CN115014571A - Submarine cable risk event identification system and method - Google Patents
Submarine cable risk event identification system and method Download PDFInfo
- Publication number
- CN115014571A CN115014571A CN202210574302.4A CN202210574302A CN115014571A CN 115014571 A CN115014571 A CN 115014571A CN 202210574302 A CN202210574302 A CN 202210574302A CN 115014571 A CN115014571 A CN 115014571A
- Authority
- CN
- China
- Prior art keywords
- submarine cable
- characteristic information
- risk event
- information
- grating
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 239000013307 optical fiber Substances 0.000 claims abstract description 70
- 238000013528 artificial neural network Methods 0.000 claims abstract description 52
- 238000004088 simulation Methods 0.000 claims abstract description 35
- 230000000694 effects Effects 0.000 claims abstract description 15
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 238000002474 experimental method Methods 0.000 claims abstract description 7
- 239000000835 fiber Substances 0.000 claims description 65
- 239000013598 vector Substances 0.000 claims description 40
- 238000013527 convolutional neural network Methods 0.000 claims description 29
- 238000006073 displacement reaction Methods 0.000 claims description 29
- 230000015654 memory Effects 0.000 claims description 24
- 238000012549 training Methods 0.000 claims description 19
- 239000011159 matrix material Substances 0.000 claims description 16
- 244000035744 Hura crepitans Species 0.000 claims description 15
- 239000004677 Nylon Substances 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 229920001778 nylon Polymers 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 13
- 238000004891 communication Methods 0.000 claims description 11
- 229910001285 shape-memory alloy Inorganic materials 0.000 claims description 11
- 230000006403 short-term memory Effects 0.000 claims description 11
- 230000007787 long-term memory Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 239000013535 sea water Substances 0.000 claims description 8
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 239000005995 Aluminium silicate Substances 0.000 claims description 3
- 230000001133 acceleration Effects 0.000 claims description 3
- 235000012211 aluminium silicate Nutrition 0.000 claims description 3
- NLYAJNPCOHFWQQ-UHFFFAOYSA-N kaolin Chemical compound O.O.O=[Al]O[Si](=O)O[Si](=O)O[Al]=O NLYAJNPCOHFWQQ-UHFFFAOYSA-N 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000011664 signaling Effects 0.000 claims description 3
- 210000004027 cell Anatomy 0.000 description 11
- 230000006872 improvement Effects 0.000 description 8
- 238000010606 normalization Methods 0.000 description 7
- 230000004913 activation Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 6
- 238000011176 pooling Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 230000001795 light effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 239000010802 sludge Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K11/00—Measuring temperature based upon physical or chemical changes not covered by groups G01K3/00, G01K5/00, G01K7/00 or G01K9/00
- G01K11/32—Measuring temperature based upon physical or chemical changes not covered by groups G01K3/00, G01K5/00, G01K7/00 or G01K9/00 using changes in transmittance, scattering or luminescence in optical fibres
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/16—Measuring arrangements characterised by the use of optical techniques for measuring the deformation in a solid, e.g. optical strain gauge
- G01B11/165—Measuring arrangements characterised by the use of optical techniques for measuring the deformation in a solid, e.g. optical strain gauge by means of a grating deformed by the object
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H9/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means
- G01H9/004—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means using fibre optic sensors
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses a submarine cable risk event identification system and method, which comprises the following steps: constructing a fault simulation experiment model of the submarine cable, and carrying out an experiment on a risk event of the submarine cable to obtain characteristic information when the submarine cable breaks down; establishing a submarine cable risk event recognition model based on a hybrid neural network according to characteristic information when a submarine cable fails; and judging whether the submarine cable at the current moment has a fault or not through the submarine cable risk event recognition model according to the characteristic information of the submarine cable at the current moment, recognizing the category of the risk event corresponding to the fault of the submarine cable at the current moment, and generating a fault analysis report of the submarine cable. According to the submarine cable fault early warning system, the submarine cable is monitored in real time through the plurality of quasi-distributed grating optical fiber sensors, fault judgment of the submarine cable and recognition of a submarine cable risk event are achieved through the submarine cable risk event recognition model, and the submarine cable fault early warning effect is achieved.
Description
Technical Field
The invention belongs to the technical field of ocean observation, and particularly relates to a submarine cable risk event identification system and method.
Background
Submarine cables have become a bridge link and focus of attention for long-distance cross-sea energy transmission and large-scale utilization of new energy. The method is a core and key for solving the problems of future length, depth and precision offshore engineering operation by establishing a standard system of the submarine cable, developing a novel manufacturing technology of the submarine cable, and developing a construction technology and a high-precision detection technology under the condition of high sea condition. The detection technology for improving the risk and the fault of the submarine cable is an important part, so the development of the submarine cable safety detection technology has great strategic significance for supporting and guaranteeing energy safety and social and economic sustainable development.
After the submarine cable is laid, the cable is damaged both artificially and naturally, and operation safety problems are easy to occur. According to historical data statistics, human activities cause over 90% of submarine cable failures, and one third of them are anchor hazards. The IEEE submarine cable laying specification also ranks the anchor hazard as the first of man-made disasters. The damage mode of the ship anchor to the submarine cable mainly has two types: firstly, a ship anchor falling at a certain speed penetrates through a sludge layer buried above the submarine cable, and the submarine cable is directly punctured; firstly, the ship anchor penetrates into the seabed around the submarine cable, and the submarine cable is pulled apart by hooking the submarine cable when the anchor is pulled out. ) Natural damage includes: displacement and oscillation caused by ocean currents and tides, abrasion caused by friction between cable protective pipes and rocks and cable bodies, seawater erosion and the like.
Because of the complexity and difficulty of the practical operation of the submarine cable risk event simulation experiment, many scholars usually use simulation software to perform risk event simulation analysis during the process of doing the submarine cable risk event simulation experiment, and do not perform the practical experiment, so that the research on a submarine cable risk event simulation method and a device for performing the simulation experiment is particularly important, which is beneficial to simulating the practical working environment of the submarine cable and laying a good foundation for the subsequent classified identification and analysis of risk faults.
Fiber Bragg Grating (FBG) sensors have rapidly become excellent sensor elements capable of measuring various physical quantities such as temperature, strain and pressure. It has the advantages of high sensitivity, no electromagnetic interference, good waterproof performance, small volume, light weight, high reliability, and being capable of being embedded into composite materials. Therefore, the method is widely applied to various fields, such as bridge engineering, aerospace, ocean monitoring, electric power engineering, biomedicine and the like.
Convolutional Neural Networks (CNNs) are a type of feed-forward Neural network that includes Convolutional calculation and has a deep structure, has a capability of characterizing learning, can perform supervised learning and unsupervised learning, and is commonly used for feature extraction and classification tasks; a Long Short-Term Memory neural network (Long Short-Term Memory) is a time-recursive neural network suitable for processing and predicting relatively Long-spaced and delayed events in a time series.
Disclosure of Invention
The present invention is directed to a submarine cable risk event identification system and method, which solves one or more of the problems of the prior art and provides at least one of the advantages of the present invention.
The solution of the invention for solving the technical problem is as follows: there is provided a submarine cable risk event identification system comprising: the system comprises a fault simulation experiment model, an optical fiber sensing unit, an underwater acoustic communication machine, a grating demodulator, a processing unit, a reconstruction unit and a risk event identification unit;
the fault simulation experiment model is used for simulating corresponding risk events when the submarine cable breaks down;
the optical fiber sensing unit is fixed along the axial direction of the submarine cable and is used for sensing first field information of the submarine cable when a fault occurs and second field information of the submarine cable at the current moment, and the types of the field information comprise temperature field information, strain field information, vibration field information and displacement field information;
the underwater acoustic communication machine is placed in a fault simulation experiment model and connected with the optical fiber demodulator, and the optical fiber sensing unit transmits the first field information or the second field information to the optical fiber demodulator in the form of optical fiber data through the underwater acoustic communication machine;
the grating demodulator is used for demodulating optical fiber data corresponding to the first field information to obtain a first reflected wave wavelength and a variable quantity thereof of the optical fiber sensing unit;
the processing unit is used for obtaining temperature characteristic information, vibration characteristic information and strain characteristic information when the submarine cable fails according to the wavelength of the first reflected wave and the variation of the wavelength;
the reconstruction unit is used for reconstructing the shape of the submarine cable with the fault through a reconstruction algorithm to obtain displacement characteristic information when the submarine cable has the fault;
the risk event identification unit is used for establishing a submarine cable risk event identification model based on a hybrid neural network according to the temperature characteristic information, the vibration characteristic information, the strain characteristic information and the displacement characteristic information;
the grating demodulator is further used for demodulating optical fiber data corresponding to the second field information to obtain a second reflected wave wavelength and a variable quantity thereof of the optical fiber sensing unit;
the processing unit is used for obtaining current temperature information, current vibration information and current strain information of the submarine cable at the current moment according to the wavelength of the second reflected wave and the variation of the wavelength;
the reconstruction unit is used for reconstructing the shape of the submarine cable at the current moment through a reconstruction algorithm to obtain the current displacement information of the submarine cable at the current moment;
and the risk event identification unit is also used for judging whether the submarine cable at the current moment has a fault or not according to the current temperature information, the current vibration information, the current strain information and the current displacement information through the submarine cable risk event identification model, identifying the type of the risk event corresponding to the fault of the submarine cable, and generating a fault analysis report of the submarine cable.
As a further improvement of the above technical solution, the grating sensing unit includes a shape memory alloy wire, a fiber grating vibration sensor, a fiber grating strain sensor and a fiber grating temperature sensor, the fiber grating vibration sensor, the fiber grating strain sensor and the fiber grating temperature sensor are all quasi-distributed fiber grating sensors, the fiber grating strain sensor and the fiber grating temperature sensor are all mounted on the shape memory alloy wire, and the shape memory alloy wire and the fiber grating vibration sensor are carried on the submarine cable along the axial direction of the submarine cable.
As a further improvement of the above technical solution, the fault simulation experiment model includes a sandbox, a model support, a nylon rope, a key module, a control module and a driving module, and the control module is respectively connected with the key module and the driving module; the model bracket is a scale rod with a graduated scale, and the zero point at the bottom of the scale rod and the submarine cable are positioned on the same horizontal plane; the model support is placed in the sandbox, one end of the model support is fixed with the sandbox, the other end of the model support is connected with the fixed pulley, the nylon rope is connected with the fixed pulley in a sliding mode, the ship anchor is tied at one end of the nylon rope, and the other end of the nylon rope is connected with the driving module;
the sandbox is used for simulating the environment of the submarine cable in actual use, kaolin and seawater are filled in the sandbox, and the underwater acoustic communication machine is placed in the seawater;
the driving module includes: the ship anchor driving device comprises a driving support, a guide rail sliding block, a conveyor belt and a stepping motor, wherein the driving support is positioned at the bottom of a model support, the conveyor belt is arranged on two sides of the driving support, the guide rail sliding block is arranged on the conveyor belt and is connected with one end, far away from a ship anchor, of a nylon rope, and the stepping motor is used for driving the conveyor belt to transmit along the horizontal direction so as to drive the guide rail sliding block to move along the horizontal direction;
the control module is used for detecting the condition that the key module is pressed down and outputting a first control signal or a second control signal to the driving module according to the condition that the key module is pressed down; the driving module is used for controlling the guide rail sliding block to move leftwards along the horizontal direction according to the first control signal and controlling the guide rail sliding block to move rightwards along the horizontal direction according to the second control signal;
the key module comprises a start key, a pause key, a left shift key and a right shift key, wherein the start key is used for starting the stepping motor, the pause key is used for closing the stepping motor, the left shift key is used for controlling the stepping motor to rotate forwards, and the right shift key is used for controlling the stepping motor to rotate backwards.
A submarine cable risk event identification method is applied to a submarine cable risk event identification system and comprises the following steps:
s100, constructing a fault simulation experiment model of the submarine cable, and performing an experiment on a risk event of the submarine cable to obtain characteristic information when the submarine cable breaks down;
wherein the characteristic information when the submarine cable fails comprises: temperature characteristic information, vibration characteristic information, strain characteristic information and displacement characteristic information;
s200, establishing a submarine cable risk event recognition model based on a hybrid neural network according to characteristic information when a submarine cable fails;
s300, judging whether the submarine cable at the current moment has a fault or not through the submarine cable risk event recognition model according to the characteristic information of the submarine cable at the current moment, recognizing the type of a risk event corresponding to the fault of the submarine cable at the current moment, generating a fault analysis report of the submarine cable, and sending out early warning according to the type of the risk event;
wherein the characteristic information of the submarine cable at the current moment comprises: current temperature information, current vibration information, current strain information, and current displacement information.
As a further improvement of the above technical solution, the S100 specifically is:
s110, constructing a fault simulation experiment model according to a certain scaling ratio according to the environment of the submarine cable in actual conditions;
s120, carrying an optical fiber sensing unit along the submarine cable, placing the submarine cable under the model bracket, and adjusting the model bracket to ensure that the zero point at the bottom of the model bracket and the submarine cable are in the same plane;
s130, respectively carrying out simulation experiments of an anchor event, a hook event and an ocean current event of the submarine cable to obtain characteristic information when the submarine cable breaks down.
As a further improvement of the above technical solution, the obtaining of the characteristic information when the submarine cable fails further includes the following steps:
s131, sensing field information of the submarine cables in different risk events through an optical fiber sensing unit, and transmitting the field information of the submarine cables to a grating demodulator through an underwater acoustic communication machine in the form of optical fiber data;
s132, demodulating the optical fiber data through a grating demodulator to obtain the reflected wave wavelength variation of the optical fiber sensing unit;
s133, calculating the temperature characteristic information, the strain characteristic information and the vibration characteristic information according to the wavelength variation of the reflected wave;
wherein the temperature characteristic information satisfies the following formula:
wherein μ represents the temperature characteristic information, Δ λ B Indicating a first reflected wave wavelength variation, λ, of the grating fiber temperature sensor B An initial quantity, a, representing a first reflected wave wavelength of the grating fiber temperature sensor 1 B is the thermal expansion coefficient of the grating optical fiber temperature sensor, and b is the thermo-optic effect coefficient of the grating optical fiber temperature sensor;
the strain characteristic information satisfies the following formula:
wherein ε represents the strain characteristic information, Δ λ B Indicating the variation of the wavelength of the first reflected wave, λ, of the grating fiber strain sensor B Representing said optical fiberInitial quantity, P, of the first reflected wave wavelength of the strain sensor e The effective elastic-optical coefficient of the grating optical fiber strain sensor;
the vibration characteristic information satisfies the following formula:
wherein S represents the vibration characteristic information, k eff Is equivalent stiffness, L, of the fiber grating vibration sensor f Is the package length of the fiber grating vibration sensor, a is the acceleration, m is the mass of the submarine cable, and Delta lambda B Indicating a variation of a wavelength of a first reflected wave, λ, of the grating fiber vibration sensor B An initial quantity, P, representing a first reflected wave wavelength of the grating fiber vibration sensor e The elastic-optical coefficient of the grating optical fiber vibration sensor is obtained;
s134, reconstructing the shape of the submarine cable through a reconstruction algorithm, and comparing the reconstructed shape of the submarine cable with the shape of the submarine cable before reconstruction to obtain the displacement characteristic information.
As a further improvement of the above technical solution, step S200 includes the following steps:
s210, normalizing the characteristic information when the submarine cable fails, and converting the format of the characteristic information when the submarine cable fails into a format which meets the input conditions of a convolutional neural network and a long-short term memory neural network;
s220, dividing feature information of the submarine cable in fault into a training set and a testing set according to a proportion, inputting the training sets into the convolutional neural network and the long-short term memory neural network respectively, training the convolutional neural network and the long-short term memory neural network, outputting a first feature vector by the convolutional neural network, and outputting a second feature vector by the long-short term memory neural network;
s230, setting a splicing layer, and splicing the first feature vector and the second feature vector through the splicing layer to generate high-level risk features;
s240, setting a full connection layer and a Softmax classifier, inputting the high-level risk features into the full connection layer and the Softmax classifier, and generating a submarine cable risk event recognition model and a weight value thereof;
and S250, testing the performance of the submarine cable risk event recognition model through the test set.
As a further improvement of the above technical solution, in step S220, after the training of the convolutional neural network is finished, a first feature matrix is output, where the first feature matrix is a two-dimensional matrix, a stretching layer is arranged, and the first feature matrix is stretched into a first feature vector through the stretching layer, where the first feature vector is a one-dimensional feature vector.
As a further improvement of the above technical solution, the step S200 further includes: and optimizing the hyper-parameters of the long-term and short-term memory neural network by a Bayesian optimization algorithm.
As a further improvement of the above technical solution, the step S200 further includes: calculating a loss function of the submarine cable risk event identification model, and optimizing the weight of the submarine cable risk event identification model through an Adam algorithm.
The invention has the beneficial effects that: the invention discloses a submarine cable risk event recognition system and method, which are characterized in that a fault simulation experiment model is used for acquiring characteristic information when a submarine cable fails, so that the characteristic information is closer to the actual situation, and the uncertainty of the characteristic information is reduced; the method comprises the steps of monitoring the condition of the submarine cable in real time through a plurality of quasi-distributed grating optical fiber sensors, establishing a submarine cable risk event recognition model through a hybrid neural network, judging the fault of the submarine cable, recognizing the type of a risk event corresponding to the fault of the submarine cable, and giving an alarm when the submarine cable is dangerous to a certain extent so as to achieve the effect of fault early warning.
Drawings
In order to more clearly illustrate the technical solution in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below. It is clear that the described figures are only some embodiments of the invention, not all embodiments, and that a person skilled in the art can also derive other designs and figures from them without inventive effort.
Fig. 1 is a flow chart of a submarine cable risk event identification method;
fig. 2 is a schematic structural diagram of a fault simulation experiment model of a submarine cable risk event recognition system;
FIG. 3 is a block schematic diagram of a submarine cable risk event identification system;
FIG. 4A is a cross-sectional view of a fiber optic sensing unit of a submarine cable risk event identification system piggybacked onto a submarine cable;
FIG. 4B is a schematic diagram of the fiber grating strain sensor, fiber grating temperature sensor, shape memory alloy wire configuration of a submarine cable risk event identification system;
FIG. 4C is a cross-sectional view of a fiber grating strain sensor, a fiber grating temperature sensor, a shape memory alloy wire of a submarine cable risk event identification system;
FIG. 5 is a flow chart of a method for identifying submarine cable risk events for constructing a fault simulation experimental model;
fig. 6 is a flow chart of a submarine cable risk event identification method for obtaining characteristic information of a submarine cable in the event of a fault;
fig. 7 is a flow chart of a submarine cable risk event identification method for building a submarine cable risk event identification model;
fig. 8 is a structural flowchart of a submarine cable risk event recognition model of a submarine cable risk event recognition method.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
It is noted that while a division of functional blocks is depicted in the system diagram, and logical order is depicted in the flowchart, in some cases the steps depicted and described may be performed in a different order than the division of blocks in the system or the flowchart. The terms first, second and the like in the description and in the claims, and the drawings described above, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
Referring to fig. 1 to 8, the present disclosure provides a submarine cable risk event identification system, wherein a first embodiment of the submarine cable risk event identification system comprises: the system comprises a fault simulation experiment model 100, an optical fiber sensing unit 200, an underwater acoustic communication machine, a grating demodulator 300, a processing unit 400, a reconstruction unit 500 and a risk event identification unit 600;
the fault simulation experiment model 100 is used for simulating a corresponding risk event of the submarine cable 700 when a fault occurs, so as to acquire characteristic information of the submarine cable 700 when the fault occurs;
the optical fiber sensing unit 200 is fixed along the axial direction of the submarine cable 700, the optical fiber sensing unit 200 is used for sensing first field information of the submarine cable 700 when a fault occurs and second field information of the submarine cable 700 at the current moment in real time, and the field information is temperature field information, strain field information, vibration field information and displacement field information;
the underwater acoustic communicator is placed in the fault simulation experiment model 100 and connected with the optical fiber demodulator 300, and the optical fiber sensing unit 200 transmits the first field information or the second field information to the optical fiber demodulator 300 in the form of optical fiber data through the underwater acoustic communicator;
the grating demodulator 300 is configured to demodulate optical fiber data corresponding to the first field information to obtain a first reflected wave wavelength of the optical fiber sensing unit 200 and a variation thereof;
the processing unit 400 is configured to obtain temperature characteristic information, vibration characteristic information and strain characteristic information of the submarine cable 700 when a fault occurs according to the first reflected wave wavelength and the variation thereof;
the reconstruction unit 500 is configured to reconstruct the shape of the submarine cable 700 with a fault through a reconstruction algorithm, so as to obtain displacement characteristic information of the submarine cable 700 with the fault;
the risk event identification unit 600 is configured to establish a submarine cable risk event identification model based on a hybrid neural network according to the temperature characteristic information, the vibration characteristic information, the strain characteristic information, and the displacement characteristic information;
the grating demodulator 300 is further configured to demodulate light data corresponding to the second field information to obtain a second reflected wave wavelength of the optical fiber sensing unit and a variation thereof;
the processing unit 400 is configured to obtain current temperature information, current vibration information, and current strain information of the submarine cable 700 at the current time according to the wavelength of the second reflected wave and the variation thereof;
the reconstruction unit 500 is configured to reconstruct the shape of the submarine cable 700 at the current time by using a reconstruction algorithm, and obtain current displacement information of the submarine cable 700 at the current time;
the risk event identification unit 600 is further configured to determine whether the submarine cable 700 fails at the current moment according to the current temperature information, the current vibration information, the current strain information, and the current displacement information through the submarine cable risk event identification model, identify the category of the risk event corresponding to the failure of the submarine cable 700, and generate a failure analysis report of the submarine cable 700.
Further, referring to fig. 2, the fault simulation experiment model 100 includes a sandbox 110, a model support 120, a nylon rope 130, a key module, a control module and a driving module, wherein the control module is respectively connected to the key module and the driving module; the method specifically comprises the following steps:
the sandbox 110 is used for simulating the environment of the submarine cable in actual use, kaolin and seawater are filled in the sandbox 110, and the underwater acoustic communicator is placed in the seawater;
the model support 120 is a scale rod with a graduated scale, and the zero point at the bottom of the scale rod and the submarine cable are positioned on the same horizontal plane; the model bracket 120 is placed in the sandbox 110, one end of the model bracket 120 is fixed with the sandbox, the other end of the model bracket 120 is connected with the fixed pulley 140, the nylon rope 130 is connected with the fixed pulley 140 in a sliding manner, one end of the nylon rope 130 is tied with the ship anchor 150, and the other end of the nylon rope 130 is connected with the driving module;
the driving module includes: the driving support 160 is located at the bottom of the model support 120, the conveyor belts 180 are installed on two sides of the driving support 160, the guide rail sliders 170 are placed on the conveyor belts 180, the guide rail sliders 170 are connected with one ends, away from the boat anchor 150, of the nylon ropes 130, and the stepping motor is used for driving the conveyor belts 180 to transmit along the horizontal direction so as to drive the guide rail sliders 170 to move along the horizontal direction, so that the effect of controlling the boat anchor 150 to fall and be pulled up is achieved.
The control module is used for detecting the condition that the key module is pressed down and outputting a first control signal or a second control signal to the driving module according to the condition that the key module is pressed down; the driving module is used for controlling the guide rail sliding block 170 to move leftwards along the horizontal direction according to a first control signal and controlling the guide rail sliding block 170 to move rightwards along the horizontal direction according to a second control signal.
The key module includes: the device comprises a start key, a pause key, a left shift key and a right shift key, wherein the start key is used for starting the stepping motor, the pause key is used for closing the stepping motor, the left shift key is used for controlling the stepping motor to rotate forwards, and the right shift key is used for controlling the stepping motor to rotate backwards.
In this embodiment, the rotation direction of the stepping motor determines the direction of the movement of the rail slider 170 in the horizontal direction. When the stepping motor rotates forward, the guide rail sliding block 170 moves leftwards along the horizontal direction, so that the boat anchor 150 falls; when the stepping motor rotates reversely, the guide rail sliding block 170 moves to the right along the horizontal direction, and the ship anchor 150 is pulled up, so that the effect of simulating the risk event of the submarine cable 700 is achieved.
According to the invention, the characteristic information of the submarine cable when the submarine cable fails is obtained through the fault simulation experiment model, the fault simulation experiment model is constructed according to the environment where the submarine cable is located in the actual situation, and the occurrence of the risk event is simulated through software in a contrast manner. In addition, the fault simulation experiment model is low in cost, and excessive manpower and material resources are not required to be consumed.
Further, referring to fig. 4A, 4B and 4C, the grating sensing unit 200 includes a shape memory alloy wire 210, a fiber grating vibration sensor 220, a fiber grating strain sensor 230 and a fiber grating temperature sensor 240, the fiber grating vibration sensor 220, the fiber grating strain sensor 230 and the fiber grating temperature sensor 240 are all quasi-distributed fiber grating sensors, the fiber grating strain sensor 230 and the fiber grating temperature sensor 240 are all mounted on the shape memory alloy wire 210, and the shape memory alloy wire 210 and the fiber grating vibration sensor 220 are mounted on the submarine cable 700 along the axial direction of the submarine cable.
The present disclosure proposes a submarine cable risk event identification method applied to a submarine cable risk event identification system, the second embodiment of which comprises the steps of:
s100, constructing a fault simulation experiment model of the submarine cable, and performing an experiment on a risk event of the submarine cable to obtain characteristic information when the submarine cable breaks down;
wherein the characteristic information when the submarine cable fails comprises: temperature characteristic information, vibration characteristic information, strain characteristic information and displacement characteristic information;
s200, establishing a submarine cable risk event recognition model based on a hybrid neural network according to characteristic information when a submarine cable fails;
s300, judging whether the submarine cable at the current moment has a fault or not through the submarine cable risk event recognition model according to the characteristic information of the submarine cable at the current moment, recognizing the type of a risk event corresponding to the fault of the submarine cable at the current moment, generating a fault analysis report of the submarine cable, and sending out early warning according to the type of the risk event;
wherein the characteristic information of the submarine cable at the current moment comprises: current temperature information, current vibration information, current strain information, and current displacement information.
The submarine cable fault simulation experiment model mainly simulates the following submarine cable risk events:
an anchor pound event. The anchor smashing event is an event that in actual conditions, when a ship is anchored, the ship anchor smashes the submarine cable, and impact damage is caused to the submarine cable.
A hook event. The hook event is an event that a ship drags the submarine cable at a certain speed when the ship is anchored in an actual situation, and the stretchability of the submarine cable is damaged.
An ocean current event. The ocean current event is an event that a ship passes above the submarine cable when moving in an actual situation, and seawater generates certain vibration and is transmitted to the submarine cable.
Referring to fig. 5, in step S100 of this embodiment, the method includes constructing a simulation accident experiment model of the submarine cable, performing an experiment on a risk event of the submarine cable, and obtaining characteristic information of the submarine cable when the submarine cable fails, and the method further includes the following steps:
s110, constructing a fault simulation experiment model according to a certain scaling ratio according to the environment of the submarine cable in actual conditions;
s120, carrying an optical fiber sensing unit on the submarine cable along the line, placing the submarine cable under the model bracket, and adjusting the model bracket to ensure that the zero point at the bottom of the model bracket and the submarine cable are in the same plane;
s130, respectively carrying out simulation experiments of an anchor event, a hook event and an ocean current event of the submarine cable to obtain characteristic information when the submarine cable breaks down.
The present disclosure realizes online monitoring of the submarine cable 700 through the optical fiber sensing unit 200, and obtains characteristic information of the submarine cable 700 when a fault occurs and characteristic information of the submarine cable 700 at the current time. The optical fiber sensing unit comprises an optical fiber grating vibration sensor 220, an optical fiber grating strain sensor 230 and an optical fiber grating temperature sensor 240, and the three optical fiber grating sensors are quasi-distributed optical fiber grating sensors. Compared with other fiber grating sensors, the accuracy of the quasi-distributed fiber grating sensor is higher.
Referring to fig. 6, the step of obtaining the characteristic information when the submarine cable fails includes:
s131, sensing the submarine cables in different risk events through an optical fiber sensing unit to obtain first field information, and transmitting the first field information to a grating demodulator through an underwater acoustic communication machine in the form of optical fiber data;
s132, demodulating the optical fiber data corresponding to the first field information through a grating demodulator to obtain a first reflected wave wavelength and a variable quantity of the first reflected wave wavelength of the optical fiber sensing unit;
s133, calculating the temperature characteristic information, the strain characteristic information and the vibration characteristic information according to the wavelength of the first reflected wave and the variation of the wavelength;
specifically, when the submarine cable fails, the grating optical fiber temperature sensor 240 has a thermal expansion effect and a thermo-optic effect, the thermal expansion effect affects the period of the grating optical fiber temperature sensor 240, and the thermo-optic effect affects the effective refractive index of the grating optical fiber temperature sensor 240, and the temperature characteristic information is obtained by calculating through the thermal expansion effect and the thermo-optic effect:
wherein μ represents the temperature characteristic information, Δ λ B Indicating the variation of the wavelength of the first reflected wave, lambda, of the grating fiber temperature sensor B Initial quantity, a, representing the wavelength of the first reflected wave of the grating fibre-optic temperature sensor 1 B is the thermal expansion coefficient of the grating optical fiber temperature sensor, and b is the thermal light effect coefficient of the grating optical fiber temperature sensor.
When the submarine cable fails, the grating fiber strain sensor 230 is subjected to forces in all directions, and the strain characteristic information is obtained by calculating the effective elastic-optical coefficient of the grating fiber strain sensor 230:
wherein ε represents the strain characteristics information, Δ λ B Indicating the variation of the wavelength of the first reflected wave, lambda, of the optical fibre strain sensor B Representing an initial quantity, P, of a first reflected wave wavelength of a grating fibre strain sensor e Is the effective elastic-optical coefficient of the grating optical fiber strain sensor.
When the submarine cable is in fault, the vibration characteristic information can be calculated according to the first reflected wave wavelength variation, and the vibration characteristic information satisfies the following formula:
wherein S represents the vibration characteristic information, k eff Representing the equivalent stiffness, L, of a fiber grating vibration sensor f The packaging length of the fiber grating vibration sensor is shown, a is the acceleration of the fiber grating vibration sensor, m is the mass of the submarine cable, and delta lambda B Indicating a variation of a wavelength of a first reflected wave, λ, of the grating fiber vibration sensor B An initial quantity, P, representing a first reflected wave wavelength of the grating fiber vibration sensor e The elastic-optical coefficient of the grating optical fiber vibration sensor.
S134, reconstructing the shape of the submarine cable with the fault through a reconstruction algorithm, and comparing the reconstructed shape of the submarine cable with the shape of the submarine cable before reconstruction to obtain the displacement characteristic information.
In step S200 in this embodiment, a hybrid neural network is trained by using feature information when a submarine cable fails and three risk events, namely, an anchor event, a hook event, and an ocean current event, as training parameters, so as to generate a submarine cable risk event recognition model.
Referring to fig. 7, the generating of the submarine cable risk event identification model includes the following steps:
s210, normalizing the characteristic information when the submarine cable fails, and converting the format of the characteristic information when the submarine cable fails into a format which meets the input conditions of a convolutional neural network and a long-short term memory neural network;
s220, dividing feature information when the submarine cable fails into a training set and a testing set according to a proportion, inputting the training sets into a convolutional neural network and a long-short term memory neural network respectively, training the convolutional neural network and the long-short term memory neural network, outputting a first feature vector by the convolutional neural network, and outputting a second feature vector by the long-short term memory neural network;
s230, setting a splicing layer, and splicing the first feature vector and the second feature vector through the splicing layer to generate high-level risk features;
s240, setting a full connection layer and a Softmax classifier, inputting the high-level risk features into the full connection layer and the Softmax classifier, and generating a submarine cable risk event recognition model and a weight value thereof;
and S250, testing the performance of the submarine cable risk event recognition model through the test set.
Referring to fig. 8, the hybrid neural network used in the present invention is a convolutional neural network-long and short term memory neural network (CNN-LSTM) neural network, and the present invention inputs the feature information and the type of the risk event when the submarine cable fails into the convolutional neural network and the long and short term memory neural network, respectively trains the convolutional neural network and the long and short term memory neural network, splices the training results of the two neural networks through a splicing layer, and outputs the submarine cable risk event recognition model through a full connection layer and a Softmax classifier.
Before inputting the characteristic information when the submarine cable fails into the hybrid neural network, firstly, all the characteristic information needs to be normalized, and format conversion is carried out on the normalized characteristic information so as to enable the format of the characteristic information to accord with the data input conditions of the convolutional neural network and the long-short term memory neural network.
The normalization process maps the features of all feature information to the same size, the normalization process is beneficial to finding out a global optimal solution when a neural network is trained in the later stage so as to avoid dimension influence among indexes, and the feature information after normalization is within [0,1 ].
The normalization process satisfies the following formula:
wherein x is the characteristic information when the submarine cable fails, min (x) is the minimum value of the characteristic information when the submarine cable fails, max (x) is the maximum value of the characteristic information when the submarine cable fails, and x * The characteristic information of the submarine cable when the submarine cable fails is obtained after normalization processing.
The invention converts the normalized format of the characteristic information into an input format conforming to the convolutional neural network through zero padding operation. The zero padding operation is defined as centering the dimension of all the feature information to 0, and converting the data dimension of all the feature information to N x 1 (wherein N is the dimension of the feature information) so as to avoid the data deviation from reducing the effect of the deep learning network model.
The zero-padding operation satisfies the following formula:
wherein N is zero The number of zero padding required for each feature information is defined, m is defined as the total data amount contained in each feature information,
is defined as
Rounding to positive infinity.
After normalization processing and format conversion, dividing characteristic information of the submarine cable when a fault occurs into a training set and a test set according to a certain proportion, wherein the training set is used as an input parameter to train the hybrid neural network, and the test set is used for testing the performance of the submarine cable risk event recognition model. The invention takes the displacement characteristic information, the strain characteristic information, the temperature characteristic information and the vibration characteristic information as a channel respectively to form four-channel data, and inputs the training set into a convolutional neural network and a long-short term memory neural network in a multi-channel mode.
Training the convolutional neural network according to the set hyper-parameters of the convolutional neural network:
in the convolutional neural network, the ith convolutional layer performs the ith convolutional calculation on input data through a set convolutional core to generate an ith feature map, the ith pooling layer samples the ith feature map through a set pooling rule so as to reduce the size of a convolutional neural network model, improve the calculation speed of the convolutional neural network and improve the robustness of extracted features, and the pooling rule is a maximum pooling method. And outputting a first feature matrix after multilayer convolutional layer convolution calculation and multilayer pooling layer sampling, wherein the first feature matrix is used for connecting with a second feature vector output by the long-term and short-term memory neural network at a later stage.
Because the first feature matrix is a two-dimensional matrix which needs to be stretched into a one-dimensional vector through stretching operation, the stretching layer is added in the method, the first feature matrix is input into the stretching layer and is transformed into a first feature vector, and the first feature vector is a one-dimensional vector, so that the subsequent fusion with the output of the long-term and short-term memory neural network is facilitated.
Training the long-short term memory neural network according to the set hyper-parameters of the long-short term memory neural network while training the convolutional neural network:
and constructing a cell unit by using the LSTMBlockcell, wherein the state of the cell unit is a feature vector, and displacement feature information, temperature feature information, strain displacement information and vibration feature information of the cell unit at the t-th moment are updated or forgotten through a forgetting gate, an updating gate and an output gate. Let the state of the cell unit at the t-1 th time be a <t-1＞ The input at the t-th time is x ＜t＞ The forgetting door and the updating door meet the following formulas:
f t ＝σ(w f *[a ＜t-1＞ ,x ＜t＞ ]+b f )；
i t ＝σ(w i *[a <t-1> ,x <t＞ ]+b i )；
wherein f is t 、i t Defined as the output of the forgetting gate and the updating gate, w f 、w i Weight matrix of inputs of forgetting gate and updating gate, respectively, b f 、b i Offset term vectors of a forgetting gate and an updating gate are respectively, and sigma is a Sigmond activation function.
Update candidate value of cell unit at i-th time:
wherein,
represents a candidate value of a cell unit at the i-th time, tanh () is a tanh activation function, W c As a weight matrix input to the tanh activation function, b c Is a bias term vector;
calculate the state of the cell unit at time i:
the above formula is defined as the updated value a of the cell unit ＜t＞ By forgetting the state a of the cell unit at the previous moment ＜t-1＞ Or/and adding the candidate value of the cell unit at the i-th moment
And (4) obtaining.
The output gate is used for integrating the states of the cell units at the ith moment so as to obtain the output of the long-short term memory neural network sequence, and the output gate satisfies the following formula:
o t ＝σ(w o *[a ＜t-1＞ ,x vt> ]+b o )；
wherein o is t Defined as the output of the output gate, σ being the activation function, w o Weight matrix of inputs being output gates, b o Is a bias term vector;
final long-short term memory neural network output sequence output h t Said sequence outputs h t Is a second feature vector, where the sequence outputs h t The following formula is satisfied:
h t ＝o t *tanh(a ＜t＞ )。
the first feature vector output by the convolutional neural network and the second feature vector output by the long-short term memory neural network are jointly converged in a splicing layer, the first feature vector and the second feature vector are both features reflecting risk events, the splicing layer is used for splicing and integrating the first feature vector and the second feature vector, the splicing layer obtains high-level features reflecting the risk events after splicing and integrating the first feature vector and the second feature vector, the advanced features are converted into feature maps through calculation of an activation layer and a normalization layer, finally the feature maps are input into a full connection layer and a Softmax classifier to generate a submarine cable risk event recognition model and a weight value thereof, a loss function of the submarine cable risk event recognition model is calculated, and optimizing the weight of the submarine cable risk event identification model by using an Adam algorithm so as to enable the submarine cable risk event identification model to achieve the best classification result.
For identification of risk events, the method integrates the spatial feature extraction capability of a convolutional neural network and the time memory capability of a long-term and short-term memory neural network, wherein parameter dynamic and model dynamic are two key points, the parameter dynamic is defined as that data sensed by a plurality of fiber bragg grating sensors and data sensed by history of the data are utilized to extract change information, and a plurality of dynamic parameters, namely feature information when a submarine cable fails, are obtained; and the model dynamism is defined as that various variables of the submarine cable risk event identification model change along with the change of time.
The method comprises the steps of firstly combining a plurality of pieces of characteristic information which accord with input conditions of a neural network, wherein the characteristic information can represent induction parameters of submarine cable faults, performing multiple convolution calculation and pooling operations by using the convolution neural network, processing data by using a long-term and short-term memory neural network capable of processing time sequence data to obtain a first characteristic vector and a second characteristic vector which reflect risk events, fusing the first characteristic vector and the second characteristic vector to obtain high-grade characteristics which reflect the risk events, and obtaining a submarine cable risk event identification model and hyper-parameters thereof by using a full connection layer and a Softmax classifier.
Preferably, when the sequence node of the long-short term memory neural network is calculated each time, the hyper-parameters stored at the previous time node (namely the t-1 th time) are loaded, the invention adopts a Bayesian optimization algorithm to optimize the hyper-parameters of the long-short term memory neural network, and the hyper-parameters of the long-short term memory neural network comprise a learning rate, the number of neurons of a hidden layer, an activation function and the like, so as to obtain the optimal hyper-parameters of the long-short term memory neural network.
Further, after the submarine cable risk event model is obtained, the test set is input into the submarine cable risk event identification model, and the performance of the submarine cable risk event identification model is tested.
In step S300 of this embodiment, the plurality of grating fiber sensors sense field information of the submarine cable 700 at the current time, obtain second field information, and transmit the second field information to the grating demodulator 300 through the underwater acoustic communication device in the form of fiber data, the grating demodulator 300 demodulates the fiber data corresponding to the second field information, and obtain characteristic information of the submarine cable 700 at the current time through the processing unit 400 and the reconstruction unit 500, where the characteristic information of the submarine cable 700 at the current time includes current temperature information, current vibration information, current strain information, and current displacement information.
And inputting the characteristic information of the submarine cable at the current moment into the submarine cable risk event recognition model. And judging whether the submarine cable 700 at the current moment has a fault or not through the submarine cable risk event identification model, identifying the type of the risk event corresponding to the fault of the submarine cable at the current moment, and generating a fault analysis report of the submarine cable so as to achieve the effect of fault early warning.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that the present invention is not limited to the details of the embodiments shown and described, but is capable of numerous equivalents and substitutions without departing from the spirit of the invention as set forth in the claims appended hereto.
Claims (10)
1. A submarine cable risk event identification system, comprising: the system comprises a fault simulation experiment model, an optical fiber sensing unit, an underwater acoustic communication machine, a grating demodulator, a processing unit, a reconstruction unit and a risk event identification unit;
the fault simulation experiment model is used for simulating corresponding risk events when the submarine cable fails;
the optical fiber sensing unit is fixed along the axial direction of the submarine cable and is used for sensing first field information of the submarine cable when the submarine cable breaks down and second field information of the submarine cable at the current moment, and the field information is temperature field information, strain field information, vibration field information and displacement field information;
the underwater acoustic communicator is placed in a fault simulation experiment model and connected with the optical fiber demodulator, and the optical fiber sensing unit transmits the first field information or the second field information to the optical fiber demodulator in the form of optical fiber data through the underwater acoustic communicator;
the grating demodulator is used for demodulating optical fiber data corresponding to the first field information to obtain a first reflected wave wavelength and a variable quantity thereof of the optical fiber sensing unit;
the processing unit is used for acquiring temperature characteristic information, vibration characteristic information and strain characteristic information when the submarine cable fails according to the wavelength of the first reflected wave and the variable quantity of the first reflected wave;
the reconstruction unit is used for reconstructing the shape of the submarine cable with the fault through a reconstruction algorithm to obtain displacement characteristic information when the submarine cable has the fault;
the risk event identification unit is used for establishing a submarine cable risk event identification model based on a hybrid neural network according to the temperature characteristic information, the vibration characteristic information, the strain characteristic information and the displacement characteristic information;
the grating demodulator is further used for demodulating optical fiber data corresponding to the second field information to obtain a second reflected wave wavelength and a variable quantity thereof of the optical fiber sensing unit;
the processing unit is used for obtaining current temperature information, current vibration information and current strain information of the submarine cable at the current moment according to the wavelength of the second reflected wave and the variation of the wavelength;
the reconstruction unit is used for reconstructing the shape of the submarine cable at the current moment through a reconstruction algorithm to obtain the current displacement information of the submarine cable at the current moment;
and the risk event identification unit is also used for judging whether the submarine cable at the current moment has a fault or not according to the current temperature information, the current vibration information, the current strain information and the current displacement information through the submarine cable risk event identification model, identifying the type of the risk event corresponding to the fault of the submarine cable, and generating a fault analysis report of the submarine cable.
2. The submarine cable risk event recognition system according to claim 1, wherein the grating sensing unit comprises a shape memory alloy wire, a fiber grating vibration sensor, a fiber grating strain sensor and a fiber grating temperature sensor, the fiber grating vibration sensor, the fiber grating strain sensor and the fiber grating temperature sensor are all quasi-distributed fiber grating sensors, the fiber grating strain sensor and the fiber grating temperature sensor are all mounted on the shape memory alloy wire, and the shape memory alloy wire and the fiber grating vibration sensor are mounted on the submarine cable along the axial direction of the submarine cable.
3. The submarine cable risk event recognition system according to claim 1, wherein the fault simulation experiment model comprises a sandbox, a model support, a nylon rope, a key module, a control module and a driving module, and the control module is connected with the key module and the driving module respectively; the model support is a scale rod with a graduated scale, and a zero point at the bottom of the scale rod and the submarine cable are positioned on the same horizontal plane; the model support is placed in the sandbox, one end of the model support is fixed with the sandbox, the other end of the model support is connected with the fixed pulley, the nylon rope is connected with the fixed pulley in a sliding mode, the ship anchor is tied at one end of the nylon rope, and the other end of the nylon rope is connected with the driving module;
the sandbox is used for simulating the environment of the submarine cable in actual use, kaolin and seawater are filled in the sandbox, and the underwater acoustic communication machine is placed in the seawater;
the driving module includes: the driving support is positioned at the bottom of the model support, the conveying belts are installed on two sides of the driving support, the guide rail sliding blocks are placed on the conveying belts, the guide rail sliding blocks are connected with one ends, far away from the ship anchor, of the nylon ropes, and the stepping motors are used for driving the conveying belts to transmit in the horizontal direction and further driving the guide rail sliding blocks to move in the horizontal direction;
the control module is used for detecting the condition that the key module is pressed down and outputting a first control signal or a second control signal to the driving module according to the condition that the key module is pressed down; the driving module is used for controlling the guide rail sliding block to move leftwards along the horizontal direction according to the first control signal and controlling the guide rail sliding block to move rightwards along the horizontal direction according to the second control signal;
the key module comprises a start key, a pause key, a left shift key and a right shift key, wherein the start key is used for starting the stepping motor, the pause key is used for closing the stepping motor, the left shift key is used for controlling the stepping motor to rotate forwards, and the right shift key is used for controlling the stepping motor to rotate backwards.
4. A submarine cable risk event identification method applied to a submarine cable risk event identification system according to any one of claims 1 to 3, the method comprising the steps of:
s100, constructing a fault simulation experiment model of the submarine cable, and performing an experiment on a risk event of the submarine cable to obtain characteristic information when the submarine cable breaks down;
wherein the characteristic information when the submarine cable fails comprises: temperature characteristic information, vibration characteristic information, strain characteristic information and displacement characteristic information;
s200, establishing a submarine cable risk event recognition model based on a hybrid neural network according to characteristic information when a submarine cable fails;
s300, judging whether the submarine cable at the current moment has a fault or not through the submarine cable risk event recognition model according to the characteristic information of the submarine cable at the current moment, recognizing the type of a risk event corresponding to the fault of the submarine cable at the current moment, generating a fault analysis report of the submarine cable, and sending out early warning according to the type of the risk event;
wherein the characteristic information of the submarine cable at the current moment comprises: current temperature information, current vibration information, current strain information, and current displacement information.
5. The submarine cable risk event identification method according to claim 4, wherein S100 is specifically:
s110, constructing a fault simulation experiment model according to a certain scaling ratio according to the environment of the submarine cable in actual conditions;
s120, carrying an optical fiber sensing unit along the submarine cable, placing the submarine cable under the model bracket, and adjusting the model bracket to ensure that the zero point at the bottom of the model bracket and the submarine cable are in the same plane;
s130, respectively carrying out simulation experiments of an anchor event, a hook event and an ocean current event of the submarine cable to obtain characteristic information when the submarine cable breaks down.
6. The submarine cable risk event identification method according to claim 5, wherein said obtaining characteristic information of a submarine cable failure further comprises the steps of:
s131, sensing field information of the submarine cables in different risk events through an optical fiber sensing unit, and transmitting the field information of the submarine cables to a grating demodulator through an underwater acoustic communication machine in the form of optical fiber data;
s132, demodulating the optical fiber data through a grating demodulator to obtain the reflected wave wavelength variation of the optical fiber sensing unit;
s133, calculating the temperature characteristic information, the strain characteristic information and the vibration characteristic information according to the wavelength variation of the reflected wave;
wherein the temperature characteristic information satisfies the following formula:
wherein μ represents the temperature characteristic information, Δ λ B Indicating the variation of the wavelength of the first reflected wave, λ, of the grating fiber temperature sensor B An initial quantity, a, representing a first reflected wave wavelength of the grating fiber temperature sensor 1 B is the thermal expansion coefficient of the grating optical fiber temperature sensor, and b is the thermo-optic effect coefficient of the grating optical fiber temperature sensor;
the strain characteristic information satisfies the following formula:
wherein ε represents the strain characteristic information, Δ λ B Indicating the variation of the wavelength of the first reflected wave, λ, of the grating fiber strain sensor B An initial quantity, P, representing a first reflected wave wavelength of the grating fiber strain sensor e The effective elastic-optical coefficient of the grating optical fiber strain sensor;
the vibration characteristic information satisfies the following formula:
wherein S represents the vibration characteristic information, k eff Is equivalent stiffness, L, of the fiber grating vibration sensor f Is the package length of the fiber grating vibration sensor, a is the acceleration, m is the mass of the submarine cable, and Delta lambda B Presentation instrumentThe first reflection wave wavelength variation of the grating fiber vibration sensor is λ B An initial quantity, P, representing a first reflected wave wavelength of the grating fiber vibration sensor e The elastic-optical coefficient of the grating optical fiber vibration sensor is obtained;
s134, reconstructing the shape of the submarine cable through a reconstruction algorithm, and comparing the reconstructed shape of the submarine cable with the shape of the submarine cable before reconstruction to obtain the displacement characteristic information.
7. The submarine cable risk event identification method according to claim 4, wherein step S200 comprises the steps of:
s210, normalizing the characteristic information when the submarine cable fails, and converting the format of the characteristic information when the submarine cable fails into a format which meets the input conditions of a convolutional neural network and a long-short term memory neural network;
s220, dividing feature information of the submarine cable in fault into a training set and a testing set according to a proportion, inputting the training sets into the convolutional neural network and the long-short term memory neural network respectively, training the convolutional neural network and the long-short term memory neural network, outputting a first feature vector by the convolutional neural network, and outputting a second feature vector by the long-short term memory neural network;
s230, setting a splicing layer, and splicing the first feature vector and the second feature vector through the splicing layer to generate high-level risk features;
s240, setting a full connection layer and a Softmax classifier, inputting the high-level risk features into the full connection layer and the Softmax classifier, and generating a submarine cable risk event recognition model and a weight value thereof;
and S250, testing the performance of the submarine cable risk event recognition model through the test set.
8. The submarine cable risk event identification method according to claim 7, wherein in step S220, after the convolutional neural network training is finished, a first feature matrix is output, wherein the first feature matrix is a two-dimensional matrix, a stretching layer is provided, and the first feature matrix is stretched into a first feature vector through the stretching layer, wherein the first feature vector is a one-dimensional feature vector.
9. The submarine cable risk event identification method according to claim 8, wherein said step S200 further comprises: and optimizing the hyper-parameters of the long-term and short-term memory neural network by a Bayesian optimization algorithm.
10. The submarine cable risk event identification method according to claim 9, wherein said step S200 further comprises: calculating a loss function of the submarine cable risk event identification model, and optimizing the weight of the submarine cable risk event identification model through an Adam algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210574302.4A CN115014571B (en) | 2022-05-25 | 2022-05-25 | Submarine cable risk event identification system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210574302.4A CN115014571B (en) | 2022-05-25 | 2022-05-25 | Submarine cable risk event identification system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115014571A true CN115014571A (en) | 2022-09-06 |
| CN115014571B CN115014571B (en) | 2023-03-24 |
Family
ID=83068398
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210574302.4A Active CN115014571B (en) | 2022-05-25 | 2022-05-25 | Submarine cable risk event identification system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115014571B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115388959A (en) * | 2022-10-31 | 2022-11-25 | 高勘(广州)技术有限公司 | Submarine cable operation and maintenance method, device, equipment and storage medium |
| CN115841742A (en) * | 2022-11-16 | 2023-03-24 | 国网福建省电力有限公司厦门供电公司 | Diversified application scene pipeline monitoring risk early warning key method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102981104A (en) * | 2012-11-19 | 2013-03-20 | 中国能源建设集团广东省电力设计研究院 | On-line monitoring method for submarine cables |
| CN206593996U (en) * | 2017-04-06 | 2017-10-27 | 国网浙江省电力公司舟山供电公司 | Optical fiber composite submarine cable wear test system |
| CN109870627A (en) * | 2017-12-05 | 2019-06-11 | 华北电力大学（保定） | Submarine cable fault alarm and diagnostic method based on distributed fiber optic temperature strain and vibration monitoring data |
| CN110186489A (en) * | 2019-06-28 | 2019-08-30 | 中国南方电网有限责任公司超高压输电公司广州局 | A kind of submarine cable temperature stress imitative experimental appliance and method on the bank |
| AU2020102819A4 (en) * | 2020-10-16 | 2020-12-10 | Beihang University | Digital-twin-driven method and system for fault diagnosis of subsea production system of offshore oil |
| CN213473473U (en) * | 2020-10-30 | 2021-06-18 | 浙江舟山海洋输电研究院有限公司 | Device for simulating penetration depth of ship anchor |
| CN113030789A (en) * | 2021-04-12 | 2021-06-25 | 辽宁工程技术大学 | Series arc fault diagnosis and line selection method based on convolutional neural network |
| CN113298110A (en) * | 2021-03-24 | 2021-08-24 | 国网河北省电力有限公司沧州供电分公司 | Submarine cable fault diagnosis method, device and equipment |
| CN113933667A (en) * | 2021-11-08 | 2022-01-14 | 国网浙江省电力有限公司电力科学研究院 | Submarine cable aging test device and method under simulated operation environment |
-
2022
- 2022-05-25 CN CN202210574302.4A patent/CN115014571B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102981104A (en) * | 2012-11-19 | 2013-03-20 | 中国能源建设集团广东省电力设计研究院 | On-line monitoring method for submarine cables |
| CN206593996U (en) * | 2017-04-06 | 2017-10-27 | 国网浙江省电力公司舟山供电公司 | Optical fiber composite submarine cable wear test system |
| CN109870627A (en) * | 2017-12-05 | 2019-06-11 | 华北电力大学（保定） | Submarine cable fault alarm and diagnostic method based on distributed fiber optic temperature strain and vibration monitoring data |
| CN110186489A (en) * | 2019-06-28 | 2019-08-30 | 中国南方电网有限责任公司超高压输电公司广州局 | A kind of submarine cable temperature stress imitative experimental appliance and method on the bank |
| AU2020102819A4 (en) * | 2020-10-16 | 2020-12-10 | Beihang University | Digital-twin-driven method and system for fault diagnosis of subsea production system of offshore oil |
| CN213473473U (en) * | 2020-10-30 | 2021-06-18 | 浙江舟山海洋输电研究院有限公司 | Device for simulating penetration depth of ship anchor |
| CN113298110A (en) * | 2021-03-24 | 2021-08-24 | 国网河北省电力有限公司沧州供电分公司 | Submarine cable fault diagnosis method, device and equipment |
| CN113030789A (en) * | 2021-04-12 | 2021-06-25 | 辽宁工程技术大学 | Series arc fault diagnosis and line selection method based on convolutional neural network |
| CN113933667A (en) * | 2021-11-08 | 2022-01-14 | 国网浙江省电力有限公司电力科学研究院 | Submarine cable aging test device and method under simulated operation environment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115388959A (en) * | 2022-10-31 | 2022-11-25 | 高勘(广州)技术有限公司 | Submarine cable operation and maintenance method, device, equipment and storage medium |
| CN115841742A (en) * | 2022-11-16 | 2023-03-24 | 国网福建省电力有限公司厦门供电公司 | Diversified application scene pipeline monitoring risk early warning key method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115014571B (en) | 2023-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115014571B (en) | Submarine cable risk event identification system and method | |
| US11774485B2 (en) | Fault diagnosis method and apparatus for submarine cable, and device | |
| CN114352947B (en) | Gas pipeline leakage detection method, system, device and storage medium | |
| CN113837477A (en) | Data dual-drive power grid fault prediction method, device and equipment under typhoon disaster | |
| CN109284863A (en) | A kind of power equipment temperature predicting method based on deep neural network | |
| Yang et al. | Pipeline safety early warning by multifeature-fusion CNN and LightGBM analysis of signals from distributed optical fiber sensors | |
| CN115688035A (en) | Time sequence power data anomaly detection method based on self-supervision learning | |
| CN117824756B (en) | Building structure monitoring method, equipment and storage medium | |
| Zhou et al. | Structural health monitoring of offshore wind power structures based on genetic algorithm optimization and uncertain analytic hierarchy process | |
| US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
| CN112528548A (en) | Self-adaptive depth coupling convolution self-coding multi-mode data fusion method | |
| Xu et al. | A hybrid model integrating CNN-BiLSTM and CBAM for anchor damage events recognition of submarine cables | |
| CN114739891A (en) | Floating type offshore wind power mooring system corrosion state detection and evaluation system | |
| CN117521512A (en) | Bearing residual service life prediction method based on multi-scale Bayesian convolution transducer model | |
| CN118153774A (en) | Method for predicting tension long-term and short-term memorizing network of mooring cable based on gravitation search optimization | |
| Tang et al. | Sensing technologies and artificial intelligence for subsea power cable asset management | |
| Yang et al. | Condition prediction of submarine cable based on CNN-BiGRU integrating attention mechanism | |
| Yang et al. | Development of ship structure health monitoring system based on IOT technology | |
| CN117554000A (en) | Intelligent detection method and system for leakage of tailing conveying pipeline | |
| CN117056845A (en) | Automatic driving automobile abnormality detection method and related system | |
| Xu et al. | Long Short Term Memory Based Status Prediction for Diesel Generator Set | |
| CN116881714A (en) | Power grid false data injection detection method and device based on double-branch CNN-LSTM | |
| CN113761795A (en) | Aircraft engine fault detection method and system | |
| Zhu et al. | Prediction of distributed optical fiber monitoring data based on GRU-BP | |
| Xu et al. | Anchor-Damage Event Recognition Based on FBG sensors and CNN-BiLSTM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |