DAS large swing rate vibration signal recovery method based on adaptive multi-stage unwinding

Abstract

The application provides a DAS large swing rate vibration signal recovery method based on adaptive multi-stage unwinding, which comprises the following steps: acquiring a phase signal output by a DAS system demodulation and calculating an instantaneous swing rate; if the instantaneous swing rate is greater than a first threshold, performing primary unwinding for basic correction; if the instantaneous swing rate is greater than a second threshold, triggering secondary unwinding, establishing a prediction model based on a historical phase trend to assist in determining an optimal compensation order; if the instantaneous swing rate is greater than a third threshold, triggering multi-stage unwinding, and utilizing spatial continuity to perform cross-channel verification to determine a missing compensation order; and finally outputting a continuous phase signal recovered by unwinding. The application matches the unwinding strategy dynamically according to the vibration intensity through a hierarchical triggering mechanism, accurately recovers the continuous phase signal under the large swing rate condition, avoids waveform distortion and spectrum leakage, expands the large amplitude vibration measurement range, and ensures the real-time processing efficiency by entering the high-order calculation only when needed.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing technology, and in particular to a method for recovering DAS large slew rate vibration signals based on adaptive multi-level unwinding. Background Technology

[0002] Distributed fiber acoustic sensing (DAS) technology senses external vibrations by demodulating the phase changes of backscattered Rayleigh light. It offers advantages such as long-distance transmission, high spatial resolution, and resistance to electromagnetic interference, and is widely used in fields such as seismic exploration, pipeline monitoring, and security. However, due to the periodicity of the arctangent function, the directly demodulated phase is wrapped within the interval [-π, π). Phase dewrapping (i.e., superimposing the correct integer multiples of 2kπ) is necessary to reconstruct the true continuous vibration waveform, which is a fundamental and crucial step in DAS signal processing.

[0003] The most mainstream untangling method currently is the single-step point-by-point correction algorithm (Itoh algorithm). Its core assumption is that the true phase difference between adjacent sampling points does not exceed π. It determines whether to add ±2π compensation by comparing the difference between the current wrapped phase and the untangled phase at the previous moment. This algorithm is simple to calculate, has good real-time performance, and can work effectively when the vibration is gentle. However, its fundamental limitation is that when the external vibration is severe and the phase yaw rate is extremely high, the phase difference between adjacent points may far exceed π. The algorithm cannot identify the missing 2kπ (k>1) multiples of tangling, resulting in abrupt distortion and spectral leakage in the untangled waveform, which seriously restricts the DAS's ability to measure vibration signals with a large dynamic range.

[0004] Furthermore, the inherent coherent fading of the DAS system causes the amplitude of the interference signal to drop into the noise floor. At this time, the demodulated phase is actually pure noise and appears as a false high-frequency large slew rate jump. Traditional algorithms cannot distinguish between real vibration and fading noise, and are prone to accidentally triggering unwinding operations, which in turn amplifies the noise.

[0005] In summary, existing technologies struggle to achieve accurate and robust phase recovery under complex conditions such as high slew rate, low signal-to-noise ratio, and coherent fading. There is an urgent need for an untangling scheme with hierarchical discrimination and adaptive capabilities. Summary of the Invention

[0006] This invention proposes a method for recovering DAS high oscillation vibration signals based on adaptive multi-level unwinding, which solves the problem that existing unwinding methods are difficult to achieve accurate and robust phase recovery under complex conditions such as high oscillation, low signal-to-noise ratio and coherent fading.

[0007] The technical solution of this invention is implemented as follows: This invention provides a method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding, comprising the following steps: Acquire the phase signal of the demodulated output of the DAS system and calculate the instantaneous slew rate of the phase signal. ; Set the first threshold. Second threshold Based on the comparison between the instantaneous slew rate and a preset threshold, the appropriate unwinding level is selected to process the phase signal at the current sampling moment: like Perform one unwinding operation to make a basic correction to the phase at the current moment; like This triggers secondary unwinding, establishes a local prediction model based on the phase change trend of historical sampling points, and uses the prediction results to help determine the optimal phase compensation order. like This triggers multiple unwinding processes, utilizes the spatial phase continuity of adjacent sensing channels to perform cross-channel verification, and determines the missing phase compensation order of the current channel. Outputs a continuous phase signal recovered after unwinding.

[0008] Specifically, the formula for calculating the instantaneous slew rate of the phase signal is: ; in, The current sampling time Instantaneous slew rate, This is the phase value output at the current sampling time. This is the phase value output at the previous sampling time.

[0009] Specifically, the method for performing one untangling operation is as follows: Wrap phase of the current sampling point Overlay The compensation amount yields the phase after one untangling. : ; ; in, This is the current sampling point number. It is an integer. The value of makes Phase after unwrapping from the previous sampling point Minimize the absolute value of the difference between them; by selecting the integer that minimizes the phase difference between adjacent points. To ensure that the distance between adjacent sampling points does not exceed Phase jump correction.

[0010] Specifically, the secondary untangling method includes the following steps: Using the previously correctly unwrapped data at the current sampling point A local prediction model is established based on historical sampling points to calculate the phase prediction value of the current sampling point. ; Calculate the package phase of the current sampling point With the phase prediction value The predicted residuals between : ); in, is an integer representing the candidate phase compensation order; By finding the predicted residual smallest integer Determine the optimal compensation order : ; in, This indicates the rounding operation; The phase value after the second unwinding is: ; in, This is the phase value output after two unwinding cycles.

[0011] Optionally, when using a first-order linear prediction model, take The predicted value of the current sampling point is calculated using the phase of the first two sampling points: ; in, For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding.

[0012] Optionally, when using a second-order parabolic prediction model, take An acceleration term is introduced, and the predicted value of the current sampling point is calculated using the phase of the previous three sampling points: ; After unfolding, we get: ; in, For the first Phase increment at time, For the first The second-order phase difference at time t; For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding.

[0013] Specifically, the method of untangling multiple times includes the following steps: Extract the target channel from the previous time step With adjacent reference channels Spatial phase gradient between : ; in, For the first Time target channel The true phase value after unwinding. For the first Time Reference Channel The true phase value after unwinding; Using the current time reference channel True phase after untangling Superimposed with the spatial phase gradient of the previous moment Calculate the target channel Spatial expected phase at the current moment : ; Build target channel Candidate solution set of true phase And calculate the phase of each candidate solution with the expected spatial phase. The residual is used to determine the optimal compensation order by finding the minimum residual. : ; in, This indicates the rounding operation. The wrapper phase of the demodulated output of the target channel at the current moment; The phase value after multiple unwindings is: ; in, This is the phase value output by the target channel after multiple unwinding processes.

[0014] Furthermore, before performing cross-channel verification, a dynamic optimization and verification step for the reference channel is also included: Calculate the target channel Adjacent channels and Each of the signal quality factors at the current moment With instantaneous slew rate ; From adjacent channels and The following channels are selected as reference channels: and ; in, This is a preset lower limit threshold for signal quality. For adjacent channels in Signal quality factor at time 10:00 For adjacent channels in The instantaneous oscillation rate at any given moment.

[0015] Preferably, the first threshold Second threshold Based on dynamic adjustment of the system signal-to-noise ratio and signal quality factor, the dynamic adjustment method is as follows: Calculate signal quality factor : ; ; in, for Signal quality factor at time 10:00 The envelope amplitude of the signal analyzed by the DAS system. The background noise amplitude threshold set for the system; The in-phase component signal at the current moment. The current time is the orthogonal component signal; The first threshold and the second threshold Adaptive adjustment is performed according to the following formula: ; in, express The time adjustment A threshold, ; For the system under fading-free, high signal-to-noise ratio conditions, the first One benchmark threshold constant; This represents the short-time signal-to-noise ratio within the current sliding time window. This serves as the system's reference signal-to-noise ratio constant. The preset lower limit threshold for signal quality, when At this time, the threshold is set to negative infinity, forcibly blocking the time axis unwinding of the current channel; and The adaptive adjustment coefficient is greater than zero. Controlling the threshold rise at low signal-to-noise ratios The nonlinear penalty force of the threshold when the control signal quality deteriorates.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention calculates the instantaneous slew rate of the demodulated phase signal of the DAS system and adaptively selects three different processing levels, namely, one-time unwinding, two-time unwinding, or multiple unwinding, based on the comparison result between the instantaneous slew rate and the preset threshold. Through the graded triggering mechanism, the system can dynamically match the corresponding unwinding strategy according to the intensity of vibration. Even under extreme conditions of large slew rate, it can still accurately recover the real continuous phase signal, effectively avoid waveform jump distortion and spectrum leakage, significantly expand the measurement range of the DAS system for large amplitude vibration signals, and only enter the high-order calculation module when needed, ensuring the real-time processing efficiency of the system. (2) In the second unwinding process, the present invention uses historical sampling points that have been correctly unwound before the current sampling point to establish a local prediction model based on Taylor expansion, calculates the phase prediction value of the current sampling point, and then calculates the residual between the wrapped phase and the prediction value. The optimal compensation order is determined by the minimum residual. Even if the actual phase difference exceeds π, the system can still extrapolate a reasonable phase expectation value based on physical inertia and accurately identify the missing winding order. This effectively solves the problem of signal distortion caused by missed winding in the traditional single algorithm. (3) In the case of single-point time series prediction failure caused by extreme large signals during multiple unwinding, the present invention introduces spatial coherence constraints, extracts the spatial phase gradient between the target channel and the adjacent reference channel at the previous moment, calculates the spatial expected phase of the target channel at the current moment, and then finds the optimal compensation order with the smallest residual with the spatial expected phase by constructing a candidate solution set. It makes full use of the extremely high spatial resampling rate of the DAS system and the spatial continuity characteristics of mechanical vibration in the fiber axial strain field, so that even in extreme working conditions where the time dimension information is completely insufficient to determine the winding order, the system can still complete the phase recovery by referring to the adjacent channel, which significantly improves the robustness of the DAS system in the severe vibration scenario. (4) In this invention, the first threshold and the second threshold are adjusted nonlinearly based on the system signal-to-noise ratio and the signal quality factor. When the signal-to-noise ratio decreases, the threshold is dynamically raised through an exponential penalty mechanism to avoid misjudging high-frequency background noise as large-swing vibration and erroneously triggering high-order unwinding. When coherent fading causes the signal quality factor to decrease, the threshold is significantly increased through a power function penalty mechanism to prevent failed phase data from participating in time axis unwinding. When the signal quality factor falls below the preset lower limit threshold, the threshold is directly set to negative infinity to forcibly block the time axis unwinding of the current channel from the physical bottom layer and force it to be handled by the cross-channel spatial verification module. This dynamic threshold adaptive mechanism enables the hierarchical unwinding strategy of this invention to maintain accurate trigger judgment under complex conditions such as low signal-to-noise ratio and coherent fading, avoid noise amplification and false repair, and effectively ensure the accuracy and stability of the unwinding process under all working conditions. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the method for recovering DAS high slew rate vibration signals based on adaptive multi-level unwinding according to the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention provides a method for recovering DAS (Dynamic Amplitude Seismic) vibration signals with high oscillation rates based on adaptive multi-level unwinding. This method uses a graded triggering mechanism to adaptively select unwinding strategies of different depths according to the severity of the vibration signal, thereby achieving accurate and robust phase recovery under complex conditions such as high oscillation rates, low signal-to-noise ratios, and coherent fading. Figure 1 As shown, a specific embodiment of the present invention includes the following steps: Step 1: Obtain the phase signal of the demodulated output of the DAS system and calculate the instantaneous slew rate of the phase signal.

[0021] In a DAS system, when external vibrations act on the optical fiber, the phase of the backscattered Rayleigh light changes accordingly. The system extracts the phase component of the analytical signal through coherent I / Q quadrature demodulation. However, due to the periodicity of the arctangent function, the directly demodulated phase is wrapped within the principal value range of [-π, π), denoted as the wrapped phase. This step first obtains the phase signal demodulated from the DAS system. This phase signal is the current sampling time. The original wrapped phase value of the demodulated output.

[0022] Then, the system calculates the instantaneous slew rate at the current sampling moment in real time. The calculation formula is as follows: ; in, The current sampling time Instantaneous slew rate, This is the phase value output at the current sampling time. This represents the phase value output at the previous sampling moment. The physical meaning of the instantaneous slew rate is the rate of phase change between two adjacent sampling points, and its magnitude directly reflects the intensity of external vibration: the larger the slew rate, the more intense the vibration, and the more likely the true phase difference between adjacent sampling points is to exceed the linear correction interval [-π, π] of the traditional single unwinding algorithm.

[0023] In addition, the system also calculates the signal quality factor in real time. The signal quality factor is used to assess the reliability of the current phase data. The formula for calculating the signal quality factor is: ; ; in, for Signal quality factor at time 10:00 The envelope amplitude of the signal analyzed by the DAS system. The background noise amplitude threshold set for the system or the historical average amplitude obtained a priori; The in-phase component signal at the current moment. The current orthogonal component signal; DAS systems inherently exhibit coherent fading: when local strain causes a specific change in the relative position of the Rayleigh scattering centers within the fiber, the amplitude of the interference light drops sharply or is even submerged by the system noise floor. In this case, the demodulated phase is pure noise, manifesting as irregular high-frequency, high-slew rate jumps. When When the envelope amplitude falls into the noise floor region, coherent fading is determined to have occurred. At this point, the demodulated phase data is unreliable, and the time-axis unwinding process should be skipped. The current channel should be marked as requiring spatial auxiliary processing and directly handed over to subsequent multiple unwinding modules for cross-channel spatial verification. The system can effectively distinguish whether the current phase jump originates from genuine high-intensity external disturbances to the fiber or from low signal-to-noise ratio degradation caused by coherent fading (i.e., false high-swing-rate signals), thus avoiding misinterpreting noise signals as real vibrations.

[0024] Step 2: Set the first threshold and the second threshold, and select the corresponding unwinding level based on the comparison result between the instantaneous oscillation rate and the preset threshold.

[0025] In a preferred embodiment, these two thresholds are not fixed constants, but are determined based on the system signal-to-noise ratio. and signal quality factor The specific adjustment method for the dynamic threshold that is adjusted in real time will be explained in detail in subsequent steps.

[0026] The system makes hierarchical decisions based on the following logic: 1) Unwinding in one step: like If the current signal is determined to be in a stable state with relatively gentle phase changes, a single unwinding operation is performed to fundamentally correct the phase of the current sampling point. This branch corresponds to normal operating conditions where vibration is minimal and the true phase difference between adjacent sampling points does not exceed π. Unwinding can be completed using a single-stage process with minimal computational load. The method for performing a single unwinding operation is as follows: Wrap phase of the current sampling point Overlay The compensation amount yields the phase after one untangling. : ; ; in, This is the current sampling point number. It is an integer. The value of makes Phase after unwrapping from the previous sampling point Minimize the absolute value of the difference between them; by selecting the integer that minimizes the phase difference between adjacent points. To ensure that the distance between adjacent sampling points does not exceed Phase jump correction.

[0027] 2) Secondary untangling: like If the current signal slew rate is determined to be high, it indicates that the external vibration is relatively severe, and the true phase difference between adjacent sampling points may be close to or slightly exceed π. In this case, a single basic correction cannot accurately identify the winding order, and the system automatically triggers the secondary unwinding module. Secondary unwinding establishes a local prediction model based on the phase change trend of historical sampling points, and uses the prediction results to help determine the optimal phase compensation order for the current sampling point. This branch corresponds to moderate-intensity vibration conditions, and compensates for the shortcomings of the basic algorithm by performing predictive extrapolation in the time dimension. The secondary unwinding method includes the following steps: Using the previously correctly unwrapped data at the current sampling point A local prediction model is established based on historical sampling points to calculate the phase prediction value of the current sampling point. The specific prediction model used can be one of the following two: Assuming that the rate of phase change (velocity) caused by vibration remains constant within a very short time, the current phase prediction value equals the phase at the previous moment plus the phase increment at the previous moment. In this case, a first-order linear prediction model (constant velocity model) can be used. The predicted value of the current sampling point is calculated using the phase of the first two sampling points: ; in, For the first The phase value of each sampling point after unwinding. For the first The phase values ​​of each sampling point after unwinding. The constant velocity model is suitable for scenarios where the instantaneous oscillation rate is in a low range and the phase change is relatively gentle. Its calculation is simple and can effectively correct the unwinding phenomenon caused by slight over-π.

[0028] When the external vibration frequency is high and the signal is in the transition region between a peak and a trough, the constant velocity assumption no longer holds, and the phase change rate itself also changes, requiring the introduction of an acceleration term (i.e., the second-order difference of the phase change). In this case, a second-order parabolic prediction model (constant acceleration model) can be used. An acceleration term is introduced, and the predicted value of the current sampling point is calculated using the phase of the previous three sampling points: ; After unfolding, we get: ; in, For the first Phase increment at time, For the first The second-order phase difference at time t; For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding.

[0029] By introducing a second-order difference term, the constant acceleration model can better track large slew rate turning signals at the peak or trough. When the vibration signal reaches the peak, the velocity drops to zero but the acceleration is not zero. Using only the constant velocity model will produce a large prediction error, while the constant acceleration model can accurately capture this turning trend, thus maintaining high prediction accuracy even in extreme cases of large slew rate.

[0030] Calculate the package phase of the current sampling point With the phase prediction value The predicted residuals between : ); in, The value is an integer representing the candidate phase compensation order; the smaller the residual, the closer the candidate compensation order is to the true value. By finding the predicted residual smallest integer Determine the optimal compensation order : ; in, This indicates rounding to the nearest integer. Due to noise and prediction errors, this ratio is usually a decimal. By rounding to the nearest integer, the optimal integer compensation order corresponding to the minimum residual can be obtained.

[0031] The optimal compensation order Substitute the values ​​to complete the secondary untangling compensation: ; in, This is the phase value output after two unwinding cycles.

[0032] At this point, the output phase value has crossed the ±π limit and can accurately restore the true phase. Even if the true phase difference between adjacent sampling points exceeds π, as long as the prediction model is accurate enough, the system can correctly identify the missing multiples of 2kπ, thereby effectively solving the problem of signal distortion caused by missed wrapping in traditional single-shot algorithms.

[0033] 3) Untangling multiple times: like If the current signal slew rate is determined to have reached its limit, it indicates extremely severe external vibrations and a true phase difference between adjacent sampling points far exceeding π. The information from a single-point time series is insufficient to determine the correct winding order, and the system automatically triggers a multiple unwinding module. Multiple unwinding utilizes the spatial phase continuity of adjacent sensing channels for cross-channel verification, using the correct phase information from spatial neighbors to determine the missing phase compensation order for the current channel. This branch corresponds to extreme large-signal operating conditions, introducing redundant information in the spatial dimension to compensate for the lack of information in the temporal dimension.

[0034] The principle of multiple unwinding is that the strain field of mechanical vibration along the fiber optic axis has spatial continuity, and DAS systems typically have extremely high spatial resampling rates (e.g., nominal spatial resolution of 10m, but spatial sampling interval of only 0.4m). This means that even if the current channel has an extremely high phase swirl on the time axis, causing the time prediction model to fail, its spatial phase difference (spatial gradient) with adjacent channels at the same moment is usually still much smaller than π. Therefore, the correct phase information of adjacent reference channels can be used to help recover the true phase of the current target channel.

[0035] Before performing cross-channel verification, dynamic optimization and verification of the reference channel are required first: Calculate the target channel Adjacent channels and Each of the signal quality factors at the current moment With instantaneous slew rate ; From adjacent channels and The following channels are selected as reference channels: (No severe coherent fading occurred) and (No unwinding error occurred); in, This is a preset lower limit threshold for signal quality. For adjacent channels in Signal quality factor at time 10:00 For adjacent channels in The instantaneous oscillation rate at any given moment.

[0036] Through the above verification steps, the system can ensure the data quality of the reference channel itself and avoid error propagation due to the same problem in the reference channel.

[0037] After selecting a reliable reference channel, perform cross-channel spatial assisted unwinding according to the following steps: Extract the previous time step Target Channel With adjacent reference channels Spatial phase gradient between : ; in, For the first Time target channel The true phase value after unwinding. For the first Time Reference Channel The true phase value after unwinding; due to the spatial distribution of the external vibration wave in an extremely short time step (from arrive The spatial phase gradient does not change abruptly within a time interval (usually on the order of microseconds), so the spatial phase gradient is gradually changing and the spatial gradient of the previous time interval can be used as a reasonable approximation of the spatial gradient of the current time interval.

[0038] Using the current time reference channel True phase after untangling Superimposed with the spatial phase gradient of the previous moment Calculate the target channel Spatial expected phase at the current moment : ; By utilizing the physical constraint of gradually varying spatial gradients, the correct phase value of the reference channel at the current moment is translated along the spatial direction to the target channel position, thereby extrapolating the expected phase that the target channel should have at the current moment. Since the phase of the reference channel at the current moment is a true continuous phase (already correctly unwrapped), and the historical spatial gradient reflects the inherent spatial phase difference between the target channel and the reference channel, the superposition of the two yields a reliable expected value.

[0039] Build target channel Candidate solution set of true phase And calculate the phase of each candidate solution with the expected spatial phase. The residual is used to determine the optimal compensation order by finding the minimum residual. : ; in, This indicates the rounding operation. The wrapper phase of the demodulated output of the target channel at the current moment; The optimal compensation order Substitute the values ​​and perform multiple untangling compensation steps: ; in, This is the phase value output by the target channel after multiple unwinding processes.

[0040] Even in extreme cases where the time dimension information is completely insufficient to determine the winding order, the system can still complete the phase recovery of the target channel by using the correct phase information of spatially adjacent channels, which significantly improves the robustness of the DAS system under severe vibration scenarios.

[0041] Through the above three-level diversion mechanism, the system only enters the computationally intensive high-order unwinding module when needed. While ensuring accurate unwinding under high slew rate conditions, it minimizes the average computational load of the system and ensures real-time processing capability.

[0042] Step 3: Dynamically adjust the first and second thresholds.

[0043] The first threshold in this embodiment Second threshold It is not a fixed constant, but a dynamic threshold that is adjusted in real time based on the system signal-to-noise ratio and signal quality factor. In a DAS system, when the signal-to-noise ratio... When signal degradation or slight coherent fading occurs, the random phase noise of the signal itself increases. If the unwinding trigger threshold remains fixed, this noise will be mistaken for high-rate oscillations, thus incorrectly triggering secondary or even multiple unwinding events, causing the noise to be amplified exponentially and severely degrading system performance. Therefore, the core logic of dynamic thresholding is: the more severe the environment (lower signal-to-noise ratio, smaller signal quality factor), the higher the threshold for triggering higher-order unwinding should be to prevent false triggering. The specific adjustment method for dynamic thresholding is as follows: Calculate signal quality factor : ; ; in, for Signal quality factor at time 10:00 The envelope amplitude of the signal analyzed by the DAS system. The background noise amplitude threshold set for the system; The in-phase component signal at the current moment. The current time is the orthogonal component signal; The first threshold and the second threshold Adaptive adjustment is performed according to the following formula: ; in, express The time adjustment A threshold, ; For the system under fading-free, high signal-to-noise ratio conditions, the first A baseline threshold constant (e.g., can be set) , ); This represents the short-time signal-to-noise ratio within the current sliding time window. This is a system reference signal-to-noise ratio constant (e.g., the normal signal-to-noise ratio of 20dB as specified by the system manufacturer). When the actual... Much larger When the exponential term approaches 0, the threshold is unaffected by the signal-to-noise ratio; The preset lower limit threshold for signal quality, when When this occurs, it indicates that severe coherent fading has occurred in the current channel, and the phase data is completely unreliable. At this time, the threshold is set to negative infinity to forcibly block the time axis unwinding of the current channel and force it to be handed over to the cross-channel spatial verification processing of the multiple unwinding module. and The adaptive adjustment coefficient is greater than zero. Controlling the threshold rise at low signal-to-noise ratios The nonlinear penalty force of the threshold when the control signal quality deteriorates.

[0044] When the signal-to-noise ratio When it decreases, The term increases, prompting the threshold to... Dynamic lifting; this is equivalent to the system automatically tightening the triggering conditions of higher-order unwinding in harsh environments, avoiding misjudging high-frequency background noise as a large vibration signal, thereby preventing noise from being amplified incorrectly.

[0045] When slight coherent fading occurs... Decrease (but not below) )hour, The exponential function increases sharply, forcing the threshold to rise significantly, further tightening the triggering conditions, so that the failed phase data does not participate in the time axis unwinding.

[0046] when Break At that time, the threshold is directly set to negative infinity, and the unwinding of the time axis of the current channel is completely disabled, forcing it to be handled by spatial auxiliary processing.

[0047] By employing a dynamic threshold, the hierarchical unwinding strategy of this invention can maintain accurate triggering judgment under complex operating conditions such as low signal-to-noise ratio and coherent fading, effectively ensuring the accuracy and stability of the unwinding process under all operating conditions.

[0048] Step 4: Output the unwound and recovered continuous phase signal.

[0049] After the corresponding levels of unwinding processing in the preceding steps, the system has determined and superimposed the correct 2kπ compensation amount on the phase signal at the current sampling time, so that the wrapped phase is restored to the true continuous phase. The system outputs the true phase value of each level of unwinding as the final result.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding, characterized in that, Includes the following steps: Acquire the phase signal of the demodulated output of the DAS system and calculate the instantaneous slew rate of the phase signal. ; Set the first threshold. Second threshold Based on the comparison between the instantaneous slew rate and a preset threshold, the appropriate unwinding level is selected to process the phase signal at the current sampling moment: like Perform one unwinding operation to make a basic correction to the phase at the current moment; like This triggers secondary unwinding, establishes a local prediction model based on the phase change trend of historical sampling points, and uses the prediction results to help determine the optimal phase compensation order. like This triggers multiple unwinding processes, utilizes the spatial phase continuity of adjacent sensing channels to perform cross-channel verification, and determines the missing phase compensation order of the current channel. Outputs a continuous phase signal recovered after unwinding.

2. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 1, characterized in that, The formula for calculating the instantaneous slew rate of a phase signal is: ； in, The current sampling time Instantaneous slew rate, This is the phase value output at the current sampling time. This is the phase value output at the previous sampling time.

3. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 1, characterized in that, The method for performing one untangling operation is as follows: Wrap phase of the current sampling point Overlay The compensation amount yields the phase after one untangling. : ； ； in, This is the current sampling point number. It is an integer. The value of makes Phase after unwrapping from the previous sampling point Minimize the absolute value of the difference between them; by selecting the integer that minimizes the phase difference between adjacent points. To ensure that the distance between adjacent sampling points does not exceed Phase jump correction.

4. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 1, characterized in that, The secondary untangling method includes the following steps: Using the previously correctly unwrapped data at the current sampling point A local prediction model is established based on historical sampling points to calculate the phase prediction value of the current sampling point. ; Calculate the wrap phase of the current sampling point With the phase prediction value The predicted residuals between : )； in, is an integer representing the candidate phase compensation order; By finding the predicted residual smallest integer Determine the optimal compensation order : ； in, This indicates the rounding operation; The phase value after the second unwinding is: ； in, This is the phase value output after two unwinding cycles.

5. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 4, characterized in that, When using a first-order linear prediction model, take The predicted value of the current sampling point is calculated using the phase of the first two sampling points: ； in, For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding.

6. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 4, characterized in that, When using a second-order parabolic prediction model, take An acceleration term is introduced, and the predicted value of the current sampling point is calculated using the phase of the previous three sampling points: ； After unfolding, we get: ； in, For the first Phase increment at time, For the first The second-order phase difference at time t; For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding. For the first The phase value of each sampling point after unwinding.

7. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 1, characterized in that, The method of untangling multiple times includes the following steps: Extract the target channel from the previous time step With adjacent reference channels Spatial phase gradient between : ； in, For the first Time target channel The true phase value after unwinding. For the first Time Reference Channel The true phase value after unwinding; Using the current time reference channel True phase after untangling Superimposed with the spatial phase gradient of the previous moment Calculate the target channel Spatial expected phase at the current moment : ； Build target channel The set of candidate solutions for the true phase And calculate the phase of each candidate solution with the expected spatial phase. The residual is used to determine the optimal compensation order by finding the minimum residual. : ； in, This indicates the rounding operation. The wrapper phase of the demodulated output of the target channel at the current moment; The phase value after multiple unwindings is: ； in, This is the phase value output by the target channel after multiple unwinding processes.

8. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 7, characterized in that, Before performing cross-channel verification, a dynamic optimization and verification step for the reference channel is also included: Calculate the target channel Adjacent channels and Each of the signal quality factors at the current moment With instantaneous slew rate ; From adjacent channels and The following channels are selected as reference channels: and ; in, This is a preset lower limit threshold for signal quality. For adjacent channels in Signal quality factor at time 10:00 For adjacent channels in The instantaneous oscillation rate at any given moment.

9. The method for recovering DAS large oscillation rate vibration signals based on adaptive multi-level unwinding as described in claim 1, characterized in that, The first threshold Second threshold Based on dynamic adjustment of the system signal-to-noise ratio and signal quality factor, the dynamic adjustment method is as follows: Calculate signal quality factor : ； ； in, for Signal quality factor at time 10:00 The envelope amplitude of the signal analyzed by the DAS system. The background noise amplitude threshold set for the system; The in-phase component signal at the current moment. The current time is the orthogonal component signal; The first threshold and the second threshold Adaptive adjustment is performed according to the following formula: ； in, express The time adjustment A threshold, ; For the system under fading-free, high signal-to-noise ratio conditions, the first One benchmark threshold constant; This represents the short-time signal-to-noise ratio within the current sliding time window. This serves as the system's reference signal-to-noise ratio constant. The preset lower limit threshold for signal quality, when At this time, the threshold is set to negative infinity, forcibly blocking the time axis unwinding of the current channel; and The adaptive adjustment coefficient is greater than zero. Controlling the threshold rise at low signal-to-noise ratios The nonlinear penalty force of the threshold when the control signal quality deteriorates.

Images