CN113657280A - Power transmission line target defect detection warning method and system - Google Patents
Power transmission line target defect detection warning method and system Download PDFInfo
- Publication number
- CN113657280A CN113657280A CN202110949156.4A CN202110949156A CN113657280A CN 113657280 A CN113657280 A CN 113657280A CN 202110949156 A CN202110949156 A CN 202110949156A CN 113657280 A CN113657280 A CN 113657280A
- Authority
- CN
- China
- Prior art keywords
- target
- module
- transmission line
- power transmission
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 90
- 230000005540 biological transmission Effects 0.000 title claims abstract description 72
- 230000007547 defect Effects 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000007689 inspection Methods 0.000 claims description 76
- 238000013507 mapping Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 13
- 230000003416 augmentation Effects 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 230000003190 augmentative effect Effects 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 230000002776 aggregation Effects 0.000 claims description 2
- 238000004220 aggregation Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 6
- 230000008859 change Effects 0.000 abstract description 5
- 239000012212 insulator Substances 0.000 description 13
- 238000000605 extraction Methods 0.000 description 4
- 238000011176 pooling Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/20—Checking timed patrols, e.g. of watchman
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The application discloses transmission line target defect detection warning method and system, through the position and the type of the power consumption part in the identification of target recognition model patrols and examines the image, and add corresponding type mark and detection frame, still judge the damage condition of power consumption part according to the coincidence degree of power consumption part and target template through target defect recognition model, wherein, the target template predetermines acquiescence frame unanimous with the scaling ratio of the detection frame of power consumption part, the problem of the defect detection error that unmanned aerial vehicle's flight state change brought has been solved, the defect detection efficiency has been improved. Meanwhile, the geographical position and the shooting time corresponding to the damaged component are obtained, and the early warning information is generated according to the type mark, the geographical position and the shooting time of the damaged component, so that the early warning effect is improved.
Description
Technical Field
The application relates to the technical field of power inspection, in particular to a power transmission line target defect detection warning method and system.
Background
The unmanned aerial vehicle carries out the transmission line and patrols and examines the in-process, can carry out the target detection to the power consumption part in the transmission line to judge whether the power consumption part damages.
However, because unmanned aerial vehicle's developments flight, its target detection is a dynamic process too, and the distance between unmanned aerial vehicle and the power consumption part changes along with unmanned aerial vehicle's flight state to lead to unmanned aerial vehicle to clap the size of the detection target also different, this defect detection brings the error just, has also reduced transmission line's defect detection efficiency. Meanwhile, after the unmanned aerial vehicle finds the part defects, the early warning effect on the corresponding defective parts is poor, and therefore operation and maintenance efficiency is reduced.
Disclosure of Invention
The application provides a power transmission line target defect detection warning method and system, which are used for solving the technical problems of low defect detection error and defect detection efficiency caused by flight state change of an unmanned aerial vehicle and poor early warning effect.
In view of this, the first aspect of the present application provides a method for warning detection of target defects of a power transmission line, including the following steps:
acquiring a patrol inspection image of the power transmission line through an unmanned aerial vehicle, and decomposing the patrol inspection image into a plurality of patrol inspection images according to frames so as to establish a patrol inspection image set;
identifying the position and the type of the power utilization component in each routing inspection image in the routing inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
judging whether the contact ratio of the electric component and the target template in each routing inspection image is greater than a preset threshold value or not through a pre-trained target defect identification model, if so, judging that the electric component is a damaged component, and adding a damage mark to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
acquiring a geographical position and shooting time corresponding to the damaged part added with the damage mark;
and generating early warning information according to the type mark, the geographic position and the shooting time of the damaged part, and sending the early warning information to a specified terminal.
Preferably, the identifying the position and the type of the power consumption component in each patrol inspection image in the patrol inspection image set through a pre-trained target identification model specifically includes, before the step of adding the corresponding type mark and the detection frame to the corresponding position of the identified power consumption component:
acquiring a plurality of image samples containing power utilization components, and marking the position and the type of the power utilization components in each image sample;
carrying out augmentation operation processing on the marked image sample to obtain an augmented image sample set;
and training the augmented image sample set by a deep convolution neural network based on a YOLOv3 model so as to train and obtain a target recognition model.
Preferably, the convolutional neural network comprises a plurality of convolutional modules, short links are set between output ends of adjacent convolutional modules, the short links are convolutional layers, and features output by the short links and features output by the adjacent convolutional modules are subjected to feature aggregation through an adder.
Preferably, the convolution module includes an initial module and a convergence module, a jump connection channel is provided between an output end of the initial module and an output end of the convergence module, and features output by the jump connection channel and features output by the convergence module are subjected to feature convergence by the adder;
the jump connecting channel is a convolution layer, the step length of the jump connecting channel is equal to the step length of the short link, and the step length of the short link is equal to the step length of the convergence module.
Preferably, the step of acquiring the geographical position and the shooting time corresponding to the damaged part added with the damage mark comprises the following steps:
acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with adjacent towers;
sequentially numbering the towers and the transmission lines connected with the adjacent towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
the step of acquiring the geographical position and the shooting time corresponding to the damaged part added with the damage mark specifically includes:
acquiring geographical coordinate information of the tower or the power transmission line closest to the damaged part through an unmanned aerial vehicle;
based on the mapping relation between the serial numbers and the geographic coordinate information, matching according to the geographic coordinate information of the tower or the power transmission line which is closest to the damaged part to obtain the serial number of the tower or the power transmission line which is closest to the damaged part;
and acquiring the shooting time of the inspection image based on a clock system of a camera carried by the unmanned aerial vehicle, and marking the serial number and the shooting time into the corresponding inspection image.
In a second aspect, the present invention provides a power transmission line target defect detection warning system, including:
the system comprises a first image acquisition module, a second image acquisition module and a control module, wherein the first image acquisition module is used for acquiring a patrol inspection image of the power transmission line through an unmanned aerial vehicle and decomposing the patrol inspection image into a plurality of patrol inspection images according to frames so as to establish a patrol inspection image set;
the target identification module is used for identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
the defect identification module is used for judging whether the contact ratio of the electric component and the target template in each routing inspection image is greater than a preset threshold value or not through a pre-trained target defect identification model, if the contact ratio is smaller than the preset threshold value, the electric component is judged to be a damaged component, and a damage mark is added to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
the information acquisition module is used for acquiring the geographic position and the shooting time corresponding to the damaged part added with the damage mark;
and the early warning module is used for generating early warning information according to the type mark, the geographic position and the shooting time of the damaged part and sending the early warning information to a specified terminal.
Preferably, the system further comprises:
the marking module is used for acquiring a plurality of image samples containing the power utilization components and marking the positions and types of the power utilization components in each image sample;
the augmentation processing module is used for carrying out augmentation operation processing on the marked image samples so as to obtain an augmentation image sample set;
and the target recognition training module is used for training the augmented image sample set based on a deep convolutional neural network of a YOLOv3 model so as to obtain a target recognition model through training.
Preferably, the system further comprises:
the first geographic coordinate acquisition module is used for acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with the adjacent towers;
the mapping module is used for sequentially numbering the towers and the transmission lines connected with the adjacent towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
the information acquisition module specifically includes:
the second geographic coordinate acquisition module is used for acquiring the geographic coordinate information of the tower or the power transmission line closest to the damaged part through the unmanned aerial vehicle;
the matching module is used for matching according to the geographical coordinate information of the tower or the power transmission line closest to the damaged part to obtain the number of the tower or the power transmission line closest to the damaged part based on the mapping relation between the number and the geographical coordinate information;
and the time marking module is used for acquiring the shooting time of the inspection image based on a clock system of a camera carried by the unmanned aerial vehicle, and marking the serial number and the shooting time into the corresponding inspection image.
According to the technical scheme, the invention has the following advantages:
according to the method, the position and the type of the power utilization component in the inspection image are identified through the target identification model, the corresponding type mark and the detection frame are added, and the damage condition of the power utilization component is judged according to the contact ratio of the power utilization component and the target template through the target defect identification model, wherein the preset default frame of the target template is consistent with the scaling ratio of the detection frame of the power utilization component, so that the problem of defect detection errors caused by the change of the flight state of the unmanned aerial vehicle is solved, and the defect detection efficiency is improved. Meanwhile, the geographical position and the shooting time corresponding to the damaged component are obtained, and the early warning information is generated according to the type mark, the geographical position and the shooting time of the damaged component, so that the early warning effect is improved.
Drawings
Fig. 1 is a flowchart of a method for warning detection of a target defect of a power transmission line according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a power transmission line target defect detection warning system according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
For easy understanding, please refer to fig. 1, the method for warning detection of target defects of a power transmission line provided by the present invention includes the following steps:
s1, acquiring the inspection image of the power transmission line through the unmanned aerial vehicle, and decomposing the inspection image into a plurality of inspection images according to frames so as to establish an inspection image set;
s2, identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
s3, judging whether the coincidence degree of the electric component and the target template in each routing inspection image is larger than a preset threshold value or not through a pre-trained target defect recognition model, if the coincidence degree is smaller than the preset threshold value, judging that the electric component is a damaged component, and adding a damage mark to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
s4, acquiring the geographical position and the shooting time corresponding to the damaged part added with the damaged mark;
and S5, generating early warning information according to the type mark, the geographical position and the shooting time of the damaged part, and sending the early warning information to a specified terminal.
It should be noted that according to the power transmission line target defect detection warning method provided by the invention, the position and the type of the power utilization component in the inspection image are identified through the target identification model, the corresponding type mark and the detection frame are added, and the damage condition of the power utilization component is judged according to the contact ratio of the power utilization component and the target template through the target defect identification model, wherein the preset default frame of the target template is consistent with the scaling ratio of the detection frame of the power utilization component, so that the problem of defect detection errors caused by the change of the flight state of the unmanned aerial vehicle is solved, and the defect detection efficiency is improved. Meanwhile, the geographical position and the shooting time corresponding to the damaged component are obtained, and the early warning information is generated according to the type mark, the geographical position and the shooting time of the damaged component, so that the early warning effect is improved.
The following is a detailed description of an embodiment of the method for detecting and warning the target defect of the power transmission line provided by the invention.
The invention provides a method for detecting and warning the target defects of a power transmission line, which comprises the following steps:
s100, acquiring a patrol inspection image of the power transmission line through an unmanned aerial vehicle, and decomposing the patrol inspection image into a plurality of patrol inspection images according to frames so as to establish a patrol inspection image set;
set up camera equipment on unmanned aerial vehicle, can set up unmanned aerial vehicle flight route, unmanned aerial vehicle records the video along flight route is automatic, obtains video image, in unmanned aerial vehicle's treater, unframes video image, acquires the picture image. For example, the duration of the recorded video image is 300 seconds, and the frame rate of the video is 10 frames/second, so that the video image of 300 seconds can be deframed into 3000 picture images.
S200, identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
the detection frame is a minimum-sized detection frame that can include the electric component.
Before step S200, the method includes:
s201, obtaining a plurality of image samples containing power utilization components, and marking the position and the type of the power utilization components in each image sample;
the electric components can be poles, wires, insulators, vibration dampers and the like.
S202, carrying out augmentation operation processing on the marked image samples to obtain an augmented image sample set;
s203, training the augmented image sample set by a deep convolution neural network based on a YOLOv3 model, and thus training to obtain a target recognition model.
S300, judging whether the contact ratio of the electric component and the target template in each routing inspection image is larger than a preset threshold value or not through a pre-trained target defect identification model, if so, judging that the electric component is a damaged component, and adding a damage mark to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
it should be noted that, in actual detection, because along with the flight of unmanned aerial vehicle, detection is a dynamic process, and the distance of unmanned aerial vehicle apart from the power transmission tower is different, and the size of shooing the detection target is also different, and neural network when detecting, along with picture image through the feature extraction of convolution module, can lead to the disappearance of detail characteristics such as shallow texture, corner certainly. The features in the picture image are less and less, so that the features of the detection target are easily lost when the picture image is subjected to feature extraction by the convolution module, and the loss is irreversible, which brings errors to detection.
In this embodiment, the convolutional neural network includes a plurality of convolutional modules, short links are set between output ends of adjacent convolutional modules, the short links are convolutional layers, and features output by the short links and features output by the adjacent convolutional modules are feature aggregated by an adder.
The detail features extracted by the convolution module at the front end of the neural network can be transmitted to the convolution module at the rear end of the neural network through short link, namely, the detail features obtained by the convolution module at the front end are gathered into the convolution module at the rear end, and the loss of the detail features in the features extracted by the neural network is compensated. Therefore, the feature expression capability in the image is stronger, and the detection of the detection target is facilitated.
The convolution module comprises an initial module and a convergence module, a jump connecting channel is arranged between the output end of the initial module and the output end of the convergence module, and features output by the jump connecting channel and features output by the convergence module are converged by an adder;
the number of the initial modules is 2, and the initial modules are marked as initial modules C1 and sequentially comprise a convolutional layer, an active layer, a convolutional layer and a pooling layer. The convergence module is provided with a plurality of. The convergence module comprises a convolution layer, an activation layer, a convolution layer and a pooling layer in sequence, wherein the convergence module comprises C2/C3/C4 and the like.
The jump connecting channel is a convolution layer, the step length of the jump connecting channel is equal to the step length of the short link, and the step length of the short link is equal to the step length of the convergence module.
The convolution layers all adopt 3-by-3 convolution kernels, and the extraction efficiency is high when feature extraction is carried out on the picture images. And performing down-sampling through the pooling layer, removing unimportant features in the picture image, reducing the number of the features, extracting the features of the picture image, and outputting the features of the picture image by the convergence module. The step length of the jump connection channel is equal to the step length of the short link, the step length of the short link is equal to the step length of the convergence module, the feature graph output by the convergence module and the feature graph output by the short link can be kept consistent in dimension and resolution, and convergence of input and output features of the convolution module is achieved.
Meanwhile, the bottom layer features in the initial module can be converged into the deep layer features in the convergence module through the jumping connection channel, and the bottom layer features are allowed to be converged into the deep layer features directly. The expression capability of deep features is stronger, detection errors caused by information loss are relieved, and the detection of a small target such as a detection target is facilitated.
And the bottom layer features are converged into the deep layer features through short links and jump connecting channels, so that better feature representation is achieved, and the performance of the neural network for detecting the detection target is improved.
Wherein, the output result after the short chain-connection convolution operation is as follows:
wherein:
to represent
The output result after the convolution operation of the short link,
represents the output characteristics of the i-1 th convolution module, and ω sc is the short-chained convolution operation.
The output result after the jump connection channel convolution operation is:
wherein:
to represent
The output result after the convolution operation of the hopping connection channel,
the output feature vector of the 1 st convolution module is shown, and ω sk is the convolution operation of the jump connection channel.
Output characteristics of convolution Module i
Comprises the following steps:
wherein,
representing the output characteristics of the ith convolution module, and P representing the pooling operation in the convolution module or the convolution operation in the additional module; f is a function of the convolution layer in the convolution module,
representing the input characteristics of the ith convolution module; e t represents the output of the convolution layer in the convolution module.
The features output by the convolution module and the features output by the short link are input into an adder, the adder converges the features output by the convolution module and the features output by the short link and outputs the converged features, and the output of the adder is the input of the convolution module immediately behind the adder.
The adder includes an overlay layer and an active layer.
The confidence coefficient can determine that when the convolution module i is connected with a short link in a detection frame in the data image, the output of the adder is the input of the convolution module immediately behind the adder
Comprises the following steps:
when the convolution module i is connected to the short link and jump link channels, the output of the adder is the input of the convolution module
Comprises the following steps:
wherein,
representing the output characteristics of the i-1 th convolution module,
to represent
The output result after the convolution operation of the short link,
representing the output characteristics of the i-2 th convolution module.
To represent
And outputting the result after the convolution operation of the jump connection channel. i is more than or equal to 2, when i is 2,
is equivalent to
Only need to exist at this time
In (1)
One is needed.
In the neural network, default frames are set, and in the convergence module, default frames with different sizes can be set according to the size of the detection target.
The width and height of the default box are noted as:
wherein, wmThe width of the default frame, hm is the height of the default frame, and m is the serial number of the convolution module for obtaining the default frame. I.e. each convolution module can set a separate default box.
Aspect ratio a of the default framerIs recorded as:
ar∈{1，2，3，1/2，1/3}；
when the aspect ratio of the default frame is 1, wm＝hm。
In order to adapt to the situation that detection targets with the same size have different sizes when different distances exist according to the distance between the unmanned aerial vehicle and the power transmission tower, the offset is obtained by the neural network to correct the default frame, and the detection frame is obtained.
Obtaining a detection frame marking the detection target according to the default frame, wherein the size of the detection frame is represented as:
wherein,
a coordinate point of x, y representing a center point of the detection frame;
indicating the width of the detection frame;
indicating the height of the detection box. dcx，dcyA coordinate point of x, y representing the center point of the default box; dwRepresenting the width of the default box; dhIndicating the height of the default box. variance denotes scaling, v0，v1，v2，v3Representing the preset scaling parameters in the variance. lcx，lcyIndicates the offset of the default box center point in x, y, lwOffset, l, representing default box widthhAn offset representing the default box height.
After the category in the detection frame is determined, the detection target in the detection frame is compared with the target templates of the same category, for example, the category in the detection frame is an insulator, and the insulator in the target template is a size of a vertical shot at a distance of 5 meters. After the insulators in the detection frame are zoomed by a certain multiple, the sizes of the detection frame of the insulators in the detection target are consistent with the size of the default frame of the target template, so that the sizes of the insulators in the detection frame and the insulators in the target template are the same, the contact ratio of the insulators in the detection frame and the insulators in the target template is compared, the threshold value of the contact ratio is set to be 90%, if the contact ratio of the insulators in the detection frame and the insulators in the target template is compared to be more than 90%, the insulators in the detection target are normal and not damaged, and if the contact ratio is less than 90%, the insulators are damaged.
S400, acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with adjacent towers;
it should be noted that the unmanned aerial vehicle can carry a GPS positioning system to acquire longitude and latitude coordinate information of all towers in the target inspection area and the power transmission lines connected to the adjacent towers.
S500, sequentially numbering the pole towers and the power transmission lines connected with the adjacent pole towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
it should be noted that, in this embodiment, the transmission lines connected to the tower and the adjacent towers thereof are named by numbers, and the numbering mode may be tower 001, transmission line 001 and tower 002.
S600, acquiring the geographical position and the shooting time corresponding to the damaged part added with the damaged mark;
specifically, step S600 includes:
s601, acquiring geographical coordinate information of a tower or a power transmission line closest to the damaged part through an unmanned aerial vehicle;
s602, based on the mapping relation between the numbers and the geographic coordinate information, matching the geographic coordinate information of the tower or the power transmission line closest to the damaged component to obtain the number of the tower or the power transmission line closest to the damaged component;
it should be noted that after the serial numbers of the towers or the transmission lines are obtained through matching, the corresponding geographical position information can be quickly located.
S603, acquiring the shooting time of the inspection image based on a clock system of a camera carried by the unmanned aerial vehicle, and marking the serial number and the shooting time into the corresponding inspection image.
It should be noted that, the serial numbers and the shooting time are marked to the corresponding patrol images, so that the detecting personnel can quickly determine the geographical position and the damage time of the damaged detection target, and the method is more concise.
S700, generating early warning information according to the type mark, the geographic position and the shooting time of the damaged part, and sending the early warning information to a specified terminal.
It should be noted that the type mark, the geographic position and the shooting time of the damaged component are used for generating early warning information and sending the early warning information to a designated terminal, such as a mobile phone app, a webpage and the like, so that operation and maintenance personnel can quickly respond, and effective information can be obtained, so that the operation and maintenance efficiency is improved.
The foregoing is a detailed description of an embodiment of a method for detecting and warning a target defect of a power transmission line provided by the present invention, and the following is a detailed description of an embodiment of a system for detecting and warning a target defect of a power transmission line provided by the present invention.
For easy understanding, please refer to fig. 2, the present invention provides a power transmission line target defect detection warning system, including:
the first image acquisition module 100 is configured to acquire a patrol inspection image of the power transmission line through the unmanned aerial vehicle, and decompose the patrol inspection image into a plurality of patrol inspection images according to frames, so as to establish a patrol inspection image set;
the target identification module 200 is used for identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
the defect identification module 300 is used for judging whether the coincidence degree of the electric component and the target template in each routing inspection image is greater than a preset threshold value through a pre-trained target defect identification model, if the coincidence degree is smaller than the preset threshold value, judging the electric component to be a damaged component, and adding a damage mark to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
an information acquisition module 400, configured to acquire a geographical location and a shooting time corresponding to the damaged component to which the damage flag is added;
and the early warning module 500 is used for generating early warning information according to the type mark, the geographical position and the shooting time of the damaged part and sending the early warning information to a specified terminal.
Further, the system also includes:
the marking module is used for acquiring a plurality of image samples containing the power utilization components and marking the positions and types of the power utilization components in each image sample;
the augmentation processing module is used for carrying out augmentation operation processing on the marked image samples so as to obtain an augmentation image sample set;
and the target recognition training module is used for training the augmented image sample set based on a deep convolutional neural network of the YOLOv3 model so as to obtain a target recognition model through training.
Further, the system also includes:
the first geographic coordinate acquisition module is used for acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with adjacent towers;
the mapping module is used for sequentially numbering the pole towers and the power transmission lines connected with the adjacent pole towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
the information acquisition module specifically comprises:
the second geographic coordinate acquisition module is used for acquiring the geographic coordinate information of the nearest tower or power transmission line close to the damaged part through the unmanned aerial vehicle;
the matching module is used for matching and obtaining the number of the tower or the transmission line closest to the damaged component according to the geographical coordinate information of the tower or the transmission line closest to the damaged component based on the mapping relation between the number and the geographical coordinate information;
and the time marking module is used for acquiring the shooting time of the inspection image based on a clock system of the camera carried by the unmanned aerial vehicle and marking the shooting time to the corresponding inspection image.
It should be noted that the working process of the power transmission line target defect detection warning system provided by the present invention is consistent with the flow of the power transmission line target defect detection warning method provided by the above embodiment, and is not described herein again.
According to the power transmission line target defect detection warning system, the position and the type of the power utilization component in the inspection image are identified through the target identification model, the corresponding type mark and the detection frame are added, and the damage condition of the power utilization component is judged through the target defect identification model according to the contact ratio of the power utilization component and the target template, wherein the preset default frame of the target template is consistent with the scaling ratio of the detection frame of the power utilization component, so that the problem of defect detection errors caused by the change of the flight state of an unmanned aerial vehicle is solved, and the defect detection efficiency is improved. Meanwhile, the geographical position and the shooting time corresponding to the damaged component are obtained, and the early warning information is generated according to the type mark, the geographical position and the shooting time of the damaged component, so that the early warning effect is improved.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.
Claims (8)
1. A method for detecting and warning target defects of a power transmission line is characterized by comprising the following steps:
acquiring a patrol inspection image of the power transmission line through an unmanned aerial vehicle, and decomposing the patrol inspection image into a plurality of patrol inspection images according to frames so as to establish a patrol inspection image set;
identifying the position and the type of the power utilization component in each routing inspection image in the routing inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
judging whether the contact ratio of the electric component and the target template in each routing inspection image is greater than a preset threshold value or not through a pre-trained target defect identification model, if so, judging that the electric component is a damaged component, and adding a damage mark to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
acquiring a geographical position and shooting time corresponding to the damaged part added with the damage mark;
and generating early warning information according to the type mark, the geographic position and the shooting time of the damaged part, and sending the early warning information to a specified terminal.
2. The method for detecting and warning the target defect of the power transmission line according to claim 1, wherein the step of identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model and adding a corresponding type mark and a corresponding detection frame to the corresponding position of the identified power utilization component specifically comprises the following steps:
acquiring a plurality of image samples containing power utilization components, and marking the position and the type of the power utilization components in each image sample;
carrying out augmentation operation processing on the marked image sample to obtain an augmented image sample set;
and training the augmented image sample set by a deep convolution neural network based on a YOLOv3 model so as to train and obtain a target recognition model.
3. The method according to claim 1, wherein the convolutional neural network comprises a plurality of convolutional modules, short links are set between output ends of adjacent convolutional modules, the short links are convolutional layers, and features output by the short links and features output by the adjacent convolutional modules are subjected to feature aggregation through an adder.
4. The method for warning detection of the target defect of the power transmission line according to claim 3, wherein the convolution module comprises an initial module and a convergence module, a jump connection channel is arranged between the output end of the initial module and the output end of the convergence module, and the features output by the jump connection channel and the features output by the convergence module are subjected to feature convergence by the adder;
the jump connecting channel is a convolution layer, the step length of the jump connecting channel is equal to the step length of the short link, and the step length of the short link is equal to the step length of the convergence module.
5. The method for detecting and warning the target defect of the power transmission line according to claim 1, wherein the step of obtaining the geographic position and the shooting time corresponding to the damaged part added with the damaged mark comprises the following steps:
acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with adjacent towers;
sequentially numbering the towers and the transmission lines connected with the adjacent towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
the step of acquiring the geographical position and the shooting time corresponding to the damaged part added with the damage mark specifically includes:
acquiring geographical coordinate information of the tower or the power transmission line closest to the damaged part through an unmanned aerial vehicle;
based on the mapping relation between the serial numbers and the geographic coordinate information, matching according to the geographic coordinate information of the tower or the power transmission line which is closest to the damaged part to obtain the serial number of the tower or the power transmission line which is closest to the damaged part;
and acquiring the shooting time of the inspection image based on a clock system of a camera carried by the unmanned aerial vehicle, and marking the serial number and the shooting time into the corresponding inspection image.
6. The utility model provides a transmission line target defect detects warning system which characterized in that includes:
the system comprises a first image acquisition module, a second image acquisition module and a control module, wherein the first image acquisition module is used for acquiring a patrol inspection image of the power transmission line through an unmanned aerial vehicle and decomposing the patrol inspection image into a plurality of patrol inspection images according to frames so as to establish a patrol inspection image set;
the target identification module is used for identifying the position and the type of the power utilization component in each inspection image in the inspection image set through a pre-trained target identification model, and adding a corresponding type mark and a corresponding detection frame at the corresponding position of the identified power utilization component;
the defect identification module is used for judging whether the contact ratio of the electric component and the target template in each routing inspection image is greater than a preset threshold value or not through a pre-trained target defect identification model, if the contact ratio is smaller than the preset threshold value, the electric component is judged to be a damaged component, and a damage mark is added to the damaged component; the preset default frame of the target template is consistent with the scaling of the detection frame of the power utilization component;
the information acquisition module is used for acquiring the geographic position and the shooting time corresponding to the damaged part added with the damage mark;
and the early warning module is used for generating early warning information according to the type mark, the geographic position and the shooting time of the damaged part and sending the early warning information to a specified terminal.
7. The system for detecting and warning the target defect of the power transmission line according to claim 6, further comprising:
the marking module is used for acquiring a plurality of image samples containing the power utilization components and marking the positions and types of the power utilization components in each image sample;
the augmentation processing module is used for carrying out augmentation operation processing on the marked image samples so as to obtain an augmentation image sample set;
and the target recognition training module is used for training the augmented image sample set based on a deep convolutional neural network of a YOLOv3 model so as to obtain a target recognition model through training.
8. The system for detecting and warning the target defect of the power transmission line according to claim 6, further comprising:
the first geographic coordinate acquisition module is used for acquiring geographic coordinate information of all towers in the target inspection area and power transmission lines connected with the adjacent towers;
the mapping module is used for sequentially numbering the towers and the transmission lines connected with the adjacent towers so as to establish a mapping relation between the numbers and the geographic coordinate information;
the information acquisition module specifically includes:
the second geographic coordinate acquisition module is used for acquiring the geographic coordinate information of the tower or the power transmission line closest to the damaged part through the unmanned aerial vehicle;
the matching module is used for matching according to the geographical coordinate information of the tower or the power transmission line closest to the damaged part to obtain the number of the tower or the power transmission line closest to the damaged part based on the mapping relation between the number and the geographical coordinate information;
and the time marking module is used for acquiring the shooting time of the inspection image based on a clock system of a camera carried by the unmanned aerial vehicle, and marking the serial number and the shooting time into the corresponding inspection image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110949156.4A CN113657280B (en) | 2021-08-18 | 2021-08-18 | Power transmission line target defect detection warning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110949156.4A CN113657280B (en) | 2021-08-18 | 2021-08-18 | Power transmission line target defect detection warning method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113657280A true CN113657280A (en) | 2021-11-16 |
| CN113657280B CN113657280B (en) | 2024-06-21 |
Family
ID=78480957
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110949156.4A Active CN113657280B (en) | 2021-08-18 | 2021-08-18 | Power transmission line target defect detection warning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113657280B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115187880A (en) * | 2022-07-20 | 2022-10-14 | 无锡科技职业学院 | Communication optical cable defect detection method and system based on image recognition and storage medium |
| CN116027798A (en) * | 2022-09-30 | 2023-04-28 | 三峡大学 | Unmanned aerial vehicle power inspection system and method based on image correction |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015131462A1 (en) * | 2014-03-07 | 2015-09-11 | 国家电网公司 | Centralized monitoring system and monitoring method for unmanned aerial vehicle to patrol power transmission line |
| CN109977943A (en) * | 2019-02-14 | 2019-07-05 | 平安科技（深圳）有限公司 | A kind of images steganalysis method, system and storage medium based on YOLO |
| CN110378221A (en) * | 2019-06-14 | 2019-10-25 | 安徽南瑞继远电网技术有限公司 | A kind of power grid wire clamp detects and defect identification method and device automatically |
| CN110555513A (en) * | 2019-08-16 | 2019-12-10 | 国电南瑞科技股份有限公司 | Deep learning-based power equipment defect integrated diagnosis method |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
| CN111311569A (en) * | 2020-02-12 | 2020-06-19 | 江苏方天电力技术有限公司 | Pole tower defect identification method based on unmanned aerial vehicle inspection |
| CN111353413A (en) * | 2020-02-25 | 2020-06-30 | 武汉大学 | Low-missing-report-rate defect identification method for power transmission equipment |
-
2021
- 2021-08-18 CN CN202110949156.4A patent/CN113657280B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015131462A1 (en) * | 2014-03-07 | 2015-09-11 | 国家电网公司 | Centralized monitoring system and monitoring method for unmanned aerial vehicle to patrol power transmission line |
| CN109977943A (en) * | 2019-02-14 | 2019-07-05 | 平安科技（深圳）有限公司 | A kind of images steganalysis method, system and storage medium based on YOLO |
| CN110378221A (en) * | 2019-06-14 | 2019-10-25 | 安徽南瑞继远电网技术有限公司 | A kind of power grid wire clamp detects and defect identification method and device automatically |
| CN110555513A (en) * | 2019-08-16 | 2019-12-10 | 国电南瑞科技股份有限公司 | Deep learning-based power equipment defect integrated diagnosis method |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
| CN111311569A (en) * | 2020-02-12 | 2020-06-19 | 江苏方天电力技术有限公司 | Pole tower defect identification method based on unmanned aerial vehicle inspection |
| CN111353413A (en) * | 2020-02-25 | 2020-06-30 | 武汉大学 | Low-missing-report-rate defect identification method for power transmission equipment |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115187880A (en) * | 2022-07-20 | 2022-10-14 | 无锡科技职业学院 | Communication optical cable defect detection method and system based on image recognition and storage medium |
| CN116027798A (en) * | 2022-09-30 | 2023-04-28 | 三峡大学 | Unmanned aerial vehicle power inspection system and method based on image correction |
| CN116027798B (en) * | 2022-09-30 | 2023-11-17 | 三峡大学 | Unmanned aerial vehicle power inspection system and method based on image correction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113657280B (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113657280B (en) | Power transmission line target defect detection warning method and system | |
| CN112232279B (en) | Personnel interval detection method and device | |
| CN112767391A (en) | Power grid line part defect positioning method fusing three-dimensional point cloud and two-dimensional image | |
| CN112508865B (en) | Unmanned aerial vehicle inspection obstacle avoidance method, unmanned aerial vehicle inspection obstacle avoidance device, computer equipment and storage medium | |
| CN110892760B (en) | Positioning terminal equipment based on deep learning | |
| CN111986172A (en) | Infrared image fault detection method and device for power equipment | |
| CN113192646B (en) | Target detection model construction method and device for monitoring distance between different targets | |
| US9208555B1 (en) | Method for inspection of electrical equipment | |
| CN112270755B (en) | Three-dimensional scene construction method and device, storage medium and electronic equipment | |
| CN115359239A (en) | Wind power blade defect detection and positioning method and device, storage medium and electronic equipment | |
| CN110110684A (en) | Foreign object identification method and device for power transmission line equipment and computer equipment | |
| CN112132070A (en) | Driving behavior analysis method, device, equipment and storage medium | |
| CN110827314B (en) | Single-target tracking method and related equipment | |
| CN114880730A (en) | Method and device for determining target equipment and photovoltaic system | |
| CN117765420B (en) | Terrain surveying method and system based on remote sensing data | |
| CN114119528A (en) | Defect detection method and device for distribution line insulator | |
| CN112260402B (en) | Monitoring method for state of intelligent substation inspection robot based on video monitoring | |
| CN117830216A (en) | Lightning damage defect detection method and device for insulated terminal of power transmission line | |
| CN114677859B (en) | Unmanned aerial vehicle route automatic correction method and device | |
| CN114592411B (en) | Carrier parasitic type intelligent inspection method for highway damage | |
| CN113537194B (en) | Illumination estimation method, illumination estimation device, storage medium, and electronic apparatus | |
| CN113689485B (en) | Method and device for determining depth information of unmanned aerial vehicle, unmanned aerial vehicle and storage medium | |
| CN113870361A (en) | Calibration method, device and equipment of depth camera and storage medium | |
| CN113283285A (en) | Method for accurately positioning address based on image recognition technology | |
| CN114093051A (en) | Communication line inspection method, device and system, and computer-readable storage medium |
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 |