Seabed crustal activity monitoring method based on differential phase sensitive optical time domain reflectometer

Abstract

The invention relates to the technical field of optical signal processing and analysis, and particularly provides a seabed crustal activity monitoring method based on a differential phase sensitive optical time domain reflectometer, which comprises the following steps: carrying out differential processing on an echo optical pulse signal to obtain a first differential echo matrix; and setting a dynamic window constraint to obtain a second differential echo matrix. The second differential echo matrix includes a low rank term, a sparse term, and a noise term. And constructing a global optimization problem according to the second differential echo matrix, and solving a sparse item estimation matrix corresponding to the sparse item. And separating irrelevant activities in the sparse item estimation matrix to obtain a target activity matrix, and then obtaining a monitoring result. According to the method, non-low-rank trend features such as slow-varying interference are converted into low-rank features through differential processing, periodic and regular components in received signals are clearer, periodic events and sudden events can be separated through a low-rank sparse decomposition model, human-related activities are further separated, and the accuracy of the received signals is improved. Accurate monitoring of the sudden movement of the seabed crust is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of optical signal processing and analysis, and particularly to a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer. Background Art

[0002] When an external event acts on an optical fiber, it will cause slight changes in the refractive index, length, etc. of the optical fiber, thereby changing the phase. The Phase-sensitive Optical Time Domain Reflectometer (phase OTDR) technology can measure and determine the occurrence location of an external event based on the phase change of a signal, so it is widely used in geological monitoring.

[0003] In the prior art, the phase OTDR technology is often used to monitor submarine plate activities. Submarine plate activities include periodic events and sudden events. Among them, sudden submarine plate activities often cause serious natural disasters such as earthquakes and tsunamis, which need to be closely monitored. However, the existing monitoring methods are affected by trend characteristics such as slow-varying interference in the marine environment, unable to separate periodic events, difficult to obtain accurate disaster warnings, with a high false alarm rate and insufficient reliability. Summary of the Invention

[0004] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by the embodiments of the present invention can eliminate slow-varying interference and separate irrelevant noise, and at least solve the problem of insufficient accuracy in monitoring submarine crustal activities.

[0005] In a first aspect, the embodiments of the present invention provide a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer, and the method includes: Obtain a backscattered optical pulse signal, and perform differential processing on the backscattered optical pulse signal to obtain a first differential echo matrix; Set a dynamic window to obtain a second differential echo matrix of the first differential echo matrix under the constraint of the dynamic window; wherein, the second differential echo matrix includes a low-rank term, a sparse term, and a noise term; Construct a global optimization problem according to the second differential echo matrix, and solve the sparse term estimation matrix corresponding to the sparse term in the global optimization problem; Separate the irrelevant activities in the sparse term estimation matrix to obtain a target activity matrix; Obtain the monitoring result of the target activity according to the target activity matrix.

[0006] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention for creating an invention obtains an echo optical pulse signal, and the differential processing of the echo optical pulse signal to obtain a first differential echo matrix includes: Receiving the echo optical pulse signal through a spectral analyzer, and obtaining a sampled Rayleigh curve corresponding to each echo optical pulse signal; Performing differential processing on two adjacent sampled Rayleigh curves to obtain a first differential signal; The first differential signals form the first differential echo matrix.

[0007] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention for creating an invention sets a dynamic window, and obtaining a second differential echo matrix of the first differential echo matrix under the constraint of the dynamic window includes: Obtaining a corresponding signal main frequency according to the sampled Rayleigh curve of the echo optical pulse signal; Setting a window length of the dynamic window according to the signal main frequency; wherein, the dynamic window, the signal main frequency, and the sampled Rayleigh curve are in one-to-one correspondence; Calculating differential signal components of the first differential signal at each sampling point one by one; wherein, the differential signal components are calculated within the length range of the dynamic window; Arranging the calculated differential signal components to obtain the second differential signal; The second differential signals form the second differential echo matrix.

[0008] Before constructing a global optimization problem according to the second differential echo matrix by the method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention for creating an invention, the method further includes: Setting a weight coefficient of the low-rank term according to the dynamic window; Calculating a weight coefficient of the sparse term according to the weight coefficient of the low-rank term; wherein, the sum of the weight coefficient of the low-rank term and the weight coefficient of the sparse term is 1.

[0009] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention for creating an invention setting the weight coefficient of the low-rank term according to the dynamic window includes: Calculating an average value of the window length of the dynamic window; Calculating a standard deviation of the window length of the dynamic window according to the average value; Creating a local weight function according to the average value and the standard deviation; wherein, the local weight function is set as a piecewise function, and the segmentation points of the local weight function are determined according to the average value and the standard deviation; Calculate the local weight corresponding to each of the dynamic windows according to the local weight function; Take the average of the local weights as the weight coefficient of the low-rank term.

[0010] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention creates a global optimization problem according to the second differential echo matrix, including: Perform singular value decomposition on the low-rank term, and sum the singular values to obtain the nuclear norm of the low-rank term; Take the 1-norm of the elements in the sparse term; Perform weighted summation on the nuclear norm of the low-rank term and the 1-norm of the sparse term to obtain the objective function of the global optimization problem; Calculate the F-norm of the noise term and square it, and limit the squared result to be less than or equal to a noise suppression threshold as a constraint condition of the global optimization problem; wherein, the noise suppression threshold is preset; Obtain the global optimization problem according to the objective function and the constraint condition.

[0011] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention creates a target activity matrix by separating irrelevant activities in the sparse term estimation matrix, including: Set a time-domain target activity set, perform time-dimensional extraction on the sparse term estimation matrix, and classify the columns corresponding to the time-domain target activities in the sparse term estimation matrix into the time-domain target activity set; Set a spatial-domain target activity set, perform spatial-dimensional extraction on the sparse term estimation matrix, and classify the columns corresponding to the spatial-domain target activities in the sparse term estimation matrix into the spatial-domain target activity set; Set the columns corresponding to the irrelevant activities in the sparse term estimation matrix to zero to obtain the target activity matrix; wherein, the columns corresponding to the irrelevant activities in the sparse term estimation matrix are the columns that are neither classified into the time-domain target activity set nor classified into the spatial-domain target activity set.

[0012] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by an embodiment of the present invention creates a time-domain target activity set by performing time-dimensional extraction on the sparse term estimation matrix and classifying the columns corresponding to the time-domain target activities in the sparse term estimation matrix into the time-domain target activity set, including: Set a sliding window for the elements of the sparse term estimation matrix; wherein, the sliding window is in the same row as the element in the sparse term estimation matrix, the sliding window is centered on the element, and the length of the sliding window is preset; Calculate the average value of all elements within the sliding window; For the element at the center of the sliding window, calculate the difference between it and the average value of all elements within the sliding window and take the two-norm to obtain the smoothing error of the element; Compare the smoothing error and the smoothness threshold of each element of the sparse term estimation matrix respectively; wherein, the smoothness threshold is preset; When the smoothing error of any element is less than the smoothness threshold, classify the column where the element is located into the time-domain target activity set.

[0013] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time-domain reflectometer provided by an embodiment of the present invention creates a spatial dimension extraction for the sparse term estimation matrix, and classifies the columns corresponding to the spatial domain target activities of the sparse term estimation matrix into the spatial domain target activity set, including: Calculate the sum of the absolute values of the column elements for each column of the sparse term estimation matrix respectively as the sparsity parameter; Compare the sparsity parameter and the sparsity threshold of each column of the sparse term estimation matrix respectively; wherein, the sparsity threshold is preset; When the sparsity parameter of any column is greater than the sparsity threshold, classify the corresponding column into the spatial domain target activity set.

[0014] In a second aspect, an embodiment of the present invention also provides an electronic device, including: a processor, and a memory storing a program, the program including instructions that, when executed by the processor, cause the processor to execute the method for monitoring submarine crustal activities based on a differential phase-sensitive optical time-domain reflectometer provided in any of the above embodiments.

[0015] The method for monitoring submarine crustal activities based on a differential phase-sensitive optical time-domain reflectometer provided by an embodiment of the present invention creates converts non-low-rank features such as slow-varying interference, high-dimensional complex structures, and long-term trends in the received signal into low-rank features through differential signal processing, making the periodic and regular components in the received signal clearer. It can separate periodic events and sudden events through a low-rank sparse decomposition model, and further separate human-unrelated activities in sudden events, realizing precise monitoring of sudden submarine crustal activities. The monitoring results have good accuracy and robustness. Further analysis shows that the false alarm rate of natural disaster early warnings is lower and the reliability is higher. Description of the Drawings

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other embodiments can be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of the method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer in the embodiments of the present invention.

[0018] Figure 2 It is a schematic structural diagram of the electronic device of the present invention. Detailed implementation manners

[0019] The embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not used to limit the protection scope of the present invention.

[0020] Optical Time Domain Reflectometer (OTDR) technology can be classified according to the measurement principle and application scenarios. Among them, Phase-sensitive Optical Time Domain Reflectometer (phase OTDR) technology is a distributed optical fiber sensing technology based on Rayleigh scattering. When factors such as vibration, strain, and temperature act on the optical fiber, they will cause slight changes in the refractive index and length of the optical fiber, thereby changing the phase. The phase OTDR technology detects the phase change, obtains the parameter change information along the optical fiber, and realizes the detection of the vibration or strain position. In addition, the phase OTDR technology also has advantages such as long-distance and distributed monitoring, and is widely used in geological monitoring.

[0021] Submarine crustal activities can trigger natural disasters such as earthquakes and tsunamis and a series of environmental changes, and are interrelated with the land crust. Therefore, the monitoring of small submarine crustal activities is of great significance. The phase OTDR technology can capture the small deformations of the geology by detecting the phase change of the optical fiber echo signal, monitor the activities of the submarine crust, and provide early warnings for natural disasters. However, the echo signal contains a large amount of noise and slow-varying interference, which affects the monitoring process and results in a high false alarm rate.

[0022] Therefore, the embodiments of the present invention provide a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer. By analyzing the backscattered optical pulse signal based on the phase OTDR technology, through differential signal processing, construction and solution of the global optimization problem, a large number of irrelevant impurities in the signal are separated, and the sudden activities of the submarine crust are accurately monitored to provide early warnings for natural disasters such as tsunamis and earthquakes.

[0023] It should be noted that according to the monitoring principle of phase OTDR technology, events that may cause fiber optic phase changes in the seabed environment will all exist in the backscattered optical pulse signal. Events that may cause vibration, strain or temperature changes, resulting in fiber optic phase changes, include periodic events, sudden events and interference events.

[0024] Among them, periodic events repeat at regular time intervals and have relatively stable periods and rules. For example, the periodic activities of marine organisms and the seabed crustal activities caused by tides, etc. Sudden events are different from periodic events. They do not have stable rules and cannot be predicted, and they are sudden and uncertain. Sudden events include human activities and natural activities. Human activities include signal changes caused by human factors such as man-made faults in submarine communication systems, vessel fishing activities, and seabed excavation. Natural activities are sudden movements of seabed plates, including fractures and dislocations of the seabed crust and changes in seabed topography, etc., which may trigger natural disasters such as tsunamis and earthquakes.

[0025] The monitoring method provided by the embodiments of the present invention focuses on natural activities in seabed sudden events, that is, sudden movements of seabed crustal plates, and uses natural activities that may trigger natural disasters as target activities for monitoring. This monitoring method can exclude interference, separate periodic events and sudden events, and then separate human activities in sudden events, and finally accurately warn of possible natural disasters, with high reliability.

[0026] Specifically, the monitoring method provided by the embodiments of the present invention includes the following steps.

[0027] Step S100, obtain the backscattered optical pulse signal, and perform differential processing on the backscattered optical pulse signal to obtain the first differential echo matrix.

[0028] Step S100 includes: Step S110, in a distributed fiber optic sensing system based on phase OTDR, receive the backscattered optical pulse signal through an optical spectrum analyzer. Obtain the sampled Rayleigh curve corresponding to each backscattered optical pulse signal.

[0029] In the embodiments of the present invention, a total of P + 1 backscattered optical pulse signals are received during monitoring. Each backscattered optical pulse signal corresponds to a monitoring moment. Set the initial monitoring moment as t0. When the light source emits optical pulses, the emission interval between adjacent optical pulses is . For the pth backscattered optical pulse signal received, the corresponding monitoring moment is t0 + . Wherein, p is an integer, and 1 ≤ p ≤ P + 1.

[0030] Each reflected optical pulse signal corresponds to a sampled Rayleigh curve. The sampled Rayleigh curve contains several different positions x on the optical fiber, that is, different sampling points. The sampled Rayleigh curve corresponding to the p-th reflected optical pulse signal is 。

[0031] The matrix formed by the sampled Rayleigh curves corresponding to the received P + 1 reflected optical pulse signals is the reflected optical pulse matrix R. The reflected optical pulse matrix R is expressed as: 。

[0032] Step S120, perform a difference processing on two adjacent sampled Rayleigh curves in the reflected optical pulse matrix R to obtain a first difference signal. The difference processing process is calculated with reference to the following formula: 。

[0033] In the formula, is the p-th first difference signal.

[0034] Step S130, form the first difference echo matrix D with the first difference signals in step S102. The first difference echo matrix D is as follows: 。

[0035] As shown above, events that may cause phase changes in the optical fiber in the marine environment include periodic events, sudden events, and interference events. Among them, periodic events make the data corresponding to the received signal show a low-rank structure. In related geological monitoring, a low-rank sparse decomposition model is established based on the low-rank characteristics of periodic events to separate periodic events and identify sudden events.

[0036] However, in the marine environment, interference events include a large amount of noise and slow-varying interference. Slow-varying interference includes temperature changes and crustal trend movements, etc. Different from the regular changes of periodic events and the accidental characteristics of sudden events, slow-varying interference usually changes slowly over time to form a certain trend, rather than a simple periodic component. Therefore, slow-varying interference such as temperature changes does not have low-rank characteristics in the received signal, and also blurs the low-rank characteristics of the received signal, thereby making the low-rank sparse decomposition model unable to effectively separate and identify, and the false alarm rate is relatively high.

[0037] The monitoring method provided by the embodiments of the present invention can eliminate various slow-varying interferences typified by temperature drift through differential processing of the sampled Rayleigh curve. Non-low-rank features such as slow-varying interferences, high-dimensional complex structures, and long-term trends in the received backlight pulse signals are transformed into low-rank features. The non-low-rank features that originally change continuously over time no longer dominate the backlight pulse matrix R. After differential operation, the periodic components with low-rank characteristics dominate the first-difference echo matrix D. Differential processing can better present the periodic structure of the received signal and exclude the blurring and influence of trend components such as slow-varying interferences on the data structure.

[0038] Further, in the embodiments of the present invention, as shown in the following formula, the first-difference echo matrix D is disassembled into a low-rank term L, a sparse term S, and a noise term N.

[0039] 。

[0040] In the formula, the low-rank term L corresponds to the periodic component after the received signal eliminates trend features such as slow-varying interferences, that is, it corresponds to periodic events in the marine environment. The sparse term S corresponds to sudden events, that is, it includes human activities and natural activities. After separating human activities from the sparse term S, the target activity required for monitoring is obtained. The target activity can reflect the sudden activities of the seabed crust and is of great significance for early warning of natural disasters. The noise term N corresponds to other noises.

[0041] Step S200: Set a dynamic window and use the dynamic window alignment method to obtain the second-difference echo matrix of the first-difference echo matrix D under the constraint of the dynamic window.

[0042] As an implementable manner, step S200 includes the following steps: Step S210: Use the frequency analysis method to estimate the frequency characteristics of the backlight pulse signal. The corresponding signal main frequency is obtained through the sampled Rayleigh curve of the backlight pulse signal. For the th received backlight pulse signal, the corresponding signal main frequency is 。The signal main frequency is calculated with reference to the following formula: 。

[0043] In the formula, is the sampled Rayleigh curve corresponding to the pth backlight pulse signal. represents the Fourier transform of. represents the energy distribution of the spectrum.

[0044] By obtaining the signal main frequency , it is possible to, based on the signal main frequency Compare the size with the empirical threshold to determine whether the reflected light pulse signal corresponds to low-frequency crustal activity or high-frequency sudden crustal activity. After adjusting the dynamic window according to the main frequency of the signal, the constrained second-difference echo matrix can adapt to crustal activities of different frequencies, thereby improving the accuracy, sensitivity, reliability, and stability of the monitoring results.

[0045] Step S220, according to the main frequency of the signal obtained in step S210 Set the window length of the dynamic window. The window length of the dynamic window is calculated with reference to the following formula: .

[0046] In the formula, is the window length after the dynamic window is adjusted; is the first adjustment parameter; is the second adjustment parameter. When the main frequency of the signal is determined, the first adjustment parameter and the second adjustment parameter are used to comprehensively control the window length range of the dynamic window.

[0047] The first adjustment parameter is used to control the change range of the window length . In the same range of the main frequency of the signal , the larger the first adjustment parameter , the larger the change range of the window length of the dynamic window. The smaller the first adjustment parameter , the smaller the change range of the length of the dynamic window. In the embodiment of the present invention, the range of the first adjustment parameter is set to [10, 1000].

[0048] When the main frequency of the signal is higher than the empirical threshold, the corresponding reflected light pulse signal corresponds to high-frequency crustal activity. At this time has a small value. To avoid the window length from being too small, set the second adjustment parameter to ensure the minimum length of the dynamic window. In the embodiment of the present invention, the range of the second adjustment parameter is set to [100, 5000].

[0049] The main frequency of the signal is in one-to-one correspondence with the p-th reflected light pulse signal and the sampling Rayleigh curve . Each main frequency of the signal corresponds to a dynamic window, and the window length of the dynamic window is based on the main frequency of the signal Dynamically adjusted.

[0050] Step S230: Calculate the differential signal components of the first differential signal at each sampling point one by one; among them, the differential signal components are calculated within the length range of the dynamic window.

[0051] Under the constraint of the dynamic window, the differential signal component of the first differential signal at position x is calculated with reference to the following formula: .

[0052] In the formula, x is a sampling position on the optical fiber, that is, a sampling point. The range of i in the formula is restricted so that the operation is carried out within the window length of the dynamic window. The obtained by performing pulse alignment differential processing is the differential signal component corresponding to the sampling point x.

[0053] Step S240: Perform window constraint differential adjustment on all sampling points on the optical fiber as in Step S230. The differential signal component obtained in Step S230 is a single-point value corresponding to the sampling position x. After arranging the differential signal components corresponding to each sampling position in the order of the original sampling position, the second differential signal is obtained.

[0054] Step S250: Repeat the dynamic window setting and constraint differential adjustment in Steps S210 to S240 for each first differential signal. Each first differential signal obtains a corresponding second differential signal, and the corresponding second differential signals form the second differential echo matrix . The second differential echo matrix is expressed as: .

[0055] The second differential echo matrix is the processing result of the first differential echo matrix D under the constraint of the dynamic window, and is used for dynamic adjustment to adapt to crustal activity monitoring at different frequencies. Similar to the first differential echo matrix D, the second differential echo matrix is also decomposed into: including the low-rank term L, the sparse term S, and the noise term N.

[0056] Step S300: Construct a global optimization problem according to the second differential echo matrix and solve the sparse term estimation matrix corresponding to the sparse term S in the global optimization problem.

[0057] Before constructing the global optimization problem in step S300, the monitoring method provided by the embodiments of the present invention further includes setting weight coefficients for the low-rank term L and the sparse term S. The weight coefficients weight the low-rank term L and the sparse term S respectively when constructing the global optimization problem, which can prevent overfitting, facilitate fine-tuning of the global optimization problem, improve flexibility, and balance the contributions of the data low-rank characteristic and the data sparse characteristic in the global optimization problem.

[0058] The weight coefficients of the low-rank term L and the sparse term S are set according to the dynamic window length in step S200 In the embodiments of the present invention, first, according to the window length of the dynamic window calculate the weight coefficient of the low-rank term L, and then calculate the weight coefficient of the sparse term S according to the weight coefficient of the low-rank term L.

[0059] Specifically, first calculate the mean value of the window length according to the dynamic window in step S200.

[0060] Receive P + 1 backlight optical pulse signals, corresponding to P + 1 sampling Rayleigh curves, and then corresponding to P + 1 dynamic windows . The mean value of the window lengths of the P + 1 dynamic windows is: .

[0061] Calculate the variance of the window lengths of the P + 1 dynamic windows as: .

[0062] Then the standard deviation is σ W .

[0063] Subsequently, according to the mean value and the standard deviation σ W create a local weight function . Among them, the local weight function is set as a piecewise function, and the segmentation points of the piecewise function are determined according to the mean value and the standard deviation σ W . The local weight function is shown as follows: .

[0064] After inputting the window length of the dynamic window into the local weight function , output the local weight corresponding to this dynamic window.

[0065] As shown in the following formula, take the average value of the local weights output corresponding to the P + 1 dynamic windows as the weight coefficient of the low-rank term L .

[0066] 。

[0067] It should be noted that the segmentation point of the local weight function is set according to the mean value and the standard deviation σ W to set the interval. For the dynamic window corresponding to a backscattered optical pulse signal, when the length of the dynamic window is large and exceeds this interval, the local weight output by the local weight function is 1, indicating that the backscattered optical pulse signal corresponding to this dynamic window is more inclined to the periodic event dominated by the low-rank characteristic. Therefore, by using the local weight function the local weight is increased, and then the weight coefficient of the low-rank term L is increased

[0068] For the dynamic window corresponding to a backscattered optical pulse signal, when the length of the dynamic window is small and exceeds this interval, the local weight output by the local weight function is 0, indicating that the backscattered optical pulse signal corresponding to this dynamic window is more inclined to the sudden event dominated by the sparse characteristic. Therefore, by using the local weight function the local weight is reduced, and then the weight coefficient of the low-rank term L is controlled to be even lower

[0069] Setting the local weight function to calculate the weight coefficient of the low-rank term L can comprehensively consider the window lengths of each dynamic window, reflect the global distribution of the dynamic window as a whole, rationalize the low-rank and sparse characteristics in the data, balance the low-rank and sparse characteristics, and obtain a relatively accurate weight coefficient of the low-rank term L

[0070] Subsequently, according to the obtained weight coefficient of the low-rank term L the weight coefficient of the sparse term is calculated. Among them, the sum of the weight coefficient of the sparse term and the weight coefficient of the low-rank term L is 1. The weight coefficient of the sparse term

[0071] When classifying events under ideal conditions, the assumption is based on noiselessness, without considering the existence of noise, and only considering periodic events dominated by low-rank characteristics and bursty events dominated by sparse characteristics. When performing weighted processing, it is considered that all except the low-rank term L are sparse terms S, and all except the sparse term S are low-rank terms L. Therefore, in the embodiments of the present invention, the sum of the weight coefficients of the two is set to 1 to consider the balance between sparse characteristics and low-rank characteristics.

[0072] After calculating the weight coefficient of the low-rank term L and the sparse term weight coefficient of, as an implementable manner, step S300 specifically includes: Step S310, performing singular value decomposition on the low-rank term L and summing the singular values to obtain the nuclear norm of the low-rank term L . The nuclear norm of the low-rank term L is used to characterize the low-rank characteristics of the low-rank term L.

[0073] Taking the 1-norm of the elements in the sparse term S . The 1-norm is the maximum value of the sum of the absolute values of each column element in the matrix. The sparse term 1-norm is used to characterize the sparse characteristics of the sparse term S.

[0074] Step S320, using weighted summation of the nuclear norm of the low-rank term L and the 1-norm of the sparse term S to obtain the objective function of the global optimization problem. Among them, the weights during weighted summation are the weight coefficient of the low-rank term L calculated above and the weight coefficient of the sparse term S .

[0075] Taking the F-norm of the noise term N and squaring it, restricting the squared result to be less than or equal to the noise suppression threshold as the constraint condition of the global optimization problem. Among them, the noise suppression threshold is preset.

[0076] According to the objective function and the constraint condition, the global optimization problem is obtained as shown in the following formula: .

[0077] In the formula, is the F-norm, also known as the Frobenius norm, and is used to measure the reconstruction error.

[0078] Step S330, solving the sparse term estimation matrix corresponding to the sparse term S in the global optimization problem obtained in step S320.

[0079] Preferably, in the embodiments of the present invention, the Alternating Direction Method of Multipliers (ADMM) is used to solve the global optimization problem, and the sparse term estimation matrix corresponding to the sparse term S is obtained. The ADMM algorithm is an algorithm for solving optimization problems, which combines the ideas of the augmented Lagrangian function and the dual ascent algorithm, and has been widely used in many fields such as machine learning, signal processing, image processing, and statistical analysis.

[0080] Step S400: Separate the irrelevant activities in the sparse term estimation matrix to obtain the target activity matrix.

[0081] The sudden events corresponding to the sparse term include human activities and natural activities that can cause changes in the optical fiber phase. Among them, human activities are irrelevant activities and not the monitoring targets of submarine crustal activities. Natural activities are the target activities required for monitoring, which can reflect sudden submarine crustal activities and are of great significance for natural disaster early warning. Step S400 is used to separate human activities and natural activities in the sparse term estimation matrix.

[0082] As an implementable manner, step S400 includes: Step S410: Set the time-domain target activity set. Perform time-domain extraction on the sparse term estimation matrix obtained by solving step S330, and classify the columns corresponding to the time-domain target activities in the sparse term estimation matrix into the time-domain target activity set.

[0083] The time-domain target activity refers to the target activity obtained based on the time-domain decision condition when judging the target activity from the time-domain condition. The time-domain target activity set refers to the set of sparse Rayleigh spectra corresponding to the time-domain target activities.

[0084] Before performing the time-domain extraction, the time-domain target activity set is an empty set. Define the time-domain target activity set , and after classifying all the time-domain target activities into the time-domain target activity set, the setting of the time-domain target activity set is completed.

[0085] Specifically, set a sliding window for the element of the sparse term estimation matrix. Among them, the element is the element in the x-th row and p-th column of the sparse term estimation matrix. The x-th row is the optical fiber sampling position, and the p-th column corresponds to the p-th backscattered optical pulse signal, which is associated with the sampling time.

[0086] The sliding window is in the same row as the element in the sparse term estimation matrix. The sliding window is centered on the element , and the length of the sliding window is preset. Calculate the smoothing error of the row where the element is located through the sliding window. The calculation of the smoothing error refers to the following formula: 。

[0087] In the formula, S(x, p + j) is an element of the same row as other elements. By setting the value of j, S(x, p + j) is defined as an element within the sliding window of the same row. u t is a preset length, and there are a total of 2u +1 elements centered around t in the sliding window.

[0088] Calculate the average value of all elements within the sliding window. For the element at the center of the sliding window, calculate the difference between it and the average value of all elements within the sliding window and take the two-norm to obtain the smoothing error of the element 。

[0089] Compare the smoothing error and the smoothness threshold of each element in the sparse term estimation matrix respectively. Among them, the smoothness threshold is preset.

[0090] When the smoothing error of any element is less than the smoothness threshold , include the column where this element is located in the time-domain target activity set. For example, for the smoothing error of the element , if there exists , then the sparse component in the p-th sampled Rayleigh curve, that is, the p-th column in the sparse term estimation matrix is identified as the time-domain target activity , and is included in the time-domain target activity set 。

[0091] Step S420, set the spatial-domain target activity set, perform spatial dimension extraction on the sparse term estimation matrix obtained by solving in step S330, and include the columns corresponding to the spatial-domain target activity in the sparse term estimation matrix in the spatial-domain target activity set.

[0092] Spatial-domain target activity refers to the target activity obtained based on the spatial-domain decision condition when judging the target activity from the spatial-domain conditions. The spatial-domain target activity set refers to the set corresponding to the spatial-domain target activity.

[0093] Before performing the spatial dimension extraction, the spatial-domain target activity set is an empty set. Define the spatial-domain target activity set , after classifying all airspace target activities into the airspace target activity set, the setting of the airspace target activity set is completed. Steps S410 and S420 judge target activities from the time domain and airspace dimensions respectively. One can first set the time-domain target activity set and perform time dimension extraction, and then set the airspace target activity set and perform space dimension extraction. Or one can first set the airspace target activity set and perform space dimension extraction, and then set the time-domain target activity set and perform time dimension extraction. The execution order is not limited to this.

[0094] Specifically, for each column vector of the sparse term estimation matrix calculate the 1-norm respectively, that is, the sum of the absolute values of the elements in each column, as the sparsity parameter . The sparsity parameter is used to characterize sparsity and is calculated with reference to the following formula: .

[0095] Compare the sparsity parameters of each column of the sparse term estimation matrix and the sparsity threshold respectively. Among them, the sparsity threshold is preset.

[0096] When the sparsity parameter of any column is greater than the sparsity threshold, that is , the corresponding column is identified as an airspace target activity , and is classified into the airspace target activity set .

[0097] Step S430, reconstruct the sparse term estimation matrix according to the time-domain target activity set in step S410 and the airspace target activity set in step S420. Specifically, set the columns corresponding to irrelevant activities to zero to obtain the target activity matrix . The reconstruction process refers to the following formula: .

[0098] Retain the columns classified into the time-domain target activity set and the columns classified into the airspace target activity set. Set the columns that are neither classified into the time-domain target activity set nor classified into the airspace target activity set to zero. The columns that are neither classified into the time-domain target activity set nor classified into the airspace target activity set are the columns corresponding to irrelevant activities in the sparse term estimation matrix.

[0099] Step S400 separates the submarine sudden crustal activities of the target and the irrelevant human activities based on the time dimension and space dimension through the joint analysis of the time domain and airspace. The fusion of the spatio-temporal judgment results can improve the accuracy and robustness of target activity monitoring.

[0100] Step S500, obtain the monitoring result of the target activity according to the target activity matrix .

[0101] Preferably, in the embodiments of the present invention, the target activities in the target activity matrix can be further classified by sample training or building an empirical dictionary to output a monitoring result, and further analyze the monitoring result to give an early warning of natural disasters.

[0102] Sample training means pre-collecting a large amount of sample data related to target activities. When the target activity is a sudden crustal activity on the seabed, the sample data can include seabed topography data, geothermal flow data, physical and chemical property data of rock samples, etc. Clean and correct the large amount of collected data, and then extract features. For example, for submarine earthquakes, features such as amplitude, frequency, and waveform are extracted. Certain specific frequencies of seismic waves may be related to the types of crustal activities.

[0103] Further classify the target activities and select a suitable model, use the sample data with extracted features to train the selected model, evaluate, optimize and improve the model, and then further analyze the target activity matrix through the final model optimization result for further analysis.

[0104] Building an empirical dictionary is to collect information on sudden crustal activities on the seabed, extract typical feature patterns and descriptive information, and perform digital representation. Match and further classify the target activity matrix with the empirical dictionary, and regularly evaluate and screen the dictionary.

[0105] When further classifying the target activity matrix and outputting the monitoring result, in addition to sample training and building an empirical dictionary, other methods can also be selected according to the requirements of the monitoring result output, and it is not limited to this.

[0106] The method for monitoring seabed crustal activities based on a differential phase-sensitive optical time domain reflectometer provided by the embodiments of the present invention converts non-low-rank features such as slow-varying interference, high-dimensional complex structures, and long-term trends in the received signal into low-rank features through differential signal processing, and solves the technical problem that the recognition of target activities by related models cannot be applied to scenarios with slow-varying interference. Through differential signal processing, the periodic and regular components in the received signal become clearer, and the periodic events and sudden events can be separated by a low-rank sparse decomposition model, and further separate the human-unrelated activities in the sudden events, realizing the precise monitoring of sudden seabed crustal activities, and the monitoring result has better accuracy and robustness. The false alarm rate of the natural disaster early warning obtained by further analysis is lower and the reliability is higher.

[0107] An embodiment of the present invention also provides a non-transitory machine-readable medium storing a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to execute a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer.

[0108] An embodiment of the present invention also provides a computer program product, including a computer program, wherein the computer program, when executed by a processor of a computer, is used to cause the computer to execute a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer.

[0109] An embodiment of the present invention also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program capable of being executed by the at least one processor, and the computer program, when executed by the at least one processor, is used to cause the electronic device to execute a method for monitoring submarine crustal activities based on a differential phase-sensitive optical time domain reflectometer.

[0110] Reference Figure 2 , a block diagram of an electronic device that can be a server or a client and can be an embodiment of the present invention will now be described. It is an example of a hardware device that can be applied to various aspects of the present invention. The electronic device is intended to represent various forms of digital electronic computer devices, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0111] As Figure 2 shown, the electronic device includes a computing unit 101, which can execute various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 102 or a computer program loaded from a storage unit 108 into a random access memory (RAM) 103. In the RAM 103, various programs and data required for the operation of the electronic device can also be stored. The computing unit 101, the ROM 102, and the RAM 103 are connected to each other through a bus 104. An input / output (I / O) interface 105 is also connected to the bus 104.

[0112] Multiple components in the electronic device are connected to the I / O interface 105, including: an input unit 106, an output unit 107, a storage unit 108, and a communication unit 109. The input unit 106 can be any type of device capable of inputting information into the electronic device. The input unit 106 can receive input digital or character information and generate key signal inputs related to the user settings and / or function controls of the electronic device. The output unit 107 can be any type of device capable of presenting information and can include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 108 can include, but is not limited to, magnetic disks and optical discs. The communication unit 109 allows the electronic device to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks and can include, but is not limited to, a modem, a network card, an infrared communication device, and / or a wireless communication transceiver, such as a Bluetooth device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0113] The computing unit 101 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 101 include, but are not limited to, a CPU, a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing units, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 101 executes the various methods and processes described above. For example, in some embodiments, the method embodiments of the present invention can be implemented as a computer program tangibly contained in a machine-readable medium, such as the storage unit 108. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device via the ROM 102 and / or the communication unit 109. In some embodiments, the computing unit 101 can be configured to execute the above methods in any other suitable manner (e.g., by means of firmware).

[0114] The computer program for implementing the method of the embodiments of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the computer programs are executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The computer programs can be executed entirely on the machine, partially on the machine, executed partially on the machine as an independent software package and partially on a remote machine, or executed entirely on a remote machine or server.

[0115] In the context of embodiments of the present inventive concept, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable signal medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, or infrared system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0116] It should be noted that the term "including" and its variations used in the embodiments of the present inventive concept are open-ended, i.e., "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "a plurality" mentioned in the embodiments of the present inventive concept are illustrative rather than restrictive, and those skilled in the art should understand that, unless clearly stated otherwise in the context, it should be understood as "one or more".

[0117] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the embodiments of the present inventive concept are all information and data that have been authorized by the user or fully authorized by all parties, and the collection, use, and processing of the relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for the user to select authorization or rejection.

[0118] The various steps described in the method embodiments provided by the embodiments of the present inventive concept may be executed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The protection scope of the present inventive concept is not limited in this regard.

[0119] As used in this specification, the term "embodiment" means that the specific features, structures, or characteristics described in connection with an embodiment may be included in at least one embodiment of the present invention. The phrase appears at various positions in the specification and does not necessarily mean the same embodiment, nor does it mean being independent or alternative to other embodiments and mutually exclusive. The various embodiments in this specification are described in a related manner, and the same or similar parts among the embodiments are cross-referred to. In particular, for embodiments of devices, equipment, and systems, since they are basically similar to embodiments of methods, the description is relatively simple, and the relevant parts refer to the partial description of the method embodiments.

[0120] The above-described embodiments merely represent several implementation manners of the present invention, and the description thereof is relatively specific and detailed. However, it should not be construed as a limitation on the protection scope. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the appended claims.

Claims

1. A method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry, characterized in that: The method comprises: Acquire an echo optical pulse signal, and perform differential processing on the echo optical pulse signal to obtain a first differential echo matrix; Setting a dynamic window to obtain a second differential echo matrix of the first differential echo matrix under the constraint of the dynamic window; wherein the second differential echo matrix includes a low-rank term, a sparse term and a noise term; constructing a global optimization problem according to the second differential echo matrix, and solving a sparse item estimation matrix corresponding to the sparse item in the global optimization problem; Separating irrelevant activities in the sparse item estimation matrix to obtain a target activity matrix; The monitoring result of the target activity is obtained according to the target activity matrix.

2. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 1, characterized in that: Acquiring an echo optical pulse signal, and performing differential processing on the echo optical pulse signal to obtain a first differential echo matrix comprises: Receiving the echo light pulse signal through a spectrum analyzer, and obtaining a sampled Rayleigh curve corresponding to each echo light pulse signal; Performing differential processing on two adjacent sampled Rayleigh curves to obtain a first differential signal; The first differential signals constitute the first differential echo matrix.

3. The method for monitoring seafloor crust activity based on differential phase sensitive optical time domain reflectometry according to claim 2, characterized in that: Setting a dynamic window to obtain a second differential echo matrix of the first differential echo matrix under the constraint of the dynamic window includes: Acquire the corresponding signal main frequency according to the sampled Rayleigh curve of the echo optical pulse signal; The window length of the dynamic window is set according to the main frequency of the signal; wherein the dynamic window, the main frequency of the signal and the sampled Rayleigh curve are in one-to-one correspondence; Calculating the differential signal components of the first differential signal at each sampling point one by one; wherein the differential signal components are calculated within the length range of the dynamic window; Arranging the calculated differential signal components to obtain the second differential signal; The second differential signals constitute the second differential echo matrix.

4. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 1, characterized in that: Before constructing the global optimization problem according to the second differential echo matrix, the method further includes: Setting a weight coefficient of the low-rank term according to the dynamic window; The weight coefficient of the sparse item is calculated according to the weight coefficient of the low-rank item; wherein the sum of the weight coefficient of the low-rank item and the weight coefficient of the sparse item is 1.

5. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 4, characterized in that: Setting the weight coefficient of the low-rank term according to the dynamic window includes: Calculating a mean value of the window lengths of the dynamic windows; Calculate the standard deviation of the window length of the dynamic window according to the mean; Creating a local weight function according to the mean and the standard deviation; wherein the local weight function is set as a piecewise function, and the piecewise points of the local weight function are determined according to the mean and the standard deviation; Calculating a local weight corresponding to each of the dynamic windows according to the local weight function; An average value of the local weights is taken as the weight coefficient of the low-rank term.

6. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 1, characterized in that: Constructing a global optimization problem according to the second differential echo matrix includes: Performing singular value decomposition on the low-rank term and summing the singular values ​​to obtain a nuclear norm of the low-rank term; Taking the 1-norm of the elements in the sparse item; Taking a weighted sum of the nuclear norm of the low-rank term and the 1-norm of the sparse term to obtain an objective function of the global optimization problem; Calculating the F norm of the noise term and squaring it, and limiting the square result to be less than or equal to a noise suppression threshold as a constraint condition of the global optimization problem; wherein the noise suppression threshold is preset; The global optimization problem is obtained according to the objective function and the constraint conditions.

7. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 1, characterized in that: Separating irrelevant activities in the sparse item estimation matrix to obtain a target activity matrix includes: Setting a time domain target activity set, performing time dimension extraction on the sparse item estimation matrix, and classifying the columns of the sparse item estimation matrix corresponding to the time domain target activity into the time domain target activity set; Setting a spatial target activity set, performing spatial dimension extraction on the sparse item estimation matrix, and classifying the columns of the sparse item estimation matrix corresponding to the spatial target activity into the spatial target activity set; The columns corresponding to irrelevant activities in the sparse item estimation matrix are set to zero to obtain the target activity matrix; wherein the columns corresponding to irrelevant activities in the sparse item estimation matrix are columns that are neither included in the time domain target activity set nor in the spatial domain target activity set.

8. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 7, characterized in that: Extracting the time dimension of the sparse item estimation matrix and classifying the columns of the sparse item estimation matrix corresponding to the time domain target activity into the time domain target activity set includes: Setting a sliding window for the elements of the sparse item estimation matrix; wherein the sliding window and the elements are in the same row of the sparse item estimation matrix, the sliding window is centered on the elements, and the length of the sliding window is preset; Calculate the average value of all elements in the sliding window; For the element at the center of the sliding window, subtract the average value of all elements in the sliding window and take the second norm to obtain the smoothing error of the element; Comparing the smoothing error and the smoothness threshold of each element of the sparse term estimation matrix respectively; wherein the smoothness threshold is preset; When the smoothing error of any element is less than the smoothness threshold, the column where the element is located is included in the time-domain target activity set.

9. The method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to claim 7, characterized in that: Extracting the spatial dimension of the sparse item estimation matrix and classifying the columns of the sparse item estimation matrix corresponding to the spatial target activity into the spatial target activity set includes: For each column of the sparse item estimation matrix, the sum of the absolute values ​​of the column elements is calculated as a sparsity parameter; Comparing the sparsity parameter and the sparsity threshold of each column of the sparse item estimation matrix respectively; wherein the sparsity threshold is preset; When the sparsity parameter of any column is greater than the sparsity threshold, the corresponding column is included in the spatial target activity set.

10. An electronic device, comprising: A processor, and a memory storing a program, characterized in that the program includes instructions, which, when executed by the processor, cause the processor to execute the method for monitoring seafloor crust activity based on differential phase-sensitive optical time-domain reflectometry according to any one of claims 1 to 9.

Images