Submarine Optical Cable Disturbance Identification Method and System Based on Time-Frequency-Space and Neural Network

Abstract

The present invention relates to a method and system for identifying submarine optical cable disturbances based on time-frequency-space and neural networks. The method includes: obtaining the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable; inputting the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable into a trained convolutional neural network to obtain the disturbance category of the target submarine optical cable. By using the neural network to extract the features of multi-dimensional data in time, frequency, and space, since the features are based on disturbance feature information such as short-time average energy, spectral density, and duration, the present disclosure can extract multi-dimensional event information such as the disturbed range of the optical cable, the relative position of the disturbance source, and the moving speed of the disturbance source from the three-dimensional composite data of time, frequency, and space, effectively identify hazard events, and improve the early warning efficiency of system threat events.

Description

Technical Field

[0001] The present invention belongs to the technical field of submarine optical cable disturbance identification, and particularly relates to a submarine optical cable disturbance identification and system based on three-dimensional composite detection of time, frequency and space and convolutional neural network. Background Art

[0002] During the operation and use of submarine optical cables, various disturbance events such as mechanical cutting, biological biting, and anchor damage will damage the submarine optical cables. Since the submarine optical cables are laid on the seabed, accurately identifying the types of submarine optical cable disturbance events is very important for timely determining the disposal plan and improving the disposal efficiency.

[0003] Although human activities in the deep and far seas are relatively few, there are still many situations that can cause disturbance effects on submarine optical cables, and the disturbance situation of submarine optical cables is still complex. Disturbance events usually cause different types of disturbances to submarine optical cables in three stages. These disturbances have a strict sequence in time, significant differences in the disturbance frequency domain distribution, and obvious characteristics in the vertical distance from the submarine optical cable in space. Therefore, in order to more effectively and accurately monitor eavesdropping behavior, it is necessary to establish a composite detection model by combining the three dimensions of time, frequency domain, and space, and preprocess the disturbance signals. In the prior art, there is no comprehensive consideration of multi-dimensional data in time, frequency, and space for analysis and identification of submarine optical cable disturbances based on this. Summary of the Invention

[0004] In order to comprehensively consider multi-dimensional detection data in the time domain, frequency domain, and spatial domain related to submarine optical cable disturbances and improve the sensitivity of submarine optical cable disturbance identification, in the first aspect of the present invention, a method for identifying submarine optical cable disturbances based on time, frequency, space, and neural network is provided, including: obtaining time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable; inputting the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable into a trained convolutional neural network to obtain the disturbance category of the target submarine optical cable.

[0005] In some embodiments of the present invention, the convolutional neural network includes an input layer, four parallel convolutional layers, a pooling layer, a fusion layer, a fully connected layer, and an output layer. The input layer is used to preprocess the obtained time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable and convert it into multi-dimensional data. The four parallel convolutional layers are used to extract features in the time domain, frequency domain, energy, and duration dimensions from the multi-dimensional data respectively.

[0006] Furthermore, the convolutional neural network is a one-dimensional convolutional neural network.

[0007] In some embodiments of the present invention, the convolutional neural network is trained through the following steps: obtaining different types of disturbance signals of submarine optical cable vibration and their simulation signals; using the disturbance signals and the simulation signals as samples, and using the disturbance category corresponding to each sample as a label to construct a data set; using the data set to train the convolutional neural network until its error tends to be stable and is lower than a threshold, obtaining a trained convolutional neural network.

[0008] Further, the disturbance categories include traveling wave disturbances of vehicles, mechanical cutting disturbances, and marine organism biting disturbances.

[0009] In the above embodiment, the obtaining of the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable includes: using multiple long-time detections of detection pulses with multiple frequencies combined to obtain the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable.

[0010] In a second aspect of the present invention, there is provided a submarine optical cable disturbance recognition system based on time-frequency-space and neural network, including: an acquisition module for acquiring the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable; an identification module for inputting the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable into the trained convolutional neural network to obtain the disturbance category of the target submarine optical cable.

[0011] In a third aspect of the present invention, there is provided an electronic device, including: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the submarine optical cable disturbance recognition method based on time-frequency-space and neural network provided by the present invention in the first aspect.

[0012] In a fourth aspect of the present invention, there is provided a computer-readable medium, on which a computer program is stored, wherein when the computer program is executed by a processor, it implements the submarine optical cable disturbance recognition method based on time-frequency-space and neural network provided by the present invention in the first aspect.

[0013] The beneficial effects of the present invention are:

[0014] 1. The present invention proposes a technology for establishing a time-frequency-space multi-dimensional data feature model library and pattern recognition based on deep learning. By using deep learning algorithms such as convolutional neural networks for time, frequency, and space multi-dimensional data, based on disturbance feature information such as short-time average energy, spectral density, and duration, through intelligent learning and classification of disturbance events, a model database of disturbance event feature data is established, and an event detection method for three-dimensional composite data of time, frequency, and space is proposed;

[0015] 2. Combine with the established disturbance event feature model library, extract multi-dimensional event information such as the disturbed range of the optical cable, the relative position of the disturbance source, and the moving speed of the disturbance source from the three-dimensional composite data of time, frequency, and space, effectively identify hazard events, and improve the early warning efficiency of system threat events. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of the basic process of the submarine optical cable disturbance identification method based on time-frequency-space and neural network in some embodiments of the present invention;

[0017] Figure 2 Schematic diagram of the specific process of the submarine optical cable disturbance identification method based on time-frequency-space and neural network in some embodiments of the present invention;

[0018] Figure 3 Schematic diagram of the change in the length of the optical fiber disturbance;

[0019] Figure 4 Schematic diagram for calculating the vertical distance from the disturbance source to the optical fiber;

[0020] Figure 5 Schematic diagram of the waveform of the moving disturbance source;

[0021] Figure 6 Time trend graph of the number of times of submarine optical cable damage in a certain sea area;

[0022] Figure 7 Stress curve of the submarine optical cable;

[0023] Figure 8 Vibration waveform graph of the ship anchor impacting the rigid bottom sea floor;

[0024] Figure 9 Vibration frequency domain graph of the ship anchor impacting the rigid bottom sea floor;

[0025] Figure 10 Time domain waveform graph of seismic waves;

[0026] Figure 11 Frequency domain waveform graph of seismic waves;

[0027] Figure 12 Force waveform graph of the submarine optical cable affected by ship movement disturbance;

[0028] Figure 13 Schematic diagram of the principle of the convolutional neural network in some embodiments of the present invention;

[0029] Figure 14 Schematic diagram of the structure of the submarine optical cable disturbance identification system based on time-frequency-space and neural network in some embodiments of the present invention;

[0030] Figure 15Schematic structural diagram of an electronic device in some embodiments of the present invention. Detailed implementation manners

[0031] The principles and features of the present invention will be described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0032] Refer to Figure 1 and Figure 2 In a first aspect of the present invention, a method for identifying submarine cable disturbances based on time-frequency-space and neural networks is provided, including: S100. Obtaining time-domain, frequency-domain, energy, and duration information of the vibration of a target submarine cable; S200. Inputting the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine cable into a trained convolutional neural network to obtain the disturbance category of the target submarine cable.

[0033] Refer to Figure 2 and

[0034] It can be understood that in the embodiments of the present invention, the extraction or acquisition of the time, frequency-domain, and space information of submarine cable disturbances is achieved through the following steps respectively:

[0035] 1. Time information

[0036] Time is an important dimension for analyzing disturbance information. The three-dimensional composite monitoring time dimension analysis mainly includes two aspects: a. Analyzing the duration of the disturbance event: Due to the complexity of the marine environment, the disturbance duration of multiple different disturbance sources on the submarine cable is not a fixed value. The duration of different disturbance events can range from a few seconds to several hours. Therefore, the duration of the disturbance is an important characteristic parameter of the disturbance event, and collecting and recording the disturbance duration is conducive to classifying and identifying the disturbance source; b. The time sequence relationship of the disturbance events: Analyzing the time sequence relationship of different disturbance sources and the relationship between the intensity of the same disturbance source and time is helpful for classifying and identifying different disturbance events. In the traveling wave disturbance of a submersible, for the disturbance source close to the cable, the intensity of the disturbance signal increases with time, and vice versa, the intensity decreases with time; in the eavesdropping event, there is first a traveling wave disturbance of a submersible at a relatively close distance, then a pulling disturbance of the fixed submarine cable, and finally a mechanical disturbance of cutting the outer sheath of the submarine cable.

[0037] 2. Frequency-domain information

[0038] Frequency domain analysis mainly realizes the extraction of the spectral characteristics of broadband disturbance signals. According to the spectral distribution of different disturbance signals, their frequency domain characteristics are extracted, providing an important basis for the identification and classification of disturbance sources. In a vibration optical fiber sensing system based on Rayleigh scattering, emitting a single detection pulse to the sensed optical fiber is equivalent to sampling the vibration information carried by the optical fiber once. The response frequency of the system, that is, the maximum vibration frequency that can be detected, is limited by the repetition frequency of the detection pulse. After frequency division multiplexing, the response frequency of the system can be increased. Assuming that the undersea optical cable is 1000 km long, it can be estimated from the following formula (1) that the maximum detection pulse repetition frequency is 100 Hz, that is, the system response frequency is 50 Hz. This obviously cannot meet the actual application requirements. If 40 frequencies of detection light are used and interpolation algorithm processing is performed on the disturbance signal frequencies detected for each frequency of detection light, the response frequency can be doubled, that is, the response frequency can be increased from 50 Hz to 2000 Hz, greatly improving the system performance and detection efficiency.

[0039] 3. Spatial information

[0040] Reference Figure 2 , the analysis of the spatial dimension mainly includes three aspects. First, after the interpolation operation of the data detected by frequency division multiplexing, it can be considered that the repetition frequency of a single detection pulse is increased. Therefore, along the length direction of the optical fiber, the spatial distance of the disturbed optical fiber can be detected. Assuming that the time width of a single-frequency pulse is W pluse , the time width of the filled pseudo-pulse is W false , the length of the part of the optical fiber disturbed under the action of the disturbance source is L d , then L d There is:

[0041] Formula (1),

[0042] In the formula, k represents the number of pulses detecting the disturbance, c is the speed of light in vacuum, and n is the refractive index of the optical fiber, and the disturbed length of the optical fiber at Figure 3 can be calculated to be 10 meters.

[0043] Second, after multiple detections, according to the spatial position information in the disturbance length direction of different detection traces, the vertical position of the disturbance source from the optical fiber can be analyzed, and its principle is as Figure 4 shown. According to formula (1), the vertical distance from the disturbance source to the optical fiber can be calculated .

[0044] Formula (2),

[0045] In the formula v is the propagation speed of sound in seawater, and the expression of can be solved.

[0046] Formula (3).

[0047] Thirdly, analyze the moving speed v of the disturbance source along the fiber optic cable length direction d , and the principle is as Figure 5 shown. Where the horizontal axis is the disturbance position on the fiber optic cable, and the vertical axis is the time when the detection pulse repeats. That is, at time t1, the disturbance position on the fiber optic cable is z3, and by time t5, the disturbance position has moved to z7. Thus, the moving speed v of the disturbance source can be calculated by formula (4) d .

[0048] Formula (4).

[0049] It can be understood that the above information can only obtain the distance, space, and frequency information of the disturbance source, and cannot distinguish and identify the corresponding causes of the disturbance source. Therefore, it is necessary to perform feature analysis (feature extraction) on the above time-frequency-space information. Since there are many artificial and natural factors that can cause disturbances to submarine optical cables, this article mainly analyzes the disturbances caused by common disturbance events for shallow sea and deep sea optical cables, such as anchor damage, submarine geological movement, and ship traveling waves

[0050] A. Anchor damage

[0051] Refer to Figure 6 , which shows the statistical situation of ship anchoring causing damage to submarine optical cables in a certain area. This article analyzes the disturbances caused by ship anchoring to submarine optical cables, mainly including two situations: one is that the ship anchor directly lands on the submarine optical cable, and the other is that the ship anchor lands near the submarine optical cable

[0052] (1) The ship anchor lands on the submarine optical cable

[0053] For this type of disturbance event, it usually occurs in shallow seas. Taking the double-armored shallow sea optical cable as an example, this article studies the disturbance caused by the ship anchor directly impacting the submarine optical cable. According to research, if the submarine optical cable is laid on a non-rigid substrate and a ship anchor weighing 100 kg falls from a height of 3 meters, the deformation process of the submarine optical cable is as follows three stages

[0054] Impact stage, the ship anchor falls from a high place and impacts the submarine optical cable, causing the submarine optical cable to sink and deform. Transition stage, in this stage, the sinking depth of the submarine optical cable reaches the maximum, the falling speed of the ship anchor is zero, and the submarine optical cable begins to rebound. Rebound stage, in this stage, the submarine optical cable rebounds a certain distance and bounces the ship anchor away. The above process is through Figure 7Characterized by the stress curve of the armored layer of the submarine optical cable. That is: the three stages experienced by the ship anchor hitting the submarine optical cable. Only after the ship anchor is bounced off, the armored layer of the submarine optical cable has a certain amount of plastic deformation, so there is stress residue. According to another study, the final deformation of the armored layer of the submarine optical cable is about 1.5 mm, and the deformation of the copper sheath or stainless steel is about 1 mm. According to the test, when a 100 kg ship anchor falls from a height of 3 meters and hits the submarine optical cable laid on non-rigid bottom sediment, the depth of the formed pit is about 0.94 cm, the deflection of the optical cable is about 21.30 mm, and the deformation of the optical cable is 8.85%.

[0055] (2) The ship anchor lands near the submarine optical cable

[0056] The ship anchor landing near the submarine optical cable causes disturbances, mainly due to the impact of the ship anchor on the rigid bottom sediment seabed when it falls, which causes the seabed to vibrate. Therefore, we can simulate the anchor dropping vibration according to the characteristics of the foundation vibration caused by dynamic compaction hitting the ground. Scott C.R. and Pearce R.W. predicted the response of the foundation during strong impact and gave the motion equation and the calculation formula for the average stress at the contact surface between the soil and the rammer as shown in Formulas (4) and (6) respectively:

[0057] Formula (5),

[0058] Formula (6);

[0059] Where:

[0060] Formula (7),

[0061] Formula (8),

[0062] In the above formulas, C and K represent the stiffness of the rammer and the viscous pot damping in the equivalent lumped system respectively, a represents the radius of the rammer, G, ρ , μ represent the shear modulus, density, and Poisson's ratio of the elastic half-space of the ground respectively, ω is the angular frequency, and CV is the dilatational wave velocity of the ground.

[0063] Through experiments, by analyzing the monitored acceleration of the ground vibration, the attenuation expression of its acceleration a is obtained, as shown in Formula (9).

[0064] Formula (9),

[0065] In the formula, L is the distance from the monitoring point to the impact point. According to the above theory, the vibration waveform generated when the ship anchor impacts the rigid bottom sediment seabed can be obtained as Figure 9 shown. It can be seen that the vibration frequency generated when the ship anchor impacts the rigid bottom sediment seabed is concentrated in the range of 10 - 30 Hz.

[0066] B. Submarine Geological Movements

[0067] By obtaining seismic data from publicly available seismic detection units, typical time-domain and frequency-domain waveforms of seismic waves can be obtained as shown in Figure 10 and Figure 11 . It can be seen from the figure that the frequency of the seismic waves is concentrated within 10 Hz.

[0068] C. Ship Navigation Disturbance

[0069] The disturbance of ship navigation to submarine optical cables is mainly caused by the waves generated during ship navigation. Submarine optical cables in shallow waters are buried, and it is difficult for ship navigation to cause disturbance to them. Therefore, the disturbance of ship waves to submarine optical cables should mainly be the disturbance of the ship waves formed by submersibles to deep-sea optical cables.

[0070] In the Load Code for Hydraulic Structures, the wave height formula for ship waves is given as:

[0071] Formula (10),

[0072] where represents the allowable navigation speed of the ship as required ( ), is the block coefficient of the ship, represents the relationship between the water depth of the waterway and the navigation speed:

[0073] Formula (11),

[0074] where H represents the water depth of the waterway and g is the acceleration due to gravity. According to the above principle, the impact of ship navigation on the Hong Kong-Zhuhai-Macao Bridge Island-Tunnel Project is studied. Calculations show that when the water depth of the waterway is 15 m, for a 100,000-ton ship with a block coefficient of 0.74 and a navigation speed of 15 knots, the wave height generated at a distance of 250 m from the ship can still reach 0.92 m. Research shows that the main factor affecting the wave height of ship waves is the navigation speed. This research shows that in the deep-sea area (assuming a depth of 500 m), when the submersible reaches a certain speed, ship waves will cause disturbance to submarine optical cables. According to the research, the sine wave potential expression of ship waves is:

[0075] Formula (12),

[0076] where A is the amplitude of the sine wave, g is the acceleration due to gravity, k is the number of waves, and d is the water depth. Based on this principle, assuming that the ship waves generated by the submersible can cause disturbance to the submarine optical cable, the force waveform on the submarine optical cable should be as shown in Figure 12 . According to the above analysis and Figure 12 , it can be known that the disturbance frequency of ship waves to submarine optical cables is in the range of 0.5 - 1 Hz.

[0077] d. Background noise in the deep and far sea

[0078] Deep - sea noise has many sound sources, complex sound - generating mechanisms, and the combined action of multiple noises, resulting in a relatively wide frequency range. According to the noise sources, it mainly includes ship noise, seismic disturbances, ocean turbulence, wind - generated noise, low - frequency noise caused by tides and waves, and thermal noise. In addition, marine organisms also produce noise.

[0079] In the ocean, the propagation distance of noise is related to the frequency of the noise. The higher the frequency, the faster the attenuation and the shorter the propagation distance. The following table gives the frequency ranges of the noises caused by the above - mentioned noise sources and their impacts on the detection system.

[0080] Table 1 Comparison of deep - sea noise source frequencies

[0081]

[0082] According to the above analysis, ship noise, including the noise of submersibles and underwater robots, is the disturbance source that we are concerned about. This type of noise is mainly low - frequency and has a relatively long propagation distance, up to 100 km. When using passive sonar to detect and identify ships, it is considered that when the distance is between 10 km and 100 km, the noise generated by the ship is background noise; when the distance is within 10 km, it is a short - distance situation and noise processing is not suitable. Considering the actual water - depth environment of deep - sea disturbance detection, when the submerging distance of submersibles and underwater robots is within 1 km, it should be considered as a disturbance and dealt with as an eavesdropping threat or potential threat. In addition, submersibles and underwater robots also generate mechanical noise, propeller noise, and hydrodynamic noise. When the distance is close enough, these noises will cause disturbances to the submarine cable. During eavesdropping, when stripping the outer sheath and armor layer of the submarine cable, the stripping equipment used will also cause vibration, and due to direct contact with the submarine cable, the disturbance is stronger.

[0083] In summary, the characteristics of disturbance events are compared as shown in the following table (Table 2):

[0084]

[0085] Reference Figures 6 to 12, which respectively show the waveform diagrams obtained from the above time-frequency domain information process. Through the pattern recognition of the Φ-OTDR optical fiber disturbance event based on the square difference, short-time overlevel rate, short-time Fourier transform, and duration characteristics of the waveform, the category (classification) of the disturbance event or disturbance cause corresponding to each waveform diagram can be obtained. Taking these categories as the labels or targets in the neural network training process, the pattern recognition accuracy and robustness can be further improved through the neural network. In view of this, in step S200 of some embodiments of the present invention, the convolutional neural network includes an input layer, four parallel convolutional layers, a pooling layer, a fusion layer, a fully connected layer, and an output layer. The input layer is used to preprocess the obtained time domain, frequency domain, energy, and duration information of the target submarine optical cable vibration and convert it into multi-dimensional data. The four parallel convolutional layers are used to extract features in the time domain, frequency domain, energy, and duration dimensions from the multi-dimensional data respectively.

[0086] Specifically, the convolutional layer mainly uses convolutional filters to complete the feature extraction of the input and outputs a feature map. The pooling layer is used to reduce the complexity of the deep network and output a deep feature map. Common pooling methods include max pooling, mean pooling, and stochastic pooling. The fully connected layer is responsible for combining the effective features of the convolutional layer and the pooling layer to improve the high-dimensional feature inference ability. The classification layer uses a classifier to obtain the recognition result from the data feature output. Common classifiers include SVM, Softmax, AdaBoost, etc. The key to the implementation of the convolutional neural network is the determination of the parameters of each layer. The key at this stage is the interaction between the convolutional layer and the pooling layer. The specific training process of the CNN is shown in the following table:

[0087]

[0088] Formula (13),

[0089] Formula (14),

[0090] Among them, is the learning sample, is the learning sample The result after the convolutional filtering operation, is the weight matrix of the CNN, and the activation function is the activation function, The final classification output.

[0091] In some embodiments of the present invention, the convolutional neural network is trained through the following steps: obtaining different types of disturbance signals of submarine optical cable vibration and their simulation signals; using the disturbance signals and the simulation signals as samples, and using the disturbance category corresponding to each sample as a label to construct a data set; using the data set to train the convolutional neural network until its error tends to be stable and is lower than a threshold, and obtaining a trained convolutional neural network.

[0092] According to the time characteristics shown in Table 2, MATLAB is used to generate four types of disturbance signals, namely Ship, Uvehicle, Mcutting, and Mlbite. These four types of signals are all non-stationary signals, and their relevant characteristics are shown in Table 3. 1000 groups of data are generated by MATLAB. The first 900 groups are used for learning and training, and the last 100 groups are used for testing. The test results are statistically analyzed and shown in Table 4. Through simulation testing, the convolutional neural network has a good recognition effect on the disturbance event signals of submarine optical cables. However, the disturbance signals in reality are affected by the deep-sea background noise and are more complex. Therefore, the method may need to be appropriately adjusted and improved.

[0093] Table 3 Disturbance event characteristics

[0094]

[0095] Table 4 Recognition results of convolutional neural network for simulated events

[0096]

[0097] Embodiment 2

[0098] Reference Figure 14 In the second aspect of the present invention, a submarine optical cable disturbance recognition system 1 based on time-frequency-space and neural network is provided, including: an acquisition module 11 for acquiring the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable; an identification module 12 for inputting the time domain, frequency domain, energy, and duration information of the vibration of the target submarine optical cable into the trained convolutional neural network to obtain the disturbance category of the target submarine optical cable.

[0099] Further, the neural network includes: an acquisition unit for acquiring different types of disturbance signals of submarine optical cable vibration and their simulation signals; a construction unit for using the disturbance signals and the simulation signals as samples, and using the disturbance category corresponding to each sample as a label to construct a data set; a training unit for using the data set to train the convolutional neural network until its error tends to be stable and is lower than a threshold, and obtaining a trained convolutional neural network.

[0100] Embodiment 3

[0101] Reference Figure 15, in a third aspect of the present invention, there is provided an electronic device, comprising: one or more processors; a storage device for storing one or more programs, which when executed by the one or more processors cause the one or more processors to implement the method of the first aspect of the present invention.

[0102] The electronic device 500 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 501, which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0103] Generally, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 15 the electronic device 500 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 15 Each block shown in may represent a device or, as required, multiple devices.

[0104] In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present disclosure includes a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program code for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network via the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the methods of the embodiments of the present disclosure are performed. It should be noted that the computer-readable medium described in the embodiments of the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiments of the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the embodiments of the present disclosure, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0105] The above-mentioned computer-readable medium can be included in the above-mentioned electronic device; or it can exist separately and not be assembled into the electronic device. The above-mentioned computer-readable medium carries one or more computer programs, and when the above-mentioned one or more programs are executed by the electronic device, the electronic device is caused to:

[0106] Computer program code for performing the operations of the embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, Python, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0107] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0108] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for identifying submarine optical cable disturbances based on time-frequency-space and neural networks, characterized in that, Including: Obtain the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable: The time-frequency-space includes time, frequency domain, and space; Among them, the analysis of the spatial dimension in the spatial information includes the spatial distance of the perturbed optical fiber, the vertical position of the perturbation source from the optical fiber, and the moving speed of the perturbation source in the length direction of the optical fiber; the spatial distance of the perturbed optical fiber is calculated by the following method: , L d represents the length of the perturbed optical fiber, k represents the number of pulses detecting the perturbation, c is the speed of light in vacuum, n is the refractive index of the optical fiber, W pluse represents the single-frequency pulse time width, W false represents the filling pseudo-pulse time width; the vertical distance of the optical fiber is calculated by the following method: , v is the propagation speed of sound in seawater; the moving speed of the perturbation source is calculated by the following method: t5 and t7 respectively represent the initial time and the end time of the optical fiber perturbation, z 5 and z 7 respectively represent the initial position and the end position; Use multiple long-time detections of detection pulses with multiple frequencies combined to obtain the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable; The time-domain information includes: the duration of the disturbance time and the chronological relationship of the disturbance event time; The frequency-domain information includes the response frequency of the system and the frequency-domain characteristics extracted from the spectral distribution of different disturbance signals; Input the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable into the trained convolutional neural network to obtain the disturbance category of the target submarine optical cable.

2. The method for identifying submarine optical cable disturbances based on time-frequency-space and neural network according to claim 1, wherein, The convolutional neural network includes an input layer, four parallel convolutional layers, a pooling layer, a fusion layer, a fully connected layer, and an output layer. The input layer is used to preprocess the obtained time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable and convert it into multi-dimensional data. The four parallel convolutional layers are used to extract features in the time-domain, frequency-domain, energy, and duration dimensions respectively from the multi-dimensional data.

3. The method for identifying submarine optical cable disturbances based on time-frequency-space and neural network according to claim 2, characterized in that, The convolutional neural network is a one-dimensional convolutional neural network.

4. The method for identifying submarine optical cable disturbances based on time-frequency-space and neural network according to claim 1, characterized in that The convolutional neural network is trained through the following steps: Obtain different types of disturbance signals of the submarine optical cable vibration and their simulation signals; Use the disturbance signals and the simulation signals as samples, and use the disturbance category corresponding to each sample as a label to construct a data set; Use the data set to train the convolutional neural network until its error tends to be stable and is lower than the threshold to obtain the trained convolutional neural network.

5. The method for identifying submarine optical cable disturbances based on time-frequency-space and neural network according to claim 4, characterized in that The disturbance categories include traveling wave disturbance of a vehicle, traveling wave disturbance of a submersible, mechanical cutting disturbance, and marine organism biting disturbance.

6. An undersea optical cable disturbance recognition system based on time-frequency-space and neural network, characterized in that, Including: An acquisition module, used to obtain the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable: The time-frequency-space includes time, frequency domain, and space; Among them, the analysis of the spatial dimension in the spatial information includes the spatial distance of the disturbed optical fiber, the vertical position of the disturbance source from the optical fiber, and the moving speed of the disturbance source in the length direction of the optical fiber; the spatial distance of the disturbed optical fiber is calculated by the following method: , L d represents the length of the disturbed optical fiber, k represents the number of pulses detecting the disturbance, c is the speed of light in vacuum, n is the refractive index of the optical fiber, W pluse represents the single-frequency pulse time width, W false represents the filling pseudo-pulse time width; the vertical distance of the optical fiber is calculated by the following method: , v is the propagation speed of sound in seawater; the moving speed of the disturbance source is calculated by the following method: , t 5 and t 7 respectively represent the initial time and the end time of the optical fiber disturbance, z 5 and z 7 respectively represent the initial position and the end position; Input the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable into the trained convolutional neural network to obtain the disturbance category of the target submarine optical cable; An identification module, which inputs the time-domain, frequency-domain, energy, and duration information of the vibration of the target submarine optical cable into the trained convolutional neural network to obtain the disturbance category of the target submarine optical cable.

7. The undersea optical cable disturbance identification system based on time-frequency-space and neural network according to claim 6, characterized in that The neural network includes: An acquisition unit, used to obtain different types of disturbance signals of the submarine optical cable vibration and their simulation signals; A construction unit, used to use the disturbance signals and the simulation signals as samples, and use the disturbance category corresponding to each sample as a label to construct a data set; A training unit, used to use the data set to train the convolutional neural network until its error tends to be stable and is lower than the threshold to obtain the trained convolutional neural network.

8. An electronic device, comprising: One or more processors; A storage device, used to store one or more programs, characterized in that when the one or more programs are executed by the one or more processors, the one or more processors implement the method for identifying submarine optical cable disturbances based on time-frequency-space and neural network according to any one of claims 1 to 5.

9. A computer-readable medium having a computer program stored thereon, characterized in that, Wherein, When the computer program is executed by a processor, it implements the submarine optical cable disturbance identification method based on time-frequency-space and neural network according to any one of claims 1 to 5.

Images