CN110866548A - Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line - Google Patents

Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line Download PDF

Info

Publication number
CN110866548A
CN110866548A CN201911053900.1A CN201911053900A CN110866548A CN 110866548 A CN110866548 A CN 110866548A CN 201911053900 A CN201911053900 A CN 201911053900A CN 110866548 A CN110866548 A CN 110866548A
Authority
CN
China
Prior art keywords
insulator
aerial vehicle
unmanned aerial
neural network
convolutional neural
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.)
Pending
Application number
CN201911053900.1A
Other languages
Chinese (zh)
Inventor
刘洋
高嵩
黄强
张量
毕晓甜
张迺龙
陈杰
赵恒�
邱刚
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.)
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd
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 State Grid Corp of China SGCC, Electric Power Research Institute of State Grid Jiangsu Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN201911053900.1A priority Critical patent/CN110866548A/en
Publication of CN110866548A publication Critical patent/CN110866548A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C3/00—Measuring distances in line of sight; Optical rangefinders
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S11/00—Systems for determining distance or velocity not using reflection or reradiation
    • G01S11/12—Systems for determining distance or velocity not using reflection or reradiation using electromagnetic waves other than radio waves
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Electromagnetism (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a method and a system for infrared intelligent matching identification and distance measurement positioning of an insulator of a power transmission line, which comprises the steps of automatically identifying an insulator string target by using a trained convolutional neural network; after the insulator sub-targets are confirmed, tracking and positioning the insulator string targets through position-posture outer ring and inner ring control; selecting the optimal shooting distance by monocular distance measurement based on a similar triangle principle; and the unmanned aerial vehicle and the insulator string target keep the best shooting distance to perform safety distance early warning.

Description

Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line
Technical Field
The invention relates to the technical field of target detection and artificial intelligence intersection, in particular to a method and a system for infrared intelligent matching identification and distance measurement positioning of an insulator string of a power transmission line.
Background
The insulator is used as an important component on the power transmission line, the running condition and the quality degree of the insulator directly relate to the stability and the safety of a power grid, and the insulator has very important significance for the uninterrupted power detection and fault diagnosis work of the insulator. The unmanned aerial vehicle carrying thermal infrared imager for line inspection is a field non-contact detection method which is started in recent years, and has the advantages of high efficiency, small safety risk, low cost and the like, but at present, a plurality of technical problems of infrared inspection still exist and need to be solved urgently. Firstly, the overhead transmission line is often remote, the height of the insulator string from the ground reaches tens of meters, the unmanned aerial vehicle flyers can only overlook by eyes and display infrared images of remote control terminals, and the unmanned aerial vehicle is manually operated by experience to approach the insulator sub-targets and shoot. The manual target identification mode has high labor intensity and low efficiency. Therefore, how to realize the insulator infrared image intelligent matching identification technology becomes a primary problem. Secondly, the shooting distance between unmanned aerial vehicle and the target insulator is very critical. If the distance is too far, the shooting effect is poor, and the analysis of the later-stage insulator image is not favorable; if the distance is too close, potential safety hazards such as unmanned aerial vehicle collision exist. In the actual shooting process, the unmanned aerial vehicle is very difficult to control because of the complexity of the terrain of the surrounding environment and the limitation of the sight angle of the flyer, and the safety distance is not easy to keep. How to let unmanned aerial vehicle and target insulator automatically keep best shooting distance and carry out safe distance early warning becomes another important problem.
Disclosure of Invention
The purpose of the invention is as follows: the invention provides a method and a system for infrared intelligent matching identification and distance measurement positioning of a power transmission line insulator string, aiming at the problems in the prior art.
The technical scheme is as follows: the invention provides an infrared intelligent matching identification and distance measurement positioning method for an insulator of a power transmission line, which comprises the following steps of:
step 1: adopting a trained convolutional neural network to automatically identify an insulator string target;
step 2: tracking and positioning the insulator string target by adopting position-posture outer-inner ring control;
and step 3: acquiring the optimal shooting distance according to the position of the insulator string target;
and 4, step 4: based on the best shooting distance, the unmanned aerial vehicle carries out safety distance early warning on the insulator string target.
Further, the training step of the convolutional neural network in step 1 is as follows:
constructing a convolutional neural network;
pre-training the convolutional neural network by adopting an image database, and after convergence, modifying the last layer of the convolutional neural network into C (3), wherein the C is respectively represented by an insulator string, a background and a tower type 3 target;
performing secondary training by using the image library of the power transmission line, adopting a heuristic method in the training process, and reducing the learning rate to 1/10 at the current time if the error rate of the verification set is unchanged at the current learning rate until convergence to obtain a trained convolutional neural network;
the image database is a Cifar-100 image database; the selection of the power transmission line image library comprises the following steps:
aerial photography is carried out on the power transmission line by adopting an unmanned aerial vehicle, and an image with an insulator string is obtained;
marking 3 types of areas of a background, a tower and an insulator string in the image;
partitioning an image to obtain image blocks;
and (4) carrying out rotation, translation and scale conversion on the image blocks, and expanding the number of samples to obtain an image library of the power transmission line.
Further, the convolutional neural network takes a picture as an input, takes a category label of the picture as an output, and is represented as follows:
c＝FCNN(s|P),c∈{1,2,...,C}
wherein s is an input picture, P is a parameter of the convolutional neural network, c is an output class label of the picture, and FCNN() Representing the forward operation of the convolutional neural network, and representing the calculation of a class label of the picture s according to the parameter P of the known convolutional neural network;
further, the position-posture outer-inner ring controls horizontal displacement (x) of the unmanned aerial vehicle in a world coordinate systemd,yd) As inputs, specifically include: the outer ring position control adopts a PI controller to obtain an expected angle of inner ring attitude control, and the inner ring attitude control obtains a final control quantity U according to the expected angle;
horizontal displacement (x) of the drone in the world coordinate systemd,yd) From the tracking error e of the insulator string target in the image planecObtaining the tracking error ecExpressed as:
Figure BDA0002256034490000021
in the formula (I), the compound is shown in the specification,
Figure BDA0002256034490000022
and cpRespectively representing the pixel coordinates of the center of the image visual field and the pixel coordinates of the center of the target;
the PI controller obtains a height control law U1And inner ring attitude controlDesired angle phidAnd thetad：
Figure BDA0002256034490000023
In the formula of U1Denotes the total lift of the motor, mu ═ mux,μy,μz]TFor a virtual control quantity, phid，θd，ψdRespectively the roll angle, pitch angle and yaw angle of the unmanned aerial vehicle;
the inner ring attitude control adopts LQR control to obtain a control law U2Control law U3Control law U4；
U＝[U1,U2,U3,U4]TConstitute unmanned aerial vehicle's control input.
Further, the step 3 specifically includes:
obtaining pixels occupied by the width of the insulator string target in the aerial video frame according to the position of the insulator string target;
according to the known diameter of a single insulator, the focal length of an airborne camera of the unmanned aerial vehicle and the occupied pixel of the target width of the insulator string, calculating the distance D between the target insulator string and the unmanned aerial vehicle based on the similar triangle principle:
Figure BDA0002256034490000031
in the formula, f is the focus of unmanned aerial vehicle machine carries camera, and H represents the actual diameter of single insulator, and H represents the shared pixel of insulator diameter in the formation of image.
Further, the step of obtaining the focal length of the airborne camera of the unmanned aerial vehicle is as follows:
placing the airborne camera of the unmanned aerial vehicle in a fixed position, obtaining a plurality of calibration images at different angles by changing the orientation of a calibration object, carrying out batch processing on the calibration images by utilizing a camera calibration program in MATLAB, reading the calibration images into a camera calibration toolbox, and obtaining the focal length f of the airborne camera of the unmanned aerial vehicle according to a Zhang's plane calibration method.
The invention also provides a transmission line insulator automatic matching identification and distance measurement positioning system, which comprises:
the convolutional neural network module is used for automatically identifying the insulator string target;
the position-posture outer-inner ring control module is used for tracking and positioning the identified insulator string target;
the distance measurement module is used for calculating and obtaining the optimal shooting distance between the unmanned aerial vehicle and the insulator string target which is currently tracked and positioned;
and the unmanned aerial vehicle is used for keeping the best shooting distance with the insulator string target according to the best shooting distance obtained by the ranging module to perform safety distance early warning.
Further, the convolutional neural network module comprises a 1-layer input layer, a 3-layer convolution/pooling layer and 1 fully-connected output layer; the input of the input layer is a normalized RGB three-channel picture, the number of channels of each convolution kernel of the convolution/pooling layer is determined by the number of channels output by the previous layer, the pooling is performed by adopting maximum pooling, and the number of neurons of the full-connection layer is the same as the number of output categories, and is represented as follows:
c＝FCNN(s|P),c∈{1,2,...,C}
wherein s is an input picture, P is a parameter of the convolutional neural network, c is an output class label of the picture, and FCNN() Representing the forward operation of the convolutional neural network, and representing the calculation of a class label of the picture s according to the parameter P of the known convolutional neural network;
the training process of the convolutional neural network is a process of solving the parameter P of the convolutional neural network.
Further, the position-posture outer-inner ring control module comprises an outer ring position controller and an inner ring posture controller, the outer ring position controller obtains an expected angle of the inner ring posture controller, the inner ring posture controller obtains a final control quantity U according to the expected angle, and the final control quantity U forms the control input of the unmanned aerial vehicle; the method specifically comprises the following steps:
the height control law U is directly obtained by an outer ring position controller1And inner ring attitude controllersDesired angle phidAnd thetad：
Figure BDA0002256034490000032
In the formula of U1Denotes the total lift of the motor, mu ═ mux,μy,μz]TFor a virtual control quantity, phid，θd，ψdRespectively the roll angle, pitch angle and yaw angle of the unmanned aerial vehicle.
The inner ring attitude controller is controlled by LQR to obtain a control law U2，U3，U4；
Obtaining the final control quantity U ═ U1,U2,U3,U4]TConstitute unmanned aerial vehicle's control input.
Further, the distance measuring module calculates the distance D between the target insulator string and the unmanned aerial vehicle based on the similar triangle principle:
Figure BDA0002256034490000041
in the formula, f is the focus of unmanned aerial vehicle machine carries camera, and H represents the actual diameter of single insulator, and H represents the shared pixel of insulator diameter in the formation of image.
Further, the outer ring position controller adopts a PI controller.
Furthermore, the unmanned aerial vehicle is an unmanned aerial vehicle with a thermal infrared imager.
Has the advantages that: compared with the prior art, the method can improve the safety and the intelligent level of the infrared aerial photography of the unmanned aerial vehicle, can also improve the imaging quality and the aerial photography efficiency, and has very important practical significance.
Drawings
FIG. 1 is a sample configuration of the present invention;
FIG. 2 is a model training of the present invention;
FIG. 3 is an overall structure of the tracking controller of the present invention;
fig. 4 illustrates the triangle-like distance measurement principle of the present invention.
Detailed Description
The basic idea of the invention is as follows: the convolutional neural network algorithm is characterized in that a large amount of training data is collected, and a special training skill is added in the training process to prevent overfitting, so that the classification and identification accuracy of the convolutional neural network algorithm in the computer vision field is greatly superior to that of the traditional algorithm, and the accurate identification of a target can be realized; the position-posture outer-inner ring controller realizes that the target is always positioned at the visual field center position of the airborne camera; the monocular distance measuring module based on the similar triangle can measure the distance between the camera and the target in real time.
Example (b):
based on the principle, the insulator is automatically identified by using a convolutional neural network, the insulator string is tracked and positioned by using an outer ring controller and an inner ring controller, the best shooting distance is selected by using a monocular distance measuring module based on a similar triangle principle, and when the distance between the unmanned aerial vehicle and a target is smaller than a safe distance, an alarm signal is generated by a remote controller. The method comprises the following specific steps:
1. selection and labeling of samples
As shown in fig. 1, firstly, an unmanned aerial vehicle is used to take an aerial photograph of the power transmission line, a large number of image libraries containing insulator strings are collected, and 3 types of areas including the background, the tower and the insulator strings in the original image are marked by polygonal frames in a manual mode. After the target labeling is finished, partitioning the image, and sequentially assigning class labels to the classifications: insulator string-1, background-2 and tower-3. In order to ensure the generalization of the CNN network, the image blocks are subjected to rotation, translation and scale transformation, and the number of samples is expanded.
2. Convolutional neural network structure design
According to an AlexNet model, the convolutional neural network is designed into 5 layers, namely 1 input layer, 3 convolutional/pooling layers and a fully-connected output layer. Inputting normalized pictures with 64 multiplied by 64 and RGB three channels; the sizes of the 3 convolution kernels are respectively 11 multiplied by 11, 7 multiplied by 7 and 5 multiplied by 5, the number of channels of each layer of convolution kernels is respectively determined by the number of channels output by the previous layer, namely respectively3. 64 and 128, wherein the pooling adopts maximum pooling, and the window size is 2 multiplied by 2; the number of full-connectivity layer neurons is the same as the number of output classes, where C represents the number of network output labels, and typically defines the output values as elements in the set {1,2, …, C }. Therefore, the whole convolutional neural network can be regarded as a black box with parameters inside, the input is a picture with the same size, and the output is a class label of the picture, namely c ═ FCNN(s | P), C ∈ {1, 2.., C }, wherein s is an input picture, P is a parameter of the neural network model, C is an output class label of the image, and F is the output class label of the imageCNN() And (4) representing the forward operation of the deep convolutional neural network, namely the known model parameter P, and calculating the class label of the picture s.
3. Model training
As shown in fig. 2, the learning process of the neural network is a process of solving the model parameters P. Defining C as 100, namely 100 output categories, and performing preliminary training by using a Cifar-100 image database. After the algorithm is converged, the last layer of the model is changed into C-3, which respectively represents 3 types of targets of an insulator string, a background and a tower. And then training by using the power transmission line image library until convergence. The feature extraction parameters obtained by the Cifar-100 image database training can be converted into the feature extraction parameters of the power transmission line image. On the basis, network parameter optimization is carried out, and the overfitting phenomenon is effectively avoided. After the network structure is defined, the parameters of the neural network model are trained by adopting a gradient descent method, and the training process is finished by Caffe. The learning parameters in the pre-training phase are:
(1) the training data chunk (Batch) size is 128.
(2) All weights are initialized to gaussian noise with a mean of 0 and a standard deviation of 0.01.
(3) The coefficient momentum (momentum) is 0.9 and the weight decay coefficient is 0.0005.
(4) For neuron biases at convolutional and fully-connected layers 2, 3, 4, the initialization is 1, and the neuron bias at layer 1 is initialized to 0.
(5) The learning rate of each layer remains the same, and the initial learning rate is 0.01. A heuristic method is adopted in the training process: at the current learning rate, if the validation set error rate is unchanged, the learning rate is reduced to current 1/10, and finally the algorithm converges to obtain model parameters P for the network. Therefore, the convolutional neural network model training is completed, and the insulator string is positioned.
4. Position-attitude outer and inner loop tracker design
The unmanned aerial vehicle is taken as a rigid body, and a mathematical model of the unmanned aerial vehicle is obtained according to a force balance equation of the unmanned aerial vehicle under a world coordinate system and a moment balance equation of a body coordinate system
Figure BDA0002256034490000061
Wherein η ═ phi, theta, psi]TRoll angle, pitch angle and yaw angle of the drone, ξ ═ x, y, z]TThe coordinate position of the unmanned aerial vehicle under a world coordinate system; i ═ diag [ I ═ Ix,Iy,Iz]An inertia matrix of the unmanned aerial vehicle along 3 coordinate axes of a coordinate system of the body; m is unmanned aerial vehicle's quality, and d is the distance of organism center to the motor axle center, and K is the resistance coefficient.
Unmanned aerial vehicle's control input U ═ U1,U2,U3,U4]TIs composed of
Figure BDA0002256034490000062
In the formula Fi＝1,...4Lift generated by 4 motors.
And a wireless camera arranged on a platform below the center of the unmanned aerial vehicle captures images in real time, and the horizontal displacement between the unmanned aerial vehicle and the target is obtained through the calculation of the pixel position of the target. From the knowledge of the camera geometry, the pixel position (u) of the object in the image planei,vi) And the coordinate transformation relationship between the target and the horizontal displacement (x, y) of the drone may be expressed as
Figure BDA0002256034490000063
Wherein z is the flight height of the unmanned aerial vehicle, f is the focal length of the camera, and β is the pitching angle of the video cameraAngle α is the angle between the connecting line of two points of the object image and the optical axis of the camera (u)0,v0) Is the pixel coordinate of the central point of the image.
In order to ensure that the unmanned aerial vehicle can effectively track the target, an outer ring controller and an inner ring controller of the position-posture are designed, and specific reference is made to fig. 3. Outer loop based on position reference [ xd,yd,zd]TObtaining a desired angle phi of inner ring attitudedAnd thetadWherein the position reference value [ xd,yd,zd]TRepresenting the actual spatial displacement of the drone from the target. The inner ring obtains the final control quantity U according to the expected angle. For this purpose, a virtual control variable μ ═ μ is introducedx,μy,μz]TAnd the outer ring position controller calculates:
Figure BDA0002256034490000064
the height control law U is directly obtained by an outer ring position controller1And desired angle phi of inner ring attitudedAnd thetad：
Figure BDA0002256034490000071
And the inner ring attitude control adopts LQR control. Designing a PI controller by outer ring position control:
Figure BDA0002256034490000072
in order to ensure that the unmanned aerial vehicle moves along with the ground target, the target is always positioned near the center of the image sequence obtained by the airborne camera. Tracking error e of an object in an image planecComprises the following steps:
Figure BDA0002256034490000073
in the formula (I), the compound is shown in the specification,
Figure BDA0002256034490000074
and cpRespectively, the center pixel coordinate of the image visual field and the center pixel coordinate of the target. By ecObtaining the horizontal displacement (x) of the target unmanned aerial vehicle in the world coordinate systemd,yd) As an input to the position-attitude outer-inner-loop tracker.
5. Calibration of airborne camera
Printing 10 × 10 checkerboards as calibration objects, wherein the size of each checkerboard is 25mm × 25mm, placing an airborne camera at a fixed position, continuously changing the orientation of the checkerboard, continuously shooting 20 calibration images at different angles, carrying out batch processing on the images by using a camera calibration program in MATLAB, reading the images into a camera calibration toolbox, and obtaining the focal length f of the camera according to a Zhang plane calibration method. And then placing the single insulator at the same horizontal position of the camera for 3 meters, shooting to obtain a picture, and recording the diameter H of the actual disc surface of the insulator and the diameter distance H in the image.
6. Distance measurement of insulator string
As shown in fig. 4, the distance between the target insulator string and the unmanned aerial vehicle is calculated by using the similar triangle principle according to the diameter of the single insulator measured in advance, the focal length of the airborne camera obtained by the calibration program, and the pixel occupied by the width of the insulator string obtained in the aerial photography:
Figure BDA0002256034490000075
as will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (13)

1. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line is characterized by comprising the following steps of: the method comprises the following steps:
step 1: adopting a trained convolutional neural network to automatically identify an insulator string target;
step 2: tracking and positioning the insulator string target by adopting position-posture outer-inner ring control;
and step 3: acquiring the optimal shooting distance according to the position of the insulator string target;
and 4, step 4: and controlling the unmanned aerial vehicle to perform safety distance early warning on the insulator string target based on the optimal shooting distance.
2. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 1, which is characterized in that: the training step of the convolutional neural network in the step 1 is as follows:
constructing a convolutional neural network;
pre-training the convolutional neural network by adopting an image database, and after convergence, modifying the last layer of the convolutional neural network into C (3), wherein the C is respectively represented by an insulator string, a background and a tower type 3 target;
and (3) performing secondary training by using the image library of the power transmission line, adopting a heuristic method in the training process, and reducing the learning rate to current 1/10 until convergence if the error rate of the verification set is unchanged at the current learning rate to obtain the trained convolutional neural network.
3. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 2, characterized in that: the image database is a Cifar-100 image database;
the selection of the power transmission line image library comprises the following steps:
aerial photography is carried out on the power transmission line by adopting an unmanned aerial vehicle, and an image with an insulator string is obtained;
marking 3 types of areas of a background, a tower and an insulator string in the image;
partitioning an image to obtain image blocks;
and (4) carrying out rotation, translation and scale conversion on the image blocks, and expanding the number of samples to obtain an image library of the power transmission line.
4. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 2, characterized in that: the convolutional neural network takes a picture as input and takes a class label of the picture as output, and the convolutional neural network is expressed as follows:
c＝FCNN(s|P),c∈{1,2,...,C}
wherein s is an input picture, P is a parameter of the convolutional neural network, c is an output class label of the picture, and FCNN() Representing a convolutional neural network forward operation.
5. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 1, which is characterized in that: the position-attitude outer and inner rings control horizontal displacement (x) of the unmanned aerial vehicle in a world coordinate systemd,yd) As inputs, specifically include: the outer ring position control adopts a PI controller to obtain an expected angle of inner ring attitude control, and the inner ring attitude control obtains a final control quantity U according to the expected angle;
horizontal displacement (x) of the drone in the world coordinate systemd,yd) From the tracking error e of the insulator string target in the image planecObtaining the tracking error ecExpressed as:
Figure FDA0002256034480000021
in the formula (I), the compound is shown in the specification,
Figure FDA0002256034480000022
and cpRespectively representing the pixel coordinates of the center of the image visual field and the pixel coordinates of the center of the target;
the PI controller obtains a height control law U1And desired angle phi of inner loop attitude controldAnd thetad：
Figure FDA0002256034480000023
In the formula of U1Denotes the total lift of the motor, mu ═ mux,μy,μz]TFor a virtual control quantity, phid，θd，ψdRespectively the roll angle, pitch angle and yaw angle of the unmanned aerial vehicle;
the inner ring attitude control adopts LQR control to obtain a control law U2Control law U3Control law U4；
U＝[U1,U2,U3,U4]TConstitute unmanned aerial vehicle's control input.
6. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 1, which is characterized in that: the step 3 specifically comprises the following steps:
obtaining pixels occupied by the width of the insulator string target in the aerial video frame according to the position of the insulator string target;
according to the known diameter of a single insulator, the focal length of an airborne camera of the unmanned aerial vehicle and the occupied pixel of the target width of the insulator string, calculating the distance D between the target insulator string and the unmanned aerial vehicle based on the similar triangle principle:
Figure FDA0002256034480000024
in the formula, f is the focus of unmanned aerial vehicle machine carries camera, and H represents the actual diameter of single insulator, and H represents the shared pixel of insulator diameter in the formation of image.
7. The infrared intelligent matching identification and distance measurement positioning method for the insulator of the power transmission line according to claim 6, characterized in that: the step of obtaining the focal length of the airborne camera of the unmanned aerial vehicle is as follows:
placing the airborne camera of the unmanned aerial vehicle in a fixed position, obtaining a plurality of calibration images at different angles by changing the orientation of a calibration object, carrying out batch processing on the calibration images by utilizing a camera calibration program in MATLAB, reading the calibration images into a camera calibration toolbox, and obtaining the focal length f of the airborne camera of the unmanned aerial vehicle according to a Zhang's plane calibration method.
8. The system for the infrared intelligent matching identification and distance measurement positioning method of the power transmission line insulator based on any one of claims 1 to 7 is characterized in that: comprises that
The convolutional neural network module is used for automatically identifying the insulator string target;
the position-posture outer-inner ring control module is used for tracking and positioning the identified insulator string target;
the distance measurement module is used for calculating to obtain the optimal shooting distance between the unmanned aerial vehicle and the insulator string target;
and the unmanned aerial vehicle is used for keeping the best shooting distance with the insulator string target according to the best shooting distance obtained by the ranging module to perform safety distance early warning.
9. The system of claim 8, wherein: the convolutional neural network module comprises 1 layer of input layer, 3 layers of convolution/pooling layers and 1 fully-connected output layer; the input of the input layer is a normalized RGB three-channel picture, the number of channels of each convolution kernel of the convolution/pooling layer is determined by the number of channels output by the previous layer, the pooling is performed by adopting maximum pooling, and the number of neurons of the full-connection layer is the same as the number of output categories, and is represented as follows:
c＝FCNN(s|P),c∈{1,2,...,C}
wherein s is an input picture, P is a parameter of the convolutional neural network, c is an output class label of the picture, and FCNN() Representing a convolutional neural network forward operation.
10. The system of claim 8, wherein: the position-posture outer-inner ring control module comprises an outer ring position controller and an inner ring posture controller, the outer ring position controller obtains an expected angle of the inner ring posture controller, the inner ring posture controller obtains a final control quantity U according to the expected angle, and the final control quantity U forms the control input of the unmanned aerial vehicle; the method specifically comprises the following steps:
the height control law U is directly obtained by an outer ring position controller1And inner ring attitudeDesired angle of controller phidAnd thetad：
Figure FDA0002256034480000031
In the formula of U1Denotes the total lift of the motor, mu ═ mux,μy,μz]TFor a virtual control quantity, phid，θd，ψdRespectively the roll angle, pitch angle and yaw angle of the unmanned aerial vehicle;
the inner ring attitude controller is controlled by LQR to obtain a control law U2，U3，U4；
Obtaining the final control quantity U ═ U1,U2,U3,U4]TConstitute unmanned aerial vehicle's control input.
11. The system of claim 8, wherein: the distance measurement module calculates the distance D between the target insulator string and the unmanned aerial vehicle based on the similar triangle principle:
Figure FDA0002256034480000032
in the formula, f is the focus of unmanned aerial vehicle machine carries camera, and H represents the actual diameter of single insulator, and H represents the shared pixel of insulator diameter in the formation of image.
12. The system of claim 10, wherein: and the outer ring position controller adopts a PI controller.
13. The system of claim 8, wherein: the unmanned aerial vehicle is an unmanned aerial vehicle with a thermal infrared imager.
CN201911053900.1A 2019-10-31 2019-10-31 Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line Pending CN110866548A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911053900.1A CN110866548A (en) 2019-10-31 2019-10-31 Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911053900.1A CN110866548A (en) 2019-10-31 2019-10-31 Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line

Publications (1)

Publication Number Publication Date
CN110866548A true CN110866548A (en) 2020-03-06

Family

ID=69654785

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911053900.1A Pending CN110866548A (en) 2019-10-31 2019-10-31 Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line

Country Status (1)

Country Link
CN (1) CN110866548A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449566A (en) * 2020-03-27 2021-09-28 北京机械设备研究所 Intelligent image tracking method and system for low-speed small target in human-in-loop
CN115409888A (en) * 2022-08-22 2022-11-29 北京御航智能科技有限公司 Method and device for intelligently positioning pole tower in distribution network unmanned aerial vehicle inspection
CN116503491A (en) * 2023-06-26 2023-07-28 安徽大学 Robot dog obstacle ranging and obstacle avoidance method based on camera calibration and vision

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104978580A (en) * 2015-06-15 2015-10-14 国网山东省电力公司电力科学研究院 Insulator identification method for unmanned aerial vehicle polling electric transmission line
CN106895824A (en) * 2017-01-24 2017-06-27 国网四川省电力公司检修公司 Unmanned plane localization method based on computer vision
CN107507172A (en) * 2017-08-08 2017-12-22 国网上海市电力公司 Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray
CN107644234A (en) * 2017-10-12 2018-01-30 成都思晗科技股份有限公司 A kind of insulator of transmission line of electricity comes off defect inspection method
CN108229587A (en) * 2018-02-06 2018-06-29 贵州电网有限责任公司 A kind of autonomous scan method of transmission tower based on aircraft floating state

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104978580A (en) * 2015-06-15 2015-10-14 国网山东省电力公司电力科学研究院 Insulator identification method for unmanned aerial vehicle polling electric transmission line
CN106895824A (en) * 2017-01-24 2017-06-27 国网四川省电力公司检修公司 Unmanned plane localization method based on computer vision
CN107507172A (en) * 2017-08-08 2017-12-22 国网上海市电力公司 Merge the extra high voltage line insulator chain deep learning recognition methods of infrared visible ray
CN107644234A (en) * 2017-10-12 2018-01-30 成都思晗科技股份有限公司 A kind of insulator of transmission line of electricity comes off defect inspection method
CN108229587A (en) * 2018-02-06 2018-06-29 贵州电网有限责任公司 A kind of autonomous scan method of transmission tower based on aircraft floating state

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
周筑博 等: "基于深度卷积神经网络的输电线路可见光图像目标检测", 《液晶与显示》 *
杨蔚 等: "输电线路巡检绝缘子跟踪与测距技术", 《电子科技》 *
马良 等: "四旋翼无人机目标跟踪系统设计", 《控制工程》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449566A (en) * 2020-03-27 2021-09-28 北京机械设备研究所 Intelligent image tracking method and system for low-speed small target in human-in-loop
CN113449566B (en) * 2020-03-27 2024-05-07 北京机械设备研究所 Intelligent image tracking method and system for 'low-small' target of human in loop
CN115409888A (en) * 2022-08-22 2022-11-29 北京御航智能科技有限公司 Method and device for intelligently positioning pole tower in distribution network unmanned aerial vehicle inspection
CN115409888B (en) * 2022-08-22 2023-11-17 北京御航智能科技有限公司 Intelligent positioning method and device for pole tower in inspection of distribution network unmanned aerial vehicle
CN116503491A (en) * 2023-06-26 2023-07-28 安徽大学 Robot dog obstacle ranging and obstacle avoidance method based on camera calibration and vision

Similar Documents

Publication Publication Date Title
CN109523552B (en) Three-dimensional object detection method based on viewing cone point cloud
CN111932588B (en) Tracking method of airborne unmanned aerial vehicle multi-target tracking system based on deep learning
CN110000785B (en) Agricultural scene calibration-free robot motion vision cooperative servo control method and equipment
CN106529538A (en) Method and device for positioning aircraft
CN107729808A (en) A kind of image intelligent acquisition system and method for power transmission line unmanned machine inspection
CN114248893B (en) Operation type underwater robot for sea cucumber fishing and control method thereof
CN110866548A (en) Infrared intelligent matching identification and distance measurement positioning method and system for insulator of power transmission line
CN108229587B (en) Autonomous transmission tower scanning method based on hovering state of aircraft
CN109344878B (en) Eagle brain-like feature integration small target recognition method based on ResNet
CN109858437B (en) Automatic luggage volume classification method based on generation query network
CN106339006A (en) Object tracking method of aircraft and apparatus thereof
Rodriguez-Ramos et al. Towards fully autonomous landing on moving platforms for rotary unmanned aerial vehicles
Cocoma-Ortega et al. Towards high-speed localisation for autonomous drone racing
Koutras et al. Autonomous and cooperative design of the monitor positions for a team of uavs to maximize the quantity and quality of detected objects
Ivanovas et al. Block matching based obstacle avoidance for unmanned aerial vehicle
CN117036989A (en) Miniature unmanned aerial vehicle target recognition and tracking control method based on computer vision
CN113538702A (en) Method for generating underwater scene panoramic image of marine culture area
CN115376125A (en) Target detection method based on multi-modal data fusion and in-vivo fruit picking method based on target detection model
Lei et al. Radial coverage strength for optimization of monocular multicamera deployment
Suprapto et al. The detection system of helipad for unmanned aerial vehicle landing using yolo algorithm
CN113553943B (en) Target real-time detection method and device, storage medium and electronic device
Narayanan et al. A multi-purpose realistic haze benchmark with quantifiable haze levels and ground truth
CN114429435A (en) Wide-field-of-view range target searching device, system and method in degraded visual environment
Charco et al. Transfer Learning from Synthetic Data in the Camera Pose Estimation Problem.
CN117572885A (en) Night tracking method, system and related device based on thermal infrared camera of unmanned aerial vehicle

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20200306

RJ01 Rejection of invention patent application after publication