CN115871901B - Sturgeon-imitating robot and submarine cable fault detection method - Google Patents

Sturgeon-imitating robot and submarine cable fault detection method Download PDF

Info

Publication number
CN115871901B
CN115871901B CN202310109397.7A CN202310109397A CN115871901B CN 115871901 B CN115871901 B CN 115871901B CN 202310109397 A CN202310109397 A CN 202310109397A CN 115871901 B CN115871901 B CN 115871901B
Authority
CN
China
Prior art keywords
shell
submarine cable
trunk
sturgeon
neural network
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.)
Active
Application number
CN202310109397.7A
Other languages
Chinese (zh)
Other versions
CN115871901A (en
Inventor
田群宏
杨彬
马洪波
高天宇
刘浩冉
李洪宇
刘冰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong University of Science and Technology
Original Assignee
Shandong University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong University of Science and Technology filed Critical Shandong University of Science and Technology
Priority to CN202310109397.7A priority Critical patent/CN115871901B/en
Publication of CN115871901A publication Critical patent/CN115871901A/en
Application granted granted Critical
Publication of CN115871901B publication Critical patent/CN115871901B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
    • Y02A40/80—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in fisheries management
    • Y02A40/81—Aquaculture, e.g. of fish
    • 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
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Abstract

The invention discloses a sturgeon-like robot and a submarine cable fault detection method, and relates to the technical field of marine cable detection. The novel fishing head comprises a head part, a trunk unit, a fish tail part and a control system, wherein the head part comprises a head shell, two pectoral fins are symmetrically arranged on two sides of the head shell, and a dorsal fin is arranged on the top of the head shell. The tail portion is coupled to the head housing by a plurality of torso units, including a tail housing and a tail fin. The trunk unit comprises a trunk shell and a steering driving mechanism, and the trunk units are connected end to end in sequence. The trunk units are matched with the fish heads to realize the integral swing of the sturgeon-like robot and provide the advancing power for the sturgeon-like robot. The control system comprises a controller, a sonar, an image acquisition module and a signal emission module, wherein the image acquisition module is arranged at the front end of the head shell. The invention adopts sonar to find the position of the submarine cable, and identifies the fault type of the submarine cable through image acquisition and convolutional neural network, so that the invention has the advantages of high searching speed and accurate position judgment, and is suitable for the submarine complex environment.

Description

Sturgeon-imitating robot and submarine cable fault detection method
Technical Field
The invention relates to the technical field of marine cable fault detection, in particular to a sturgeon-imitating robot and a submarine cable fault detection method.
Background
Submarine cable means cable laid on the sea floor by wrapping insulating material, and is mainly used for telecommunication transmission. Submarine cables are classified into underwater communication cables and power cables. The application fields of the method comprise international communication, electric power and communication between coastal islands and cities, offshore wind power generation, offshore oil platform electric power and information transmission and the like. Wherein, the underwater communication cable is the main mode of international information transmission, which accounts for 97% of world data transmission; submarine cables are important transmission means for electric power and communication between coastal islands and cities, the sea areas of China are wide, telecommunication transmission between islands, coastal cities and between land and islands is realized, and the needed submarine cables are huge in quantity; in the field of energy development, the submarine cable has wide application prospects in ocean platforms, offshore wind power generation and power transmission. Therefore, the submarine cable has important value in social life, economy, military and other aspects.
However, even with shielding and burial, there are more than 200 submarine cable failures per year. Damage to the submarine cable not only causes significant economic loss, but may also lead to a series of problems in the resource exploitation process, further causing marine environmental pollution. Damage to submarine cables mainly includes natural disasters and man-made damage. Natural disasters include ocean bottom earthquakes, landslides, ocean current waves, tsunamis, billows, sea level rises, extreme weather (hurricanes), volcanic activities, and the like. These natural disasters may ultimately lead to submarine cable wear, stress fatigue and failure, breakage, etc. The artificial damage includes accidental artificial damage and intentional artificial damage. Accidental artificial threats include artificial actions that cause the cable to be unexpected, such as trawl operation during trawling or anchoring in fishing, and submarine operations such as clams dredgers, scallop dredgers, etc., are highly likely to damage the submarine cable. Intentional damage to cables is mainly manifested by theft of submarine cables. These artificial damages mainly cause breakage, fracture, and the like of the submarine cable.
In summary, there are numerous threats to submarine cables that may cause varying degrees of damage. In order to discover submarine cable faults and perform maintenance in time, many related researches have been conducted, and from the detection technology perspective, the existing submarine cable detection can be divided into methods based on optics, electricity, acoustics, magnetism and multiple sensors. The visual detection based on optical detection has the characteristics of insensitivity to noise data and capability of obtaining good environmental information, but is easily influenced by light and water quality; based on electrical detection, short-distance detection is not needed, but electrical signals are loaded on the submarine cable, and the submarine cable is generally used for detecting electrical faults, such as copper core short circuit, open circuit and the like; the acoustic-based detection technology has the characteristic of long detection distance, but is easy to be interfered by external noise, and has limited detection precision; based on the magnetic detection technology, the method has the advantage of long detection distance, but current needs to be injected into the cable for active detection, and passive detection is easy to be interfered by surrounding magnetic fields; the multi-sensor fusion technology based on combination of optical, electrical, acoustic, magnetic and other sensors has the characteristic of high reliability of detection results, but has the advantages of complex data processing, large calculation amount and long time consumption. From the above analysis, various submarine cable fault detection methods in the prior art are limited by corresponding defects, and the submarine cable detection problem in actual engineering can not be effectively solved. Accordingly, there is a need for further improvements and enhancements in the art.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide the sturgeon-like robot, and solves the problems that the detection means of the submarine cable in the prior art is affected by the submarine environment, so that the corresponding defects exist, the accuracy and the reliability of the detection result are low, and the fault point and the fault type of the submarine cable cannot be effectively and practically determined.
In order to solve the technical problems, the invention adopts the following technical scheme:
the sturgeon-like robot comprises a fish head part, a trunk unit, a fish tail part and a control system, wherein the fish head part comprises a sturgeon-like head shell, two pectoral fins are symmetrically arranged on the left side and the right side of the head shell, and dorsal fins are arranged on the top of the head shell.
The fish tail part is located the rear side of fish head part to through a plurality of that arrange in proper order the trunk unit links to each other with the rear side of head shell, and the fish tail part includes sturgeon form afterbody shell, and the rear end of afterbody shell is provided with the tail fin.
The trunk unit comprises a trunk shell and a steering driving mechanism arranged inside the trunk shell, the trunk units are connected end to end in sequence, the front end of the first trunk unit is fixedly connected with the rear end of the head shell, and the rear end of the last trunk unit is fixedly connected with the tail shell.
And each trunk unit is matched with the fish head to realize the integral swing of the sturgeon-like robot and provide the advancing power for the sturgeon-like robot.
The control system comprises a controller, a sonar, an image acquisition module and a signal transmission module, wherein the image acquisition module is arranged at the front end of the head shell, and the sonar, the image acquisition module and the signal transmission module are connected with the controller through signals.
Further, the trunk shell is a section of annular shell, and gaps are reserved between any two adjacent trunk shells and between the first trunk shell and the head shell.
The steering driving mechanism comprises a double-shaft servo motor and steering brackets, wherein the double-shaft servo motor is fixed in the trunk shell, the steering brackets are positioned on the front sides of the double-shaft servo motors and fixedly connected with output shafts of the double-shaft servo motor, and connecting frames are fixed on the rear sides of the double-shaft servo motors.
Further, the front end of the steering bracket in the trunk unit of the first time is fixedly connected with the rear end of the head shell, and the front ends of the steering brackets in the trunk units of the rest times are fixedly connected with the rear ends of the connecting frames in the trunk units adjacent to the front side.
The rear end of the connecting frame in the last trunk unit is fixedly connected with the front end of the tail shell.
Further, a motor bracket is fixedly arranged on the inner side of each trunk shell, the double-shaft servo motor is fixed on the motor bracket, and two output shafts of the double-shaft servo motor are vertically arranged.
Two float blocks made of buoyancy materials are symmetrically arranged on the left side and the right side of the double-shaft servo motor, and the two float blocks are fixedly connected with the inner side wall of the trunk shell.
Further, each trunk shell comprises two shell monomers which are oppositely arranged, the upper end of each shell monomer is fixedly provided with a hinge block, and the two hinge blocks of the same shell monomer are arranged in a staggered manner and are hinged through a rotating shaft.
The adjacent sides of the bottom of the two shell monomers are respectively fixed with a bolt block, and the two bolt blocks at the bottom of the same shell monomer are opposite to each other and are fixedly connected through bolts.
Further, the motor support comprises two support monomers, each support monomer comprises a T-shaped structure formed by connecting a transverse rod and a longitudinal rod stationary phase, and the two support monomers are arranged in an axisymmetric mode.
One end that two horizontal poles kept away from each other all links to each other with the inside wall fixed of truck shell, and the other end is fixed pegged graft with corresponding support monomer longitudinal rod, encloses into the closed region that can place biax servo motor.
Further, the top and the bottom of each connecting frame are respectively provided with a connecting terminal, and each connecting terminal is connected with the rear end of the head shell through a rib wire made of elastic materials.
Each pectoral fin is close to one side of head shell all fixedly connected with cross axle, cross axle and the lateral wall rotation sealing cooperation of head shell, and each cross axle is kept away from the one end of pectoral fin and all is connected with a servo motor.
The fin is vertically arranged, the bottom of the fin is fixedly connected with a vertical shaft, the vertical shaft is in rotary sealing fit with the top wall of the head shell, and the lower end of the vertical shaft is connected with the output end of the second servo motor fixed in the head shell.
The invention further aims at providing a submarine cable fault detection method adopting the sturgeon-like robot.
The submarine cable fault detection method adopts the sturgeon-like robot, and comprises the following steps:
step one, a convolutional neural network prediction model is established, the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full-connection layer and an output layer, and a submarine cable fault detection convolutional neural network mathematical model with optimal performance is obtained through learning and training.
Placing the sturgeon-like robot in water and submerging the sturgeon-like robot to a deep water area, enabling the sonar to send out a signal in the swimming process to detect the position of the submarine cable, and enabling the sonar to send out the signal to a controller after determining the position of the submarine cable, wherein the sturgeon-like robot reaches the position of the submarine cable.
Step three, the sturgeon-like robot moves along the extending direction of the submarine cable, the image acquisition module acquires the image information of the submarine cable in real time and sends the image information to the controller, the state and the fault type of the submarine cable are judged through the convolutional neural network, and the data information is stored.
And fourthly, after the sturgeon-like robot upwards floats out of the water, the data information is sent to a receiving terminal on the ship or the ground through a signal transmitting module, and the position sent by the signal is positioned through the base station so as to determine the position coordinates of the fault point.
And fifthly, repeating the process of the step four, sequentially determining each fault point and fault type of the submarine cable, and making a fault processing scheme by staff according to the data information.
Further, the establishment of the submarine cable detection prediction model comprises the following steps:
s1: the method comprises the steps of collecting images, horizontally overturning, randomly buckling, performing scale transformation, rotating, expanding submarine cable data images, and establishing a data set containing normal submarine cables and suspended and electric faults caused by external force.
The data set of the convolutional neural network is divided into a training subset, a verification subset and a test subset randomly and averagely, the training set is used for learning model parameters of the convolutional neural network, the verification subset selects super parameters by evaluating network performance, and the test subset measures network performance by generalization errors.
S2: and training the super parameters of the convolutional neural network, namely the number of convolutional layers, the size of the convolutional kernels and the super parameters of the number of the convolutional kernels by using a particle swarm optimization algorithm, and optimizing.
S3: training the convolutional neural network by using the optimized number of convolutional layers, the convolutional kernel size and the convolutional number, and importing the data into the input layer, the pooling layer and the full-connection layer according to the data, wherein the convolutional neural network can extract the fault characteristics of the training.
S4: and calculating an output result loss function, updating weights through a gradient descent method, and realizing multiple iterations of the convolutional neural network to enable the training model to be converged.
S5: after multiple iterations, the test set is adopted to judge the convolutional neural network classification performance based on particle swarm optimization.
S6: and according to the performance evaluation result, adjusting parameters of node number, iteration times and loss function value in the convolutional neural network.
S7: repeating the steps to finally obtain the submarine cable fault detection convolutional neural network mathematical model with the optimal performance.
The image acquisition module comprises a camera provided with a light source, wherein in the first step, the input layer is a submarine cable image, and the input layer expands submarine cable data images and further enhances the images.
The convolution layer is convolved with the convolution kernel by the feature vector and achieves the feature extraction of the submarine image by activating the function response.
The mathematical expression of the convolution layer neuron is:
Figure SMS_1
(1)/>
wherein,,
Figure SMS_6
convolutional layer for convolutional neural network>
Figure SMS_4
Is>
Figure SMS_10
The outputs of the individual channels;
Figure SMS_5
Convolutional layer for convolutional neural network>
Figure SMS_9
Is>
Figure SMS_14
The outputs of the individual channels;
Figure SMS_17
For calculating->
Figure SMS_13
A subset of the input feature maps for net activation of the individual channels;
Figure SMS_16
Is a convolution operation symbol;
Figure SMS_2
Is a convolution layer->
Figure SMS_8
Input vector->
Figure SMS_7
And neuron->
Figure SMS_12
A weight matrix of connections;
Figure SMS_15
Is a convolution layer->
Figure SMS_18
First->
Figure SMS_3
Of a characteristic mapDeviation value (I)>
Figure SMS_11
To activate the function.
The pool layer chemical expression is:
Figure SMS_19
(2)
wherein,,
Figure SMS_20
pooling layer for convolutional neural network>
Figure SMS_21
First->
Figure SMS_22
The weight coefficients of the individual channels;
Figure SMS_23
Is a pooling function;
Figure SMS_24
For pooling layer->
Figure SMS_25
First->
Figure SMS_26
Offset terms for the individual channels.
The full connection layer mathematical expression is as follows:
Figure SMS_27
(3)
wherein,,
Figure SMS_29
is a full connection layer->
Figure SMS_32
An output of (2);
Figure SMS_34
Is a full connection layer->
Figure SMS_30
Is a network weight coefficient of (a);
Figure SMS_31
Is a full connection layer
Figure SMS_33
An output of (2);
Figure SMS_35
Is a full connection layer->
Figure SMS_28
Is included.
The output layer gives specific classification results and judges the normal state of the submarine cable and specific types of suspended and electric faults caused by external force.
By adopting the technical scheme, the invention has the beneficial technical effects that: the method has the advantages that the sonar detection distance of the sturgeon-like robot is long, the target distance is accurately measured, the position of the submarine cable is quickly found and reaches the vicinity of the submarine cable, the camera acquires the image information of the submarine cable in real time, the condition of the submarine cable is detected through the convolutional neural network, the fault point is judged, the sturgeon-like robot floats out of the water surface, the data information is sent to the receiving terminal of the ship or the ground, and the base station at the bottom surface determines the fault position coordinates through the signal source. The invention has the advantages of high fault point searching speed and accurate position judgment, has strong generalization capability of the convolutional neural network prediction model, can adapt to the complex submarine environment, has good identification and classification capability, effectively identifies the fault type of the submarine cable, and provides accurate basis for the scheme formulation of fault treatment.
Drawings
Fig. 1 is a schematic structural view of a sturgeon-like robot according to the present invention.
FIG. 2 is a schematic view of the structure of FIG. 1 with one side of the housing monomer removed.
Figure 3 is a schematic view of the present invention of figure 1 with the entire torso casing removed.
Fig. 4 is a schematic view of the structure of fig. 3 after further removing all the float blocks according to the present invention.
Fig. 5 is a schematic view of the pectoral fin and related components of the present invention of fig. 1.
Fig. 6 is a partially assembled schematic illustration of the present invention showing the fishtail section, steering drive mechanisms and tendons.
Fig. 7 is a schematic view of a portion of the construction of the invention of fig. 1, showing a torso unit.
Fig. 8 is a schematic view of a portion of the structure of fig. 7, showing the torso housing and motor mount.
FIG. 9 is a flow chart of the present invention for detecting a submarine cable fault based on convolutional neural network.
Detailed Description
Embodiments of the present invention are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the invention but are not intended to limit the scope of the invention.
In the description of the present invention, unless otherwise indicated, the meaning of "a plurality" is two or more; the terms "upper," "lower," "left," "right," "inner," "outer," "front," "rear," "head," "tail," and the like refer to an orientation or positional relationship based on that shown in the drawings, merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the mechanisms or elements referred to must have a particular orientation, be configured and operated in a particular orientation, and thus are not to be construed as limiting the present invention. Furthermore, the terms "first," "second," "third," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
Embodiment 1, with reference to fig. 1 to 8, a sturgeon-like robot comprises a fish head part, a trunk unit, a fish tail part and a control system, wherein the fish head part comprises a sturgeon-like head shell 1, two pectoral fins 11 are symmetrically arranged on the left side and the right side of the head shell 1, and dorsal fins 12 are arranged on the top of the head shell.
Specifically, a lateral shaft 14 is fixedly connected to one side of each pectoral fin 11 close to the head casing 1, the lateral shaft 14 is in rotary sealing fit with the side wall of the head casing 1, and one end of each lateral shaft far away from the pectoral fin 11 is connected with a first servo motor 13.
The dorsal fin 12 is vertically arranged, the bottom of the dorsal fin is fixedly connected with a vertical shaft, the vertical shaft is in rotary sealing fit with the top wall of the head shell 1, and the lower end of the vertical shaft is connected with the output end of the second servo motor fixed in the head shell 1.
The fish tail part is located at the rear side of the fish head part and is connected with the rear side of the head shell 1 through a plurality of trunk units which are sequentially arranged, the fish tail part comprises a sturgeon-shaped tail shell 2, and the tail end of the tail shell 2 is provided with tail fins 21.
The trunk unit comprises a trunk shell 3 and a steering driving mechanism 4 arranged inside the trunk shell 3, the trunk shell 3 is a section of annular shell, each trunk shell 3 comprises two shell monomers 31 which are oppositely arranged, the upper end of each shell monomer 31 is fixedly provided with a hinge block 32, and the two hinge blocks 32 of the same shell monomer 31 are arranged in a staggered manner and are hinged through a rotating shaft.
The adjacent sides of the bottoms of the two shell monomers 31 are respectively fixed with a bolt block 33, and the two bolt blocks 33 at the bottom of the same shell monomer 31 are opposite to each other and fixedly connected through bolts.
Gaps are reserved between any two adjacent trunk shells 3 and between the first trunk shell 3 and the head shell 1. The trunk units are connected end to end in sequence, the front end of the first trunk unit is fixedly connected with the rear end of the head shell 1, and the rear end of the last trunk unit is fixedly connected with the tail shell 2.
Specifically, the steering driving mechanism 4 includes a biaxial servo motor 41 and a steering bracket 42, the biaxial servo motor 41 is fixed in the trunk housing 3, specifically, the inner side of each trunk housing 3 is fixedly provided with a motor bracket 5, the biaxial servo motor 41 is fixed on the motor bracket 5, and two output shafts thereof are vertically arranged.
Preferably, the motor bracket 5 comprises two bracket units 51, the bracket units 51 comprise a T-shaped structure formed by connecting a transverse rod and a longitudinal rod fixedly, and the two bracket units 51 are arranged in an axisymmetric mode. One end of each of the two cross bars, which are far away from each other, is fixedly connected with the inner side wall of the trunk shell 3, and the other end of each of the two cross bars is fixedly inserted with the corresponding support single 51 longitudinal rod to form a closed area capable of accommodating the double-shaft servo motor 41.
Two floater blocks 6 made of buoyancy materials are symmetrically arranged on the left side and the right side of the double-shaft servo motor 41, and the two floater blocks 6 are fixedly connected with the inner side wall of the trunk shell 3.
The steering bracket 42 is positioned at the front side of the double-shaft servo motors 41 and fixedly connected with the output shafts of the double-shaft servo motors, and the connecting frame 7 is fixed at the rear side of each double-shaft servo motor 41. The top and bottom of each of the connection frames 7 are provided with connection terminals 71, and each connection terminal 71 is connected to the rear end of the head housing 1 through a rib 72 made of an elastic material.
The front end of the steering bracket 42 in the first trunk unit is fixedly connected with the rear end of the head housing 1, and the front ends of the steering brackets 42 in the remaining trunk units are fixedly connected with the rear ends of the connecting frames 7 in the trunk units adjacent to the front side. The rear end of the connecting frame 7 in the last trunk unit is fixedly connected with the front end of the tail shell 2. And each trunk unit is matched with the fish head to realize the integral swing of the sturgeon-like robot and provide the advancing power for the sturgeon-like robot.
The control system comprises a controller, a sonar, an image acquisition module and a signal transmission module, wherein the image acquisition module is arranged at the front end of the head shell 1, and the sonar, the image acquisition module and the signal transmission module are connected with the controller through signals. The sonar adopts active sonar echo to detect the general position of the submarine cable, the instrument can send sound waves into the sea, and the time spent by echo transmission back to the robot fish can be used for calculating the shape and the position of the submarine cable under the sturgeon-like robot. The active sonar can accurately measure the target distance and can detect a fixed target.
The invention is based on the principle of bionics, and is designed by simulating the sturgeon in nature from the practical application of the engineering of functional bionics. The shape of the sturgeon in nature is simulated, and the tip of the head is slightly tilted, so that the obstacle surmounting of the sturgeon-simulated robot is facilitated; the sturgeon tail has long upper end and short lower end and can closely advance along the surface of the cable; the sturgeon-like tendons 72 resemble sturgeon's keel, allowing the mechanical structure to have a stronger attachment mechanism. The active sonar is adopted to remotely sense the position of the submarine cable, the camera is used for simulating the eyes of sturgeons in a short distance, and the problem of image blurring caused by nonuniform submarine light, turbid water quality and the like is solved, so that the fault detection of the submarine cable is realized based on a convolutional neural network optimized by particle swarm. The submarine cable detection method adopts a bionic principle, effectively overcomes the defects of the traditional method, adopts active sonar detection to simulate sturgeon kiss whiskers, detects the position of the submarine cable and is further used for detecting the fault type of the submarine cable. Therefore, on one hand, the active sonar reduces the influence of external noise on the active sonar in principle, and on the other hand, the position of the submarine cable can be determined through multiple detection, so that the requirements on the aspect of realized functions are reduced.
In embodiment 2, the submarine cable detection mainly relies on the data collected by the camera to perform corresponding work.
There are various classification modes of submarine cables, and the submarine cables are classified into 3 types according to the fault handling modes so as to carry out subsequent processing: (1) The failure of hanging up of the ship anchor, dragging of the fishing net, biting of the fish and the like caused by external force needs to be performed with the operation of reconnecting the line; (2) Suspending, the pipeline is suspended along with scouring of submarine hidden flows and the like, and the faults are solved by adding a sleeve; (3) And the electrical faults, such as bulges, deformation and the like, which occur in the abnormal delivery of the cable, need to be further overhauled, so that further damage is avoided. Therefore, the invention mainly judges which type of the suspended electric faults caused by the normal state and the external force belongs to by the sturgeon-like robot.
The submarine cable fault detection method based on the sturgeon-like robot comprises the following steps:
step one, a convolutional neural network prediction model is established, the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full-connection layer and an output layer, and a submarine cable fault detection convolutional neural network mathematical model with optimal performance is obtained through learning and training. In the first step, a submarine cable detection prediction model is established by adopting the following steps:
s1: the method comprises the steps of collecting images, horizontally overturning, randomly buckling, performing scale transformation, rotating, expanding submarine cable data images, and establishing a data set containing normal submarine cables and suspended and electric faults caused by external force.
The data set of the convolutional neural network is divided into a training subset, a verification subset and a test subset randomly and averagely, the training set is used for learning model parameters of the convolutional neural network, the verification subset selects super parameters by evaluating network performance, and the test subset measures network performance by generalization errors.
S2: and training the super parameters of the convolutional neural network, namely the number of convolutional layers, the size of the convolutional kernels and the super parameters of the number of the convolutional kernels by using a particle swarm optimization algorithm, and optimizing.
S3: training the convolutional neural network by using the optimized number of convolutional layers, the convolutional kernel size and the convolutional number, and importing the data into the input layer, the pooling layer and the full-connection layer according to the data, wherein the convolutional neural network can extract the fault characteristics of the training.
S4: and calculating an output result loss function, updating weights through a gradient descent method, and realizing multiple iterations of the convolutional neural network to enable the training model to be converged.
S5: after multiple iterations, the test set is adopted to judge the convolutional neural network classification performance based on particle swarm optimization.
S6: and according to the performance evaluation result, adjusting parameters of node number, iteration times and loss function value in the convolutional neural network.
S7: repeating the steps to finally obtain the submarine cable fault detection convolutional neural network mathematical model with the optimal performance.
Specifically, the image acquisition module comprises a camera provided with a light source, wherein in the first step, the input layer is a submarine cable image, the input layer expands submarine cable data images, and the images are further enhanced.
The convolution layer is convolved with the convolution kernel by the feature vector and achieves the feature extraction of the submarine image by activating the function response.
The mathematical expression of the convolution layer neuron is:
Figure SMS_36
(1)
wherein,,
Figure SMS_41
convolutional layer for convolutional neural network>
Figure SMS_39
Is>
Figure SMS_45
The outputs of the individual channels;
Figure SMS_40
Convolutional layer for convolutional neural network>
Figure SMS_44
Is>
Figure SMS_50
The outputs of the individual channels;
Figure SMS_53
For calculating->
Figure SMS_46
A subset of the input feature maps for net activation of the individual channels;
Figure SMS_51
Is a convolution operation symbol;
Figure SMS_37
Is a convolution layer->
Figure SMS_43
Input vector->
Figure SMS_42
And neuron->
Figure SMS_48
A weight matrix of connections;
Figure SMS_49
Is a convolution layer->
Figure SMS_52
First->
Figure SMS_38
Deviation value of individual characteristic diagram, +.>
Figure SMS_47
To activate the function.
The pool layer chemical expression is:
Figure SMS_54
(2)
wherein,,
Figure SMS_55
pooling layer for convolutional neural network>
Figure SMS_56
First->
Figure SMS_57
The weight coefficients of the individual channels;
Figure SMS_58
Is a pooling function;
Figure SMS_59
For pooling layer->
Figure SMS_60
First->
Figure SMS_61
Offset terms for the individual channels.
The full connection layer mathematical expression is as follows:
Figure SMS_62
(3)
wherein,,
Figure SMS_64
is a full connection layer->
Figure SMS_67
An output of (2);
Figure SMS_69
Is a full connection layer->
Figure SMS_65
Is a network weight coefficient of (a);
Figure SMS_66
Is a full connection layer
Figure SMS_68
An output of (2);
Figure SMS_70
Is a full connection layer->
Figure SMS_63
Is included.
The output layer gives specific classification results and judges the normal state of the submarine cable and specific types of suspended and electric faults caused by external force.
Placing the sturgeon-like robot in water and submerging the sturgeon-like robot to a deep water area, enabling the sonar to send out a signal in the swimming process to detect the position of the submarine cable, and enabling the sonar to send out the signal to a controller after determining the position of the submarine cable, wherein the sturgeon-like robot reaches the position of the submarine cable.
Step three, the sturgeon-like robot moves along the extending direction of the submarine cable, the image acquisition module acquires the image information of the submarine cable in real time and sends the image information to the controller, the state and the fault type of the submarine cable are judged through the convolutional neural network, and the data information is stored.
And fourthly, after the sturgeon-like robot upwards floats out of the water, the data information is sent to a receiving terminal on the ship or the ground through a signal transmitting module, and the position sent by the signal is positioned through the base station so as to determine the position coordinates of the fault point. The submarine cable fault position is just below the position of the sturgeon-like robot floating out of the water level, the type and the state of the fault can be known through the received data information, and a maintenance scheme is further formulated.
And fifthly, repeating the process of the step four, sequentially determining each fault point of the submarine cable and the fault type of each fault point, and making a fault processing scheme by staff according to the data information.
In the process of detecting the submarine cable by adopting the convolutional neural network, compared with the traditional optical-based visual detection, the deep learning algorithm based on the convolutional neural network has good model generalization capability, can adapt to the submarine complex environment, has good identification and classification capability, and can effectively identify various faults of the submarine cable.
The parts not described in the invention can be realized by adopting or referring to the prior art.
The embodiments of the invention have been presented for purposes of illustration and description, and are not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.
It should be understood that the above description is not intended to limit the invention to the particular embodiments disclosed, but to limit the invention to the particular embodiments disclosed, and that the invention is not limited to the particular embodiments disclosed, but is intended to cover modifications, adaptations, additions and alternatives falling within the spirit and scope of the invention.

Claims (7)

1. The submarine cable fault detection method comprises a sturgeon-like robot, wherein the sturgeon-like robot comprises a fish head part, a trunk unit, a fish tail part and a control system, and is characterized in that the fish head part comprises a sturgeon-like head shell, two pectoral fins are symmetrically arranged on the left side and the right side of the head shell, and dorsal fins are arranged on the top of the head shell;
the fish tail part is positioned at the rear side of the fish head part and is connected with the rear side of the head shell through a plurality of trunk units which are sequentially arranged, the fish tail part comprises a sturgeon-like tail shell, and the rear end of the tail shell is provided with tail fins;
the trunk unit comprises a trunk shell and a steering driving mechanism arranged in the trunk shell, the trunk units are connected end to end in sequence, the front end of the first trunk unit is fixedly connected with the rear end of the head shell, and the rear end of the last trunk unit is fixedly connected with the tail shell;
the trunk units are matched with the fish heads to realize the integral swing of the sturgeon-like robot, and the forward power of the sturgeon-like robot is provided;
the control system comprises a controller, a sonar, an image acquisition module and a signal emission module, wherein the image acquisition module is arranged at the front end of the head shell, and the sonar, the image acquisition module and the signal emission module are in signal connection with the controller;
the submarine cable fault detection method comprises the following steps:
step one, a convolutional neural network prediction model is established, the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full-connection layer and an output layer, and a submarine cable fault detection convolutional neural network mathematical model with optimal performance is obtained through learning and training;
placing the sturgeon-like robot in water and submerging the sturgeon-like robot to a deep water area, wherein the sonar sends out a signal to detect the position of the submarine cable in the swimming process, and after the position of the submarine cable is determined, the sonar sends out a signal to a controller, so that the sturgeon-like robot reaches the position of the submarine cable;
step three, the sturgeon-like robot moves along the extending direction of the submarine cable, the image acquisition module acquires the image information of the submarine cable in real time and sends the image information to the controller, the state and the fault type of the submarine cable are judged through the convolutional neural network, and the data information is stored;
step four, after the sturgeon-like robot floats upwards out of the water, the data information is sent to a receiving terminal on a ship or on the ground through a signal transmitting module, and the position sent by a base station positioning signal is used for determining the position coordinates of the fault point;
step five, repeating the process of the step four, sequentially determining each fault point and fault type of the submarine cable, and making a fault processing scheme by staff according to the data information;
in the first step, the submarine cable detection prediction model is established, which comprises the following steps:
s1: collecting images, horizontally overturning, randomly buckling, scaling, rotating and expanding submarine cable data images, and establishing a data set containing normal submarine cables and suspended electrical faults caused by external force;
the data set of the convolutional neural network is divided into a training subset, a verification subset and a test subset randomly and averagely, wherein the training subset is used for learning model parameters of the convolutional neural network, the verification subset selects super parameters by evaluating network performance, and the test subset measures network performance by generalization errors;
s2: training the super parameters of the convolutional neural network, namely the number of convolutional layers, the size of the convolutional kernels and the super parameters of the number of the convolutional kernels by using a particle swarm optimization algorithm, and optimizing;
s3: training the convolutional neural network by using the optimized number of convolutional layers, the size of the convolutional kernel and the number of the convolutional kernel, and importing data into an input layer, a pooling layer and a full-connection layer according to the data, wherein the convolutional neural network can extract the fault characteristics of the training;
s4: calculating an output result loss function, updating weights through a gradient descent method, and realizing multiple iterations of the convolutional neural network to enable the training model to be converged;
s5: after multiple iterations, judging the convolutional neural network classification performance based on particle swarm optimization by adopting a test subset;
s6: according to the performance evaluation result, parameters of node number, iteration times and loss function value in the convolutional neural network are adjusted;
s7: repeating the steps to finally obtain the submarine cable fault detection convolutional neural network mathematical model with the optimal performance.
2. The submarine cable fault detection method according to claim 1, wherein the trunk shells are annular shells, and gaps are reserved between any two adjacent trunk shells and between the first trunk shell and the head shell;
the steering driving mechanism comprises a double-shaft servo motor and steering brackets, wherein the double-shaft servo motor is fixed in the trunk shell, the steering brackets are positioned on the front sides of the double-shaft servo motors and fixedly connected with output shafts of the double-shaft servo motor, and connecting frames are fixed on the rear sides of the double-shaft servo motors.
3. The submarine cable fault detection method according to claim 2, wherein the front ends of the steering brackets in the trunk units of the first order are fixedly connected to the rear ends of the head housings, and the front ends of the steering brackets in the trunk units of the remaining orders are fixedly connected to the rear ends of the connecting frames in the trunk units adjacent to the front sides;
the rear end of the connecting frame in the last trunk unit is fixedly connected with the front end of the tail shell.
4. The submarine cable fault detection method according to claim 2, wherein a motor bracket is fixedly arranged on the inner side of each trunk shell, a double-shaft servo motor is fixedly arranged on the motor bracket, and two output shafts of the double-shaft servo motor are vertically arranged;
two float blocks made of buoyancy materials are symmetrically arranged on the left side and the right side of the double-shaft servo motor, and the two float blocks are fixedly connected with the inner side wall of the trunk shell.
5. The submarine cable fault detection method according to claim 4, wherein each trunk shell comprises two shell monomers which are arranged oppositely, the upper end of each shell monomer is fixedly provided with a hinging block, and the two hinging blocks of the same trunk shell are arranged in a staggered manner and are hinged through a rotating shaft;
the adjacent one side of two shell monomer bottoms all is fixed with the bolt piece, and two bolt pieces of same truck shell bottom just link to each other through the bolt fastening.
6. The submarine cable fault detection method according to claim 4, wherein the motor support comprises two support monomers, the support monomers comprise a T-shaped structure formed by connecting a transverse rod and a longitudinal rod stationary phase, and the two support monomers are arranged in an axisymmetric mode;
one end that two horizontal poles kept away from each other all links to each other with the inside wall fixed of truck shell, and the other end is fixed pegged graft with corresponding support monomer longitudinal rod, encloses into the closed region that can place biax servo motor.
7. A submarine cable fault detection method according to claim 3, wherein the top and bottom of each of the connector frames are provided with connector terminals, each of which is connected to the rear end of the head case by a single wire made of an elastic material;
a transverse shaft is fixedly connected to one side, close to the head shell, of each pectoral fin, and is in rotary sealing fit with the side wall of the head shell, and one end, far away from the pectoral fin, of each transverse shaft is connected with a first servo motor;
the fin is vertically arranged, the bottom of the fin is fixedly connected with a vertical shaft, the vertical shaft is in rotary sealing fit with the top wall of the head shell, and the lower end of the vertical shaft is connected with the output end of the second servo motor fixed in the head shell.
CN202310109397.7A 2023-02-14 2023-02-14 Sturgeon-imitating robot and submarine cable fault detection method Active CN115871901B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310109397.7A CN115871901B (en) 2023-02-14 2023-02-14 Sturgeon-imitating robot and submarine cable fault detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310109397.7A CN115871901B (en) 2023-02-14 2023-02-14 Sturgeon-imitating robot and submarine cable fault detection method

Publications (2)

Publication Number Publication Date
CN115871901A CN115871901A (en) 2023-03-31
CN115871901B true CN115871901B (en) 2023-05-16

Family

ID=85761073

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310109397.7A Active CN115871901B (en) 2023-02-14 2023-02-14 Sturgeon-imitating robot and submarine cable fault detection method

Country Status (1)

Country Link
CN (1) CN115871901B (en)

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101913419A (en) * 2010-08-11 2010-12-15 中国科学院自动化研究所 Biomimetic robotic dolphin
CN102303700B (en) * 2011-05-26 2013-09-04 中国科学院自动化研究所 Multiple control surface robotic fish with embedded vision
CN109625219A (en) * 2018-11-01 2019-04-16 国网浙江省电力有限公司 There is cable remote underwater robot to the cruising inspection system and method for failure submarine cable
CN111340868B (en) * 2020-02-26 2023-06-02 大连海事大学 Unmanned underwater vehicle autonomous decision control method based on visual depth estimation
US20240153261A1 (en) * 2021-02-24 2024-05-09 Thales Canada Inc. Method of and system for performing object recognition in data acquired by ultrawide field of view sensors
CN112918646A (en) * 2021-03-15 2021-06-08 武汉理工大学 Pectoral fin and tail fin cooperative control system and method for bionic robot fish
CN215851811U (en) * 2021-08-17 2022-02-18 宜宾维特瑞安科技有限公司 Bionic white sturgeon robot
CN113799948A (en) * 2021-09-13 2021-12-17 广东电网有限责任公司 Portable submarine cable inspection unmanned underwater vehicle
CN114354082B (en) * 2022-03-18 2022-05-31 山东科技大学 Intelligent tracking system and method for submarine pipeline based on imitated sturgeon whisker
CN217074764U (en) * 2022-04-14 2022-07-29 武昌工学院 Bionic fish for submarine detection

Also Published As

Publication number Publication date
CN115871901A (en) 2023-03-31

Similar Documents

Publication Publication Date Title
WO2022198899A1 (en) Submarine cable fault diagnosis method and apparatus, and device
CN108254157B (en) A kind of experimental system of interior wave and submerged body interaction
CN106568496A (en) Real-time transmission multivariate vector hydrophone array subsurface buoy system
CN103518143A (en) Sonar data collection system
CN114218864A (en) Net cage netting damage detection method and device based on mathematics twinning and storage medium
Kristiansen et al. Numerical and experimental study on the seakeeping behavior of floating closed rigid fish cages
CN107172399A (en) The alarm method of seashore outdoor bathing place visitor's ascending and descending condition
CN117214904A (en) Intelligent fish identification monitoring method and system based on multi-sensor data
CN115871901B (en) Sturgeon-imitating robot and submarine cable fault detection method
CN114572347A (en) Tsunami early warning monitoring system
CN114692520A (en) Multi-scene-oriented unmanned ship virtual simulation test platform and test method
CN108828328B (en) Portable three-component submarine electric field instrument
CN111959708B (en) Ship draft simulation system
CN112046682B (en) Small buoy of ecological monitoring net of marine ranching
Sethuraman et al. Towards sim2real for shipwreck detection in side scan sonar imagery
CN111985118A (en) Analysis method for calculating overall structure mechanical characteristics of offshore tubular pile fence culture platform
CN208255317U (en) Portable three-component submarine electric field instrument
KR102121246B1 (en) Damage detection method for mooring lines of submersible structures based on deep learning
CN110065600A (en) One kind is with the fixed high underwater navigation equipment of stream
CN117454927B (en) Detection method for cooperatively checking oil gas leakage of submarine pipeline by multiple unmanned underwater vehicles
CN114662569B (en) Digital twinning-based underwater glider fault real-time diagnosis method
CN219728521U (en) Power accelerating device for deep sea detector
Person et al. ASSEM: a new concept of observatory applied to long term SEabed Monitoring of geohazards
León Hernando et al. Experimental Investigations on a Highly Flexible Submerged Structure in Waves
Liu et al. Genetic Algorithm and LSTM Based Approach for Diver Search and Path Prediction

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
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20230331

Assignee: Qingdao Zhongshi Chuanzhi Technology Co.,Ltd.

Assignor: SHANDONG University OF SCIENCE AND TECHNOLOGY

Contract record no.: X2024980005552

Denomination of invention: A Sturgeon like Robot and Submarine Cable Fault Detection Method

Granted publication date: 20230516

License type: Common License

Record date: 20240510

EE01 Entry into force of recordation of patent licensing contract