Submarine optical fiber vibration event detection method based on adaptive trust

Abstract

The present invention relates to the technical field of optical signal processing and analysis, and specifically provides a method for detecting submarine optical fiber vibration events based on adaptive trust. The method comprises: obtaining an echo matrix and decomposing it into noise terms, sparse terms and low-rank terms. A first weight matrix is ​​set corresponding to the noise terms, sparse terms and low-rank terms respectively. The trust threshold of the first weight matrix is ​​iteratively adjusted to obtain a corresponding second weight matrix. The second weight matrix is ​​weighted to the noise terms, sparse terms and low-rank terms to construct a joint optimization problem, the sparse term estimation matrix is ​​solved and the sparse non-zero elements therein are analyzed, and the detection results are output. The present invention introduces a trust weight matrix and an adaptive update mechanism to achieve high-precision detection of vibration events in complex marine environments, fully consider the different characteristics of sparse terms and low-rank terms, dynamically adapt to environmental changes during the solution process, and significantly improve detection accuracy and robustness.

Description

Technical Field

[0001] The present invention relates to the technical field of optical signal processing and analysis, and in particular to a submarine optical fiber vibration event detection method based on adaptive confidence. Background Art

[0002] Optical Time Domain Reflectometer (OTDR) technology can assess optical fiber loss and fault locations by analyzing the echo information of optical pulse signals. Therefore, OTDR technology is widely used in the inspection and maintenance of optical fiber communication systems, and is specifically used to monitor the occurrence of vibration events in submarine optical fiber communication networks.

[0003] The submarine environment is plagued by numerous interference sources. Existing technologies for processing and decomposing OTDR signals cannot dynamically adapt to the varying presence of these interference sources. As the submarine environment changes, detection solutions struggle to adapt to different echo signal types, resulting in a high false detection rate and insufficient detection accuracy. Summary of the Invention

[0004] The present invention provides a method for detecting submarine optical fiber vibration events based on adaptive trust, which at least solves the problem of being unable to dynamically adapt to different submarine environments during detection and having a high false detection rate.

[0005] In a first aspect, the present invention provides a method for detecting submarine optical fiber vibration events based on adaptive confidence, which includes:

[0006] Obtaining an echo matrix according to the echo optical pulse signal; wherein the echo matrix includes a noise term, a sparse term, and a low-rank term;

[0007] Setting a first weight matrix corresponding to the noise term, the sparse term, and the low-rank term respectively;

[0008] Iteratively adjusting the trust threshold of the first weight matrix to obtain second weight matrices corresponding to the noise term, the sparse term, and the low-rank term, respectively;

[0009] Applying the second weight matrix to the noise term, the sparse term, and the low-rank term to construct a joint optimization problem, and solving the sparse term to obtain a sparse term estimation matrix;

[0010] The sparse non-zero elements of the sparse item estimation matrix are analyzed and a detection result is output.

[0011] The method for detecting submarine optical fiber vibration events based on adaptive trust provided by an embodiment of the present invention sets a first weight matrix corresponding to the noise term, the sparse term, and the low-rank term, respectively, including:

[0012] Inputting the signal value in the echo matrix into a signal confidence function to obtain a first signal weight matrix corresponding to the noise term; wherein the first weight matrix includes the first signal weight matrix, and the confidence threshold includes a first quality confidence threshold in the signal confidence function;

[0013] Inputting the sparse components of the signal in the echo matrix into a sparse trust function to obtain a first sparse trust matrix corresponding to the sparse items; wherein the first weight matrix includes the first sparse trust matrix, and the trust threshold includes a first emergency event threshold in the sparse trust function;

[0014] The low-rank variation of the signal in the echo matrix is ​​input into a low-rank trust function to obtain a first low-rank trust matrix corresponding to the low-rank item; wherein the first weight matrix includes the first low-rank trust matrix, and the trust threshold includes the first low-rank trust threshold in the low-rank trust function.

[0015] The method for detecting submarine optical fiber vibration events based on adaptive trust provided by an embodiment of the present invention iteratively adjusts the trust threshold of the first weight matrix to obtain a second weight matrix corresponding to the noise term, the sparse term, and the low-rank term, respectively, including:

[0016] Initializing the trust threshold and the first weight matrix;

[0017] Calculating a weighted error according to the first weight matrix and the echo matrix;

[0018] Iteratively updating the confidence threshold;

[0019] The iteration is stopped when the weighted error converges to obtain the updated second weight matrix.

[0020] The method for detecting submarine optical fiber vibration events based on adaptive trust provided by an embodiment of the present invention calculates a weighted error based on the first weight matrix and the echo matrix, including:

[0021] disassembling the echo matrix;

[0022] Perform element-by-element product calculation on the corresponding components of the first weight matrix and the echo matrix, and take the F norm to obtain the weighted error; wherein, the weighted error includes the data trust weighted residual error, the sparse trust weighted residual error and the comprehensive trust weighted total error.

[0023] The method for detecting submarine optical fiber vibration events based on adaptive confidence provided by an embodiment of the present invention iteratively updates the confidence threshold, including:

[0024] Set the learning rate parameter;

[0025] The trust threshold of the previous iteration is subtracted from the update term, and adaptively updated to obtain the trust threshold in the current iteration; wherein the update term is set to the product of the learning rate parameter and the corresponding operator, and the operator is calculated by the weighted error and the first weight matrix of the previous iteration.

[0026] The method for detecting submarine optical fiber vibration events based on adaptive trust provided by an embodiment of the present invention stops iteration when the weighted error converges, and obtains the updated second weight matrix, including:

[0027] Calculate a first relative change of a data trust weighted residual error of the weighted error compared to a previous iteration;

[0028] Calculating a second relative change of a sparse trust-weighted residual error of the weighted error compared to a previous iteration;

[0029] Calculating a third relative change of the integrated trust-weighted total error of the weighted error compared to a previous iteration;

[0030] When the first relative change, the second relative change, and the third relative change are all smaller than the error tolerance, the weighted error converges and the iteration is stopped;

[0031] The first weight matrix is ​​updated according to the trust threshold at the end of the iteration to obtain the second weight matrix.

[0032] The method for detecting submarine optical fiber vibration events based on adaptive trust provided by an embodiment of the present invention constructs a joint optimization problem by weighting the second weight matrix correspondingly to the noise term, the sparse term, and the low-rank term, including:

[0033] Performing an element-by-element product calculation on a second signal weight matrix in the second weight matrix and the noise term, and taking an F-norm on the calculation result; wherein the second signal weight matrix corresponds to the first signal weight matrix in the first weight matrix;

[0034] Performing an element-by-element product calculation on a second sparse trust matrix in the second weight matrix and the sparse item, and taking a zero norm on the calculation result; wherein the second sparse trust matrix corresponds to the first sparse trust matrix in the first weight matrix;

[0035] Performing an element-by-element product calculation on a second low-rank trust matrix in the second weight matrix and the low-rank item, and taking a nuclear norm on the calculation result; wherein the second low-rank trust matrix corresponds to the first low-rank trust matrix in the first weight matrix;

[0036] The joint optimization problem is obtained by multiplying the square of the F norm by one half, summing the weighted nuclear norm, and the weighted zero norm.

[0037] The method for detecting submarine optical fiber vibration events based on adaptive confidence provided by an embodiment of the present invention analyzes the sparse non-zero elements of the sparse item estimation matrix and outputs the detection results, including:

[0038] Normalizing the sparse term estimation matrix according to the maximum value and the minimum value of the sparse non-zero elements in the sparse term estimation matrix;

[0039] Obtaining amplitude features of the normalized sparse item estimation matrix; wherein the amplitude features include maximum amplitude features, root mean square amplitude features, and maximum amplitude change rate;

[0040] When the amplitude characteristics are all greater than the corresponding preset thresholds, the element is marked as an interference event;

[0041] The elements in the sparse non-zero elements other than the interference event are marked as vibration events and output as the detection result.

[0042] In a second aspect, the present invention also provides an electronic device comprising: a processor, and a memory storing a program, wherein the program comprises instructions, which, when executed by the processor, cause the processor to execute the method for detecting submarine optical fiber vibration events based on adaptive trust according to any of the above embodiments.

[0043] In a third aspect, the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the submarine optical fiber vibration event detection method based on adaptive trust as described in any of the above embodiments.

[0044] The present invention provides an adaptive trust-based submarine optical fiber vibration event detection method that separates slowly varying events from sudden events and further extracts and identifies vibration events based on the sudden event feature matrix obtained from the separation. By introducing a trust weight matrix and an adaptive update mechanism, the method fully considers the different characteristics of sparse and low-rank terms and dynamically adapts to environmental changes during the solution process, enabling detection of vibration events in complex marine environments and significantly improving detection accuracy and robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without inventive effort.

[0046] Figure 1 The flowchart of the submarine optical fiber vibration event detection method based on adaptive confidence level according to an embodiment of the present invention is shown.

[0047] Figure 2 It is a structural schematic diagram of the electronic device created by the present invention. DETAILED DESCRIPTION

[0048] The following describes embodiments of the present invention in more detail with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying 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 described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0049] Submarine optical fibers, as critical infrastructure for marine communications systems, carry massive amounts of data. Therefore, their maintenance is crucial to the proper functioning of marine facilities and activities. Submarine optical fibers vibrate under external forces. Slight vibrations can cause microbends in the fiber, increasing light signal scattering and loss, and impacting communication quality. Continuous vibrations can subject the fiber to prolonged stress, leading to fatigue damage and structural fractures, shortening its service life and, in severe cases, causing communication interruptions.

[0050] To monitor the status of submarine optical fibers promptly and accurately and prevent communication failures, optical time domain reflectometer (OTDR) technology has been introduced into submarine optical fiber communication systems for maintenance. OTDR technology transmits optical pulses into the fiber and uses the information in the echo signal to accurately locate faults and abnormalities within the fiber. It also assesses fiber loss changes in real time, ensuring the safe and stable operation of submarine optical fiber communication networks. OTDR technology is particularly useful for monitoring vibration events within submarine optical fiber communication networks.

[0051] The complexity of the marine environment and the numerous interference sources make it difficult for signal processing methods to adapt to different types of signal interference. Signal interference primarily includes periodic interference such as tides and sudden interference from natural disasters such as earthquakes and tsunamis. Because the vibration event signals to be detected coexist with the interference signals, and detection methods are difficult to dynamically adjust to environmental changes, the false detection rate is high.

[0052] Reference Figure 1 As shown, the first embodiment of the present invention provides a submarine optical fiber vibration event detection method based on adaptive confidence, which is used to eliminate interference signals and extract and identify vibration events. Furthermore, the method sets confidence weights and adaptively updates the confidence coefficients when processing them, dynamically adapting to environmental changes and improving detection accuracy and robustness.

[0053] Specifically, the submarine optical fiber vibration event detection method based on adaptive trust includes the following steps:

[0054] Step S100: Obtain an echo matrix R according to the echo optical pulse signal. The echo matrix R includes a noise term N, a sparse term S, and a low-rank term L.

[0055] In this embodiment, the received signal is received by an optical spectrum analyzer. In an OTDR-based distributed fiber optic sensing system, the OTDR echo optical pulse signal received by the optical spectrum analyzer is the received signal. The matrix formed by the sampled Rayleigh curves corresponding to the P received echo optical pulse signals is the echo matrix R.

[0056] The echo matrix R is specifically expressed as:

[0057] ;

[0058] Where, represents the sampling Rayleigh curve corresponding to the p-th echo light pulse signal, p is an integer, and 1≤p≤P. In this embodiment, the initial monitoring time of the vibration event is set to , the pulse repetition interval of the light pulse emitted by the light source is , so the monitoring time corresponding to the pth echo optical pulse signal is + .

[0059] The sampled Rayleigh curve contains position information x and time information t. The position information x corresponds to different monitoring positions on the optical fiber and is used to determine the fault and loss points on the optical fiber. The time information t reflects the monitoring time. In this embodiment, each sampled Rayleigh curve corresponds to M monitoring positions, which are expressed as The element R(i, j) in the echo matrix R represents Always at the location The signal data received on.

[0060] In submarine optical fiber vibration detection, common events that cause false alarms and misjudgments include periodic disturbances such as tides and natural disasters such as earthquakes and tsunamis. This embodiment decomposes the echo matrix R based on the differences between these two types of events. The optical pulse echo matrix R consists of a noise term N, a sparse term S, and a low-rank term L, as shown in the following formula:

[0061] ;

[0062] In the formula, L is a low-rank term, reflecting periodic, slow-changing events such as tides and some ocean currents. S is a sparse term, reflecting non-periodic, sudden events; sudden events include natural disasters such as earthquakes and tsunamis, as well as submarine optical fiber vibrations. Submarine optical fiber vibrations are the target events for the detection method in this embodiment. N is a noise term, reflecting other noise.

[0063] The method provided in this embodiment adopts a low-rank and sparse decomposition model to separate slowly varying events from sudden events, eliminate the influence of periodic events such as tides on detection accuracy, and calculate and separate natural disasters and vibration events in sparse terms.

[0064] Step S200 , setting a first weight matrix corresponding to the noise term N, the sparse term S, and the low-rank term L respectively.

[0065] Specifically, the first weight matrix includes a first signal weight matrix set corresponding to the noise term N, the sparse term S and the low rank term L respectively. , the first sparse trust matrix and the first low-rank trust matrix .

[0066] As an implementable method, step S200 includes:

[0067] Step S210: Input the signal value of the element R(i, j) in the echo matrix R into the signal trust function to define the trust of the received data, which is used to measure the trust of the received data. Always at the location The reliability of the OTDR signal received on the 10000 node is obtained by calculating the first signal weight matrix corresponding to the noise term N. . Refer to the following formula:

[0068] ;

[0069] The signal confidence function is set as the logistic function. is the first quality confidence threshold, is the sensitivity coefficient used to adjust the signal quality confidence, The quality metric used to characterize the received signal is the mean value of the received signal within the observation window and is defined as Where is the position at time k The received signal value on is the size of the observation window.

[0070] Step S220: Input the sparse component of the element R(i, j) in the echo matrix R into the sparse trust function to obtain the first sparse trust matrix corresponding to the sparse item S. The first sparse trust matrix Refer to the following formula:

[0071] ;

[0072] The sparse trust function is set as the logistic function. is the first emergency threshold, It is a sensitivity coefficient used to measure the change in the trust of the sparse matrix. A quality metric used to characterize sparse signals.

[0073] For location The maximum value of the sparse component of the received signal within the upper observation window is defined as Where is the position at time k The sparse components of the received signal on . Larger, indicating location There may be strong abnormal signals and vibration events. A smaller value indicates that there is no obvious sparse feature.

[0074] The first sparse trust matrix This measure measures the reliability of each signal within the sparse portion. During natural disasters, sparse portions often contain more valid emergency information, resulting in a relatively high degree of trust. By weighted optimization of sparse terms, the ability to respond to emergencies like earthquakes and tsunamis can be significantly improved.

[0075] Step S230: Input the low-rank variation of the element R(i, j) in the echo matrix R into the low-rank trust function to obtain the first low-rank trust matrix corresponding to the low-rank item L. The first low-rank trust matrix Refer to the following formula:

[0076] ;

[0077] The low-rank trust function is set as the logistic function. is the first low-rank trust threshold, is the low-rank trust sensitivity coefficient, Used to characterize the low-rank variation of the signal.

[0078] Defined as Where L(i, k) is the position at time k. The low-rank component of the received signal on .

[0079] The first low-rank trust matrix This method is used to reflect the reliability of data at different locations and times in slowly varying signals such as tides. During the tidal cycle, the periodic variation of the signal makes the trustworthiness of the low-rank components relatively high. Weighted optimization is used to ensure the reliability of the signal under the influence of tides.

[0080] Step S300: Iteratively adjust the trust threshold of the first weight matrix to obtain a second weight matrix corresponding to the noise term N, the sparse term S, and the low-rank term L respectively.

[0081] Specifically, the confidence threshold in the first weight matrix includes the first quality confidence threshold in the signal confidence function , the first emergency threshold in the sparse trust function , and the first low-rank trust threshold in the low-rank trust function .

[0082] Step S300 dynamically adjusts the confidence threshold in the first weight matrix, adaptively updating it based on a feedback loop to improve the accuracy and robustness of vibration event estimation. Embodiment 1 of the present invention introduces an adaptive confidence update mechanism that continuously updates statistical information about signal features and adjusts the confidence threshold based on estimation errors, thereby optimizing the confidence of low-rank events, sparse events, and received signal data. Dynamically adjusting the weights and reliability of each signal component enhances the model's ability to detect sudden events.

[0083] As an implementation method, step S300 includes the following steps:

[0084] Step S310: Initialize the trust threshold and the first weight matrix.

[0085] Initialize the first quality confidence threshold to , initialize the first emergency threshold to , initialize the first low-rank trust threshold to .

[0086] The initialized first weight matrix is ​​obtained according to the initialized trust threshold 、 and .in, Initialize the first low-rank trust threshold The first low-rank trust matrix after ; Initialize the first emergency threshold The first sparse trust matrix after ; Initialize the first quality confidence threshold The first signal weight matrix after .

[0087] In step S310, when calculating the first weight matrix, the sensitivity coefficient used to adjust the signal quality confidence , used to measure the sensitivity coefficient of the change in the trust of the sparse matrix and low-rank trust sensitivity coefficient The confidence threshold can be set based on historical data. The confidence threshold can also be set to an initial value based on historical data or prior knowledge, which includes but is not limited to past experience data, signal analysis characteristics, etc.

[0088] Step S320 , weighting the reconstruction error according to the first weight matrix and the echo matrix R to obtain a weighted error.

[0089] Specifically, the disassembled echo matrix R is Where L is the low-rank term, S is the sparse term, N is the noise term, and q is the iteration count, which is initialized to 0 in step S320.

[0090] The weighted error includes the data confidence weighted residual error , sparse trust weighted residual error and the comprehensive trust-weighted total error The weighted error is calculated by performing element-by-element product of the first weight matrix and the corresponding components of the echo matrix R, and taking the F norm. The weighted error is used to reflect the role of the first weight matrix in signal decomposition. The weighted error is shown in the following formula:

[0091] ;

[0092] ;

[0093] .

[0094] Step S330: iteratively update the trust threshold.

[0095] Set the learning rate parameter η L ,η S ,η D . Set the error tolerance , the learning rate parameter η L ,η S ,η D and error tolerance Can be set based on historical data or prior knowledge.

[0096] Starting from q = 0, the iteration is performed and the trust threshold is adaptively updated. During the update, the trust threshold is subtracted from the update term and the trust threshold in the current iteration is adaptively updated. The update of the trust threshold is shown in the following formula:

[0097] ;

[0098] Where, the first quality confidence threshold The update term is set to the learning rate parameter η D The product of the corresponding operator, the operator is weighted by the total error of the comprehensive trust in the weighted error and the first signal weight matrix of the first weight matrix of the current iteration Calculated. The operator is calculated as shown in the following formula;

[0099] .

[0100] ;

[0101] Where, the first emergency threshold The update term is set to the learning rate parameter η S The product of the corresponding operator, the operator is composed of the sparse trust weighted residual error in the weighted error and the first sparse confidence matrix of the first weight matrix of the current iteration Calculated. The operator is calculated as shown in the following formula;

[0102] .

[0103] ;

[0104] Where, the first low-rank trust threshold The update term is set to the learning rate parameter η L The product of the corresponding operator, the operator is the residual error weighted by the data trust in the weighted error and the first low-rank trust matrix of the first weight matrix of the current iteration The operator is calculated as shown in the following formula:

[0105] .

[0106] Step S340, respectively according to the first low-rank trust threshold updated in step S330 , first emergency threshold and the first quality confidence threshold Correspondingly update the first weight matrix to obtain 、 and .

[0107] Calculate the weighted error of the data trust weighted residual error First relative change compared to the previous iteration: Calculate the sparse trust-weighted residual error of the weighted error The second relative change compared to the previous iteration; calculate the comprehensive trust-weighted total error of the weighted error The third relative change compared to the previous iteration.

[0108] Error tolerance As the convergence threshold, the first relative change, the second relative change, and the third relative change are all less than the error tolerance ,Right now and and When , the weighted error meets the convergence condition. The first weight matrix after the trust threshold is updated at the end of this iteration is used as the second weight matrix.

[0109] It should be noted that when judging whether the iteration has converged, if q = 0, the trust threshold is adaptively updated directly without judgment and the next iteration is entered. When q > 0, the iterative convergence judgment is performed. If the convergence condition is met, the second weight matrix is ​​output. If not, q = q + 1 is set to continue the iteration.

[0110] Step S400: weight the second weight matrix obtained in step S300 to the noise term N, the sparse term S, and the low-rank term L, construct a joint optimization problem for the vibration event detection process, and solve the sparse term S to obtain the sparse term estimation matrix .

[0111] Specifically, the second weight matrix includes a second signal weight matrix W D ', the second sparse trust matrix W S ' and the second low-rank trust matrix W L '. Among them, the second signal weight matrix W D ' is the first signal weight matrix The second sparse trust matrix W obtained after the corresponding update S ' is the first sparse trust matrix The second low-rank trust matrix W obtained after the corresponding update L ' is the first low-rank trust matrix Corresponding to the updated one.

[0112] The second signal weight matrix W D 'Perform element-wise product with the noise term N and take the F norm of the result.

[0113] The second sparse confidence matrix W S'Compute the element-wise product with the sparse term S and take the zero norm of the result.

[0114] The second low-rank trust matrix W L 'Compute the element-wise product with the low-rank term L and take the nuclear norm of the result.

[0115] Multiplying the square of the F norm by half, summing the weighted nuclear norm and the weighted zero norm, we get the following joint optimization problem:

[0116] .

[0117] Where, is the Frobenius norm, which is used to measure the fitting error of the model. is the nuclear norm, which is used to control the rank of the low-rank term L. is the zero norm of the sparse matrix S, which is used to control sparsity.

[0118] and is a regularization parameter that balances the weights of the low-rank term L and the sparse term S respectively.

[0119] In this embodiment, the sparse term S in the optimization model is solved by the Alternating Direction Method of Multipliers (ADMM) to obtain the sparse term estimation matrix .

[0120] ADMM algorithm is an algorithm for solving optimization problems. It combines the ideas of augmented Lagrangian function and dual ascent algorithm. It is widely used in many fields such as machine learning, signal processing, image processing, statistical analysis, etc. ADMM algorithm is used to optimize the optimization problem and the sparse item estimation matrix is ​​obtained. Includes natural disaster events and vibration events.

[0121] Step S500, estimating the matrix for sparse items Analyze the sparse non-zero elements of and output the detection results.

[0122] As an implementation method, step S500 includes the following steps:

[0123] Step S510, estimating the matrix based on the sparse terms The maximum value of the sparse non-zero elements in and minimum value , estimate the matrix for sparse terms Normalize it as shown below:

[0124] ;

[0125] Where, Estimate matrix for sparse terms The element at row n and column p in .

[0126] Step S520, obtain the normalized sparse item estimation matrix The amplitude features include the maximum amplitude feature, the root mean square amplitude feature, and the maximum amplitude change rate.

[0127] Extract the normalized sparse matrix at each observation moment The maximum amplitude characteristic of:

[0128] .

[0129] Extract the normalized sparse matrix at each observation moment The root mean square amplitude characteristics of:

[0130] .

[0131] Calculate the maximum amplitude change rate between adjacent observation times:

[0132] .

[0133] Step S530: When the amplitude characteristics are all greater than the corresponding preset thresholds, the element is marked as an interference event.

[0134] Maximum amplitude threshold , the threshold of the root mean square amplitude and peak rate of change threshold All are preset thresholds. and and When , the sparse matrix is ​​normalized The sparse non-zero elements in are marked as interference events, that is, sudden disaster events other than vibration events in the sparse items.

[0135] Normalized sparse matrix The other sparse non-zero elements except those marked as sudden disaster events are marked as vibration events, and the vibration event detection results are output.

[0136] The method provided in Example 1 of this invention significantly improves detection accuracy and robustness through multi-dimensional feature analysis, joint judgment, and real-time processing. Its flexibility and visualization capabilities make it widely applicable in practical applications, enabling effective responses to various vibration events and sudden disasters, and providing strong support for safety monitoring and risk management.

[0137] The present invention also provides a computer program product including a computer program. When the computer program is executed by a processor, the method for detecting submarine optical fiber vibration events based on adaptive confidence provided in the first embodiment is implemented.

[0138] An embodiment of the present invention further 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 the method for detecting submarine optical fiber vibration events based on adaptive confidence according to embodiment 1 of the present invention.

[0139] An embodiment of the present invention further provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, and when executed by the at least one processor, the computer program causes the electronic device to perform the method for detecting submarine optical fiber vibration events based on adaptive confidence level according to the first embodiment of the present invention.

[0140] refer to Figure 2 , a structural block diagram of an electronic device that can be used as a server or client of an embodiment of the present invention will now be described, which 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 laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices 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 required herein.

[0141] like Figure 2 As shown, the electronic device includes a computing unit 101, which can perform 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. RAM 103 may also store various programs and data required for the operation of the electronic device. Computing unit 101, ROM 102, and RAM 103 are interconnected via a bus 104. An input / output (I / O) interface 105 is also connected to bus 104.

[0142] 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 user settings and / or function control 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, a magnetic disk and an optical disk. 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.

[0143] The computing unit 101 can be any general-purpose and / or specialized processing component 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 specialized artificial intelligence (AI) computing units, various computing units that run machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 101 performs the various methods and processes described above. For example, in some embodiments, the method embodiments of the present invention may be implemented as a computer program tangibly embodied in a machine-readable medium, such as a storage unit 108. In some embodiments, part or all of the computer program may 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 may be configured to perform the above-described methods by any other suitable means (e.g., via firmware).

[0144] The computer programs for implementing the methods of the embodiments of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0145] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The 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 machine-readable storage media may 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), optical fibers, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0146] It should be noted that the term "including" and its variations used in the embodiments of the present invention are open inclusions, that is, "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 other embodiment"; the term "some embodiments" means "at least some embodiments". The modifications of "one" and "multiple" mentioned in the embodiments of the present invention are illustrative and not restrictive. Those skilled in the art should understand that unless the context clearly indicates otherwise, they should be understood as "one or more".

[0147] The various steps described in the method implementation methods provided by the embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method implementation methods may include additional steps and / or omit the steps shown. The scope of protection of the present invention is not limited in this respect.

[0148] The term "embodiment" in this specification refers to specific features, structures or characteristics described in conjunction with the embodiment that can be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily mean the same embodiment, nor does it mean that it is mutually exclusive with other embodiments and is independent or optional. The various embodiments in this specification are described in a related manner, and the same or similar parts between the various embodiments are referenced to each other. In particular, for the device, equipment, and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts refer to the partial description of the method embodiment.

[0149] The above-described embodiments merely represent several implementation methods of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of protection. It should be noted that a person of ordinary skill in the art would be able to make various modifications and improvements without departing from the scope of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A submarine optical fiber vibration event detection method based on adaptive trust, characterized in that: The method comprises: Obtaining an echo matrix according to the echo optical pulse signal; wherein the echo matrix includes a noise term, a sparse term, and a low-rank term; Setting a first weight matrix corresponding to the noise term, the sparse term, and the low-rank term respectively; Iteratively adjusting the trust threshold of the first weight matrix to obtain second weight matrices corresponding to the noise term, the sparse term, and the low-rank term, respectively; Applying the second weight matrix to the noise term, the sparse term, and the low-rank term to construct a joint optimization problem, and solving the sparse term to obtain a sparse term estimation matrix; Analyzing the sparse non-zero elements of the sparse item estimation matrix and outputting a detection result; The second weight matrix is ​​weighted correspondingly to the noise term, the sparse term, and the low-rank term to construct a joint optimization problem, including: Performing an element-by-element product calculation on a second signal weight matrix in the second weight matrix and the noise term, and taking an F-norm on the calculation result; wherein the second signal weight matrix corresponds to the first signal weight matrix in the first weight matrix; Performing an element-by-element product calculation on a second sparse trust matrix in the second weight matrix and the sparse item, and taking a zero norm on the calculation result; wherein the second sparse trust matrix corresponds to the first sparse trust matrix in the first weight matrix; Performing an element-by-element product calculation on a second low-rank trust matrix in the second weight matrix and the low-rank item, and taking a nuclear norm on the calculation result; wherein the second low-rank trust matrix corresponds to the first low-rank trust matrix in the first weight matrix; The joint optimization problem is obtained by multiplying the square of the F norm by one half, summing the weighted nuclear norm, and the weighted zero norm.

2. The submarine optical fiber vibration event detection method based on adaptive trust according to claim 1 is characterized in that: Setting a first weight matrix corresponding to the noise term, the sparse term, and the low-rank term respectively includes: Inputting the signal value in the echo matrix into a signal confidence function to obtain a first signal weight matrix corresponding to the noise term; wherein the first weight matrix includes the first signal weight matrix, and the confidence threshold includes a first quality confidence threshold in the signal confidence function; Inputting the sparse components of the signal in the echo matrix into a sparse trust function to obtain a first sparse trust matrix corresponding to the sparse items; wherein the first weight matrix includes the first sparse trust matrix, and the trust threshold includes a first emergency event threshold in the sparse trust function; The low-rank variation of the signal in the echo matrix is ​​input into a low-rank trust function to obtain a first low-rank trust matrix corresponding to the low-rank item; wherein the first weight matrix includes the first low-rank trust matrix, and the trust threshold includes the first low-rank trust threshold in the low-rank trust function.

3. The submarine optical fiber vibration event detection method based on adaptive confidence according to claim 1, characterized in that: Iteratively adjusting the trust threshold of the first weight matrix to obtain second weight matrices corresponding to the noise term, the sparse term, and the low-rank term, respectively, includes: Initializing the trust threshold and the first weight matrix; Calculating a weighted error according to the first weight matrix and the echo matrix; Iteratively updating the confidence threshold; The iteration is stopped when the weighted error converges to obtain the updated second weight matrix.

4. The method for detecting submarine optical fiber vibration events based on adaptive confidence according to claim 3, characterized in that: Calculating a weighted error according to the first weight matrix and the echo matrix includes: disassembling the echo matrix; Perform element-by-element product calculation on the corresponding components of the first weight matrix and the echo matrix, and take the F norm to obtain the weighted error; wherein, the weighted error includes the data trust weighted residual error, the sparse trust weighted residual error and the comprehensive trust weighted total error.

5. The method for detecting submarine optical fiber vibration events based on adaptive confidence according to claim 3, characterized in that: Iteratively updating the trust threshold includes: Set the learning rate parameter; The trust threshold of the previous iteration is subtracted from the update term, and adaptively updated to obtain the trust threshold in the current iteration; wherein the update term is set to the product of the learning rate parameter and the corresponding operator, and the operator is calculated by the weighted error and the first weight matrix of the previous iteration.

6. The method for detecting submarine optical fiber vibration events based on adaptive confidence according to claim 3, characterized in that: When the weighted error converges, the iteration is stopped to obtain the updated second weight matrix, including: Calculate a first relative change of a data trust weighted residual error of the weighted error compared to a previous iteration; Calculating a second relative change of a sparse trust-weighted residual error of the weighted error compared to a previous iteration; Calculating a third relative change of the integrated trust-weighted total error of the weighted error compared to a previous iteration; When the first relative change, the second relative change, and the third relative change are all smaller than the error tolerance, the weighted error converges and the iteration is stopped; The first weight matrix is ​​updated according to the trust threshold at the end of the iteration to obtain the second weight matrix.

7. The method for detecting submarine optical fiber vibration events based on adaptive confidence according to claim 1, characterized in that: Analyzing the sparse non-zero elements of the sparse item estimation matrix and outputting a detection result, including: Normalizing the sparse term estimation matrix according to the maximum value and the minimum value of the sparse non-zero elements in the sparse term estimation matrix; Obtaining amplitude features of the normalized sparse item estimation matrix; wherein the amplitude features include maximum amplitude features, root mean square amplitude features, and maximum amplitude change rate; When the amplitude characteristics are all greater than the corresponding preset thresholds, the element is marked as an interference event; The elements in the sparse non-zero elements other than the interference event are marked as vibration events and output as the detection result.

8. An electronic device comprising: A processor and a memory storing a program, wherein the program comprises instructions, which, when executed by the processor, cause the processor to perform the method for detecting submarine optical fiber vibration events based on adaptive confidence according to any one of claims 1 to 7.

9. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method for detecting submarine optical fiber vibration events based on adaptive confidence as claimed in any one of claims 1 to 7 is implemented.

Images