent perturbation length demodulation. Although the traditional phase demodulation has simple demodulation steps, it is easily affected by coherent fading, resulting in unsatisfactory demodulation effects. The advantage of the present invention is that frequency shift demodulation can be used in the traditional coherent detection Φ-OTDR system, retaining the inherent advantage of anti-coherent fading and at the same time solving the problem of complex system.

[0038] The present invention theoretically derives the quantitative relationship among the sampling rate, the signal-to-noise ratio of the detection signal, the intercepted width of the sliding window, and the demodulation strain noise power, which can more clearly understand the relationship among the parameters. This relationship provides a design direction for the optimization of such systems and also provides a reference for the popularization of the formula.

[0039] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 It is a schematic diagram of the conversion principle of the equivalent swept-frequency pulse detection signal in the present invention.

[0042] Figure 2 It is a simulation diagram of the demodulated strain signal of the equivalent swept-frequency pulse detection signal in the present invention.

[0043] Figure 3 It is a fitting diagram of the intercepted width, signal-to-noise ratio, sampling rate, and theoretical relationship formula in the present invention.

[0044] Figure 4 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0046] Embodiment 1

[0047] A heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition described in this embodiment includes:

[0048] First, an external heterodyne coherent detection Φ-OTDR system composed of a narrow linewidth laser, an acousto-optic modulator, an erbium-doped fiber amplifier, a circulator, a measurement fiber, an optical coupler, a balanced photodetector, and a data acquisition card is built.

[0049] Then, a demodulation method based on the above system is used, including the following steps:

[0050] S1: Decompose the conventional single-pulse detection signal into an equivalent swept-frequency pulse detection signal through a multi-frequency decomposition method based on the short-time Fourier transform (STFT). This signal presents a Rayleigh scattering pattern in the range and frequency dimensions.

[0051] S2: Perform cross-correlation on the Rayleigh scattering pattern of the equivalent swept-frequency pulse detection signal along the frequency axis to demodulate the frequency shift compensation amount required for dynamic strain. According to the relationship between frequency shift and strain, restore the dynamic strain amount of the vibration signal.

[0052] S3: Derive the quantitative relationship between the sampling rate, the signal-to-noise ratio of the detection signal, the sliding window truncation width, and the demodulated strain noise power.

[0053] Example 2

[0054] An heterodyne coherent detection Φ-OTDR system based on multi-frequency decomposition as described in this example, as Figure 1 shown, is a schematic diagram of the conversion principle of the equivalent swept-frequency pulse detection signal: Build an heterodyne coherent detection Φ-OTDR system consisting of a narrow-linewidth laser, an acousto-optic modulator, an erbium-doped fiber amplifier, a circulator, a measurement fiber, an optical coupler, a balanced photodetector, and a data acquisition card. First, obtain the backward Rayleigh scattering signal after beating with the local light, and then apply the short-time Fourier transform to it to obtain a two-dimensional map of time and frequency distribution. Finally, sample the frequency dimension of this map to construct an equivalent Rayleigh scattering pattern in the range and frequency dimensions.

[0055] Example 3

[0056] An heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition as described in this example. As Figure 2 shown, is a schematic diagram of simulating the demodulation of the strain signal of the equivalent swept-frequency pulse detection signal. The following is a demonstration through simulation: The sampling rate is set to 1 GHz, the pulse width is 10 ns, and the carrier frequency is set to 200 MHz. The optical fiber is 220 m in total, and a sine signal perturbation is added at 140 - 160 m. The sampling rate of the strain signal is 0.1 MHz, the frequency is 1000 Hz, the perturbation time is set to 0.002 s, and the strain amplitude is 10 nε. Sample the perturbation signal by inputting a continuous conventional single-pulse detection signal. After all the obtained backward Rayleigh scattering signals are converted into Rayleigh scattering patterns, take out the signals in the frequency dimension at 150 m on the range axis for combination to obtain the spectrogram as shown in Figure 2 (a). It can be clearly seen the frequency shift between different spectral amplitude curves in the subfigure. Perform cross-correlation on the amplitude curves by the least squares method to demodulate the frequency shift amount between the curves, and then according to the relationship between strain and frequency shift, restore the perturbation signal, as shown in Figure 2 (b), which is consistent with the applied strain.

[0057] Example 4

[0058] A heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition as described in this embodiment. As Figure 3 shown, the schematic diagram of parameter fitting affecting the noise power. The feasibility and accuracy of the theoretical relationship are verified through simulation. The blue dots in the figure are the noise standard deviations obtained through simulation calculations, and the red curve is the fitting curve calculated according to the derived theoretical relationship. Figure 3 They are respectively the fitting of the intercept width T e 3 the fitting of the signal-to-noise ratio SNR, and the fitting of the sampling rate Fs. Figure 3 The simulation results are in good agreement with the theoretical analysis, verifying the feasibility and accuracy of the theoretical relationship.

[0059] In summary, the heterodyne coherent detection Φ-OTDR frequency shift demodulation method based on multi-frequency decomposition proposed by the present invention. In heterodyne coherent detection Φ-OTDR, the multi-frequency decomposition conversion based on the short-time Fourier transform is performed on the backward Rayleigh scattering signal, and the conventional single-pulse detection signal can be converted into an equivalent swept-frequency pulse detection signal. Thus, the frequency shift demodulation technology in swept-frequency pulse detection Φ-OTDR can be applied to the traditional coherent detection Φ-OTDR, realizing the use of frequency shift demodulation to recover the dynamic disturbance signal in the conventional single-pulse detection signal. Since the equivalent swept-frequency pulse detection signal is generated based on the conversion of the conventional single-pulse detection signal, there is no need to perform a time-consuming frequency sweeping process. In addition, the present invention derives the quantitative relationship between the demodulated strain noise power, the sampling rate, the signal-to-noise ratio of the detection signal, and the sliding window intercept width, providing theoretical guidance for the optimization of such systems in the future. Compared with the frequency shift demodulation technology based on swept-frequency pulse detection, the present invention improves the speed of frequency shift demodulation and reduces the system cost of frequency sweeping.

[0060] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the relevant technical fields can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for frequency shift demodulation of heterodyne coherent detection Φ-OTDR based on multi-frequency decomposition, characterized in that: The following steps are involved: S1: The conventional single pulse detection signal is decomposed into an equivalent swept frequency pulse detection signal by a multi-frequency decomposition method based on short-time Fourier transform. The signal presents a Rayleigh scattering pattern in the distance and frequency dimensions. S2: Cross-correlate the Rayleigh scattering pattern of the equivalent swept-frequency pulse detection signal along the frequency axis to demodulate the frequency shift compensation required for dynamic strain; based on the relationship between frequency shift and strain, restore the dynamic strain of the vibration signal; S3: Derive the quantitative relationship between sampling rate, detection signal-to-noise ratio, sliding window interception width and demodulation strain noise power.

2. The heterodyne coherent detection Φ-OTDR frequency shift demodulation method according to claim 1, characterized in that: The Rayleigh scattering pattern in step S1 includes: The complex form of the electrical signal output by heterodyne coherent detection Φ-OTDR is expressed as in, is the convolution operation, m(t) is the pulse modulation signal, and f(t) is the impulse response of the heterodyne coherent detection Φ-OTDR system; Use a window function g(t w ) sliding window performs STFT transformation on v(t), and the obtained Rayleigh scattering pattern is expressed as in, In the formula, m e (f w ,t) is the equivalent modulation pulse, f w is the frequency variable in STFT, M(f w ), G(f w ) are respectively m(t w ), g(t w ) is the Fourier transform of .

3. The heterodyne coherent detection Φ-OTDR frequency shift demodulation method according to claim 1, characterized in that: The equivalent frequency sweep pulse detection signal in step S1 includes: Comparison of v(t) and V e (f w ,t), the response signal v(t) is generated by the convolution of the system response f(t) with the conventional modulation pulse m(t). After STFT transformation, it can be equivalent to the convolution of the system response f(t) with the equivalent modulation pulse m e (f w ,t) The response signal V e (f w ,t); by changing V e (f w ,t) in the frequency variable f w , the equivalent swept frequency signals of different frequencies can be obtained.

4. The heterodyne coherent detection Φ-OTDR frequency shift demodulation method according to claim 1, characterized in that: The step S2 comprises: In frequency shift demodulation, when the entire pulse effective scattering area is uniformly disturbed, the relative change in the optical path difference between scattering points caused by the disturbance strain can be linearly compensated by the frequency shift of the incident light; the relationship between the frequency shift and the disturbance is: In the formula, Δv is the frequency shift, v0 is the center frequency, and Δε is the strain; Through the relationship between frequency shift and disturbance, the backscattered Rayleigh light intensity curve is cross-correlated and the correlation peak position is estimated to obtain the frequency shift. After the above conversion, the dynamic strain of the vibration signal is restored.

5. The heterodyne coherent detection Φ-OTDR frequency shift demodulation method according to claim 1, characterized in that: The quantitative relationship between the sampling rate, the detection signal noise ratio, the sliding window interception width and the demodulation strain noise power in step S3 is: Where SNR is the signal-to-noise ratio in the time domain, T e Short-time Fourier transform interception window width, Fs is the sampling rate; demodulation strain noise power and SNR, T e 3 , is inversely proportional to Fs.

6. A system based on the method according to any one of claims 1 to 3, characterized in that: Composition includes: Narrow linewidth laser, acousto-optic modulator, erbium-doped fiber amplifier, circulator, measurement fiber, optical coupler, balanced photodetector, data acquisition card.

7. The system according to claim 6, characterized in that: The continuous wave light source generated by the narrow linewidth laser is divided into two paths through the coupler, 90% of the light power is used to generate pulse sound, and 10% of the light power is used as local oscillator light; the acousto-optic modulator modulates the continuous wave light source into a pulse light signal; The acousto-optic modulator not only modulates continuous light into pulsed light, but also introduces a fixed frequency shift; the pulsed light signal enters the erbium-doped fiber amplifier for amplification, and then the signal passes through the circulator into the measuring fiber; when the light signal is transmitted in the optical fiber, it will continuously generate back-scattered Rayleigh signals due to Rayleigh scattering; the back-scattered Rayleigh signals returning along the original path are received by the balanced photodetector after the optical coupler beats the local oscillator light, and the light signal is converted into an electrical signal; the electrical signal is digitized by the data acquisition card to obtain a digital signal.

8. The heterodyne coherent detection Φ-OTDR system according to claim 6, characterized in that: The Φ-OTDR system using narrow linewidth lasers has good optical pulse stability and coherence; the erbium-doped fiber amplifier increases the fiber power of the modulated pulse signal and increases the sensing distance of the signal in the optical fiber; the introduction of the local oscillator beat frequency can amplify the Rayleigh scattered light, effectively suppressing the influence of the broadband noise generated by the erbium-doped fiber amplifier on the signal, and further improving the signal-to-noise ratio of the system.

Images