CN117635589A - External damage detection method and system based on inspection data of monocular distribution network machine - Google Patents
External damage detection method and system based on inspection data of monocular distribution network machine Download PDFInfo
- Publication number
- CN117635589A CN117635589A CN202311678629.7A CN202311678629A CN117635589A CN 117635589 A CN117635589 A CN 117635589A CN 202311678629 A CN202311678629 A CN 202311678629A CN 117635589 A CN117635589 A CN 117635589A
- Authority
- CN
- China
- Prior art keywords
- image
- inspection
- data
- model
- monocular
- 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
Links
- 238000007689 inspection Methods 0.000 title claims abstract description 51
- 230000006378 damage Effects 0.000 title claims abstract description 25
- 238000001514 detection method Methods 0.000 title claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000007781 pre-processing Methods 0.000 claims abstract description 19
- 238000000034 method Methods 0.000 claims abstract description 17
- 238000004590 computer program Methods 0.000 claims description 10
- 238000004140 cleaning Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 5
- 230000007547 defect Effects 0.000 abstract description 4
- 238000005516 engineering process Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 102220559998 Kinesin-like protein KIF1A_M30T_mutation Human genes 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an external damage detection method and system based on inspection data of a monocular distribution network machine, and belongs to the technical field of power line inspection. The method comprises the steps of acquiring images along a power line by using unmanned aerial vehicle equipment to obtain an image dataset of the power line and surrounding environment thereof; preprocessing an image dataset; selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting the set condition; and detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden dangers of the power line and surrounding environment thereof. The invention can quickly identify and search defects in a large number of machine inspection pictures and videos through the identification model, effectively and accurately search hidden line trouble, and provide a certain help for the normal operation of the power system.
Description
Technical Field
The invention belongs to the technical field of power line inspection, and particularly relates to an external damage detection method and system based on inspection data of a monocular distribution network machine.
Background
As the complexity and coverage of electrical facilities expands, maintenance and monitoring thereof becomes increasingly critical. The traditional manual inspection method is low in efficiency and can have the risk of missing inspection. In recent years, unmanned plane technology and AI technology are combined, and a more efficient and more accurate inspection method is provided for electric power facilities. However, how to find out the hidden trouble of the line in time in massive machine inspection data is a problem which needs to be solved in the industry.
The line external force damage detection task at the present stage is mainly completed through methods such as manual inspection, video monitoring and sensor detection, but the problems that the traditional inspection efficiency is low, the video real-time monitoring input cost is high, the accuracy of sensor detection cannot be guaranteed and the like are mainly solved. The patrol data still needs to be checked manually and frame by frame, and has low efficiency and long time consumption. How to utilize good mass machine inspection data, release human resources through artificial intelligence technology's application, discover line external force destruction, potential safety hazards such as tree barrier fast, reduce the harm to power line, have important meaning to improving power line's safety and reliability.
Disclosure of Invention
In view of the above, the present invention aims to solve the above-mentioned problems of the existing line external damage detection method.
In order to solve the technical problems, the invention provides the following technical scheme:
in a first aspect, the invention provides an external damage detection method based on inspection data of a monocular distribution network machine, which comprises the following steps:
image acquisition is carried out along the power line by using unmanned aerial vehicle equipment, so that an image data set of the power line and surrounding environment thereof is obtained;
preprocessing an image dataset;
selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting the set condition;
and detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden dangers of the power line and surrounding environment thereof.
Further, preprocessing the image dataset specifically includes:
performing data cleaning on the image dataset to obtain an image meeting the set requirement;
marking the hidden danger positions of the images after data cleaning by using a marking tool, and giving corresponding labels;
and performing sample expansion processing on the marked image to finally obtain a preprocessed image dataset.
Further, in model training, the performance of the model is evaluated based on accuracy, recall, and F1 score.
Further, the method for detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle device by using the identification model specifically comprises the following steps:
acquiring images or video data obtained by patrol of unmanned aerial vehicle equipment;
if the file name is the image data, traversing all file names, judging whether the file names contain the extension names of the image files, if so, calling a recognition model to recognize the image files, outputting an image recognition result and labeling the recognition result.
In a second aspect, the present invention provides an external damage detection system based on inspection data of a monocular distribution network machine, including:
the image acquisition unit is used for storing an image data set of the power line and the surrounding environment obtained by carrying out image acquisition along the power line by using unmanned aerial vehicle equipment;
a preprocessing unit for preprocessing an image dataset;
the model training unit is used for selecting a model version of the YOLO algorithm, and performing model training by utilizing the preprocessed image data set to obtain an identification model meeting the set condition;
and the identification unit is used for detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden danger of the power line and surrounding environment thereof.
Further, in the preprocessing unit, preprocessing the image dataset specifically includes:
performing data cleaning on the image dataset to obtain an image meeting the set requirement;
marking the hidden danger positions of the images after data cleaning by using a marking tool, and giving corresponding labels;
and performing sample expansion processing on the marked image to finally obtain a preprocessed image dataset.
Further, in the model training unit, the performance of the model is evaluated based on the accuracy, recall, and F1 score.
Further, in the identifying unit, the identifying model is used for detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle device, and specifically includes:
acquiring images or video data obtained by patrol of unmanned aerial vehicle equipment;
if the file name is the image data, traversing all file names, judging whether the file names contain the extension names of the image files, if so, calling a recognition model to recognize the image files, outputting an image recognition result and labeling the recognition result.
In a third aspect, the invention provides an external damage detection device based on inspection data of a monocular distribution network machine, the device comprising a processor and a memory:
the memory is used for storing the computer program and sending the instructions of the computer program to the processor;
the processor executes the external damage detection method based on the inspection data of the monocular distribution network machine according to the instruction of the computer program.
In a fourth aspect, the present invention provides a computer storage medium, where a computer program is stored, where the computer program when executed by a processor implements a method for detecting a break in data based on a monocular distribution network machine according to the first aspect.
In summary, the invention provides an outward breaking detection method and system based on inspection data of a monocular distribution network machine, wherein the method acquires an image data set of a power line and surrounding environment thereof by utilizing unmanned aerial vehicle equipment to acquire the image along the power line; preprocessing an image dataset; selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting the set condition; and detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden dangers of the power line and surrounding environment thereof. The invention can quickly identify and search defects in a large number of machine inspection pictures and videos through the identification model, effectively and accurately search hidden line trouble, and provide a certain help for the normal operation of the power system.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of an external damage detection method based on inspection data of a monocular distribution network machine according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of image annotation according to an embodiment of the present invention;
FIG. 3 is a schematic view of image expansion according to an embodiment of the present invention;
fig. 4 is a flowchart of recognition of a recognition model according to an embodiment of the present invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is apparent that the embodiments described below are only some embodiments of the present invention, not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the embodiment provides an external damage detection method based on inspection data of a monocular distribution network machine, which includes the following steps:
step one: sample acquisition, namely performing image acquisition along a power line by using unmanned aerial vehicle equipment to obtain an image dataset of the power line and surrounding environment thereof;
step two: data preprocessing, preprocessing an image data set.
It can be understood that the image data are collected and marked, images with potential safety hazards are screened by acquiring inspection images or video data of the power distribution network of the unmanned aerial vehicle, the potential hazards in the images are marked, marking information is stored according to a data set format, and finally a power distribution network potential hazard data set of the image data and the marking file is formed.
Step three: model training, namely selecting a model version of the YOLO algorithm, and performing model training by utilizing the preprocessed image dataset to obtain an identification model meeting set conditions.
The method is characterized in that scaling, normalization, and the like are performed on the image data in the hidden danger data set, and data expansion is performed by adopting data enhancement, so that the richness of the hidden danger data is improved. And constructing a deep neural network, training the deep neural network by adopting the preprocessed data and the marking information and combining a gradient descent method, and learning the characteristics of hidden danger.
Step four: and (3) identifying, namely detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden danger of the power line and surrounding environment thereof.
The method comprises the steps of extracting hidden danger features in unmanned aerial vehicle image or video data by using a deep neural network, performing post-processing operation on the hidden danger features, acquiring hidden danger types and position information, and outputting positions of hidden troubles in patrol data; and finally, the identification of potential safety hazards of the power distribution network existing in the current unmanned aerial vehicle image is completed.
The embodiment provides an external damage detection method and system based on inspection data of a monocular distribution network machine, wherein the method acquires images along an electric power line by using unmanned aerial vehicle equipment to obtain an image dataset of the electric power line and surrounding environment thereof; preprocessing an image dataset; selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting the set condition; and detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden dangers of the power line and surrounding environment thereof. The invention can quickly identify and search defects in a large number of machine inspection pictures and videos through the identification model, effectively and accurately search hidden line trouble, and provide a certain help for the normal operation of the power system.
In one embodiment of the present invention, sample collection is to enable the YOLO algorithm to accurately identify hidden trouble near the power line, and a large number of related image samples, such as an excavator, an excessively high tree, etc., are first required. In the aspect of the type selection of the acquisition equipment, based on the consideration of image quality, flight stability and endurance, the flight path of the unmanned aerial vehicle is planned according to the power line layout, and the image acquisition is carried out by adopting unmanned aerial vehicle equipment such as M3E or M30T matched with a large airport. And image capturing is performed on the power line and the surrounding environment thereof under different time, place, weather and season conditions to capture potential hidden trouble.
In one embodiment of the invention, the data preprocessing is to perform data cleaning on massive image data acquired by the unmanned aerial vehicle, and screen out high-quality representative images. As shown in fig. 2, a labeling tool (e.g., labelImg, etc.) is used to locate the target (e.g., excavator) in each sample image and give the corresponding label. The annotation tool can provide the YOLO with the accurate, formatted data required for training. As shown in FIG. 3, the samples are added by image processing technology (such as rotation, clipping, brightness adjustment, noise addition, etc.), the data set is expanded, and the robustness of the model is improved.
In one embodiment of the invention, model training is to select an appropriate YOLO algorithm model version according to the type and number of hidden trouble to be identified and perform parameter configuration. The identification is generally performed using the newer YOLO V5 or YOLO V8 algorithms at the present stage, which have smaller and faster model structures. In the training process, in order to achieve a faster and better training effect, the GPU can be used for model training, model weights are optimized through repeated iteration, and indexes such as accuracy, recall rate, F1 score and the like are used for evaluating model performance until satisfactory accuracy is achieved.
In one embodiment of the invention, the recognition process is that after a model weight file output by YOLO is taken, inspection of machine inspection data can be performed through the file, and the method comprises analyzing image data or video files uploaded to a computer after inspection of an unmanned aerial vehicle through a YOLO algorithm, and detecting hidden danger in a picture. Besides identifying hidden dangers, YOLO can accurately mark the positions of hidden dangers in the image, and convenience is provided for subsequent processing work. For example, when the program detects hidden danger, the program can automatically output related hidden danger data to provide related information of the hidden danger. The flow of identifying hidden danger by the identification model is shown in fig. 4, wherein when identifying a picture or video, corresponding file data is traversed, and if a corresponding file extension (such as a picture file with a suffix of. ipg or.png and a video file with a suffix of.mp 4) exists, the identification model is called for identification.
The invention can utilize artificial intelligence technology, obtain a self-training model through sample collection, data preprocessing and model training, and then use the model to quickly identify and search defects in a large number of machine inspection pictures and videos, effectively and accurately search line hidden danger, and provide a certain help for the normal operation of an electric power system. Its advantages are mainly the following aspects:
1. and (5) flexible model training. Model training can retrain according to the demand, and multiple hidden danger can all be discerned at same model for hidden danger discernment has certain flexibility, can adapt to different demands.
2. The recognition efficiency is remarkably improved. The hidden danger is searched by comparing manual pictures in a video way, the one-key operation of the software is convenient, quick and efficient, the working efficiency is obviously improved, and the burden of manually checking the pictures is reduced.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The external damage detection method based on the inspection data of the monocular distribution network machine is characterized by comprising the following steps of:
image acquisition is carried out along the power line by using unmanned aerial vehicle equipment, so that an image data set of the power line and surrounding environment thereof is obtained;
preprocessing the image dataset;
selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting a set condition;
and detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden danger of the power line and surrounding environment thereof.
2. The method for detecting the external damage based on the inspection data of the monocular distribution network machine according to claim 1, wherein the preprocessing of the image dataset is specifically performed, and comprises the following steps:
performing data cleaning on the image data set to obtain an image meeting the set requirement;
marking the hidden danger positions of the images after data cleaning by using a marking tool, and giving corresponding labels;
and performing sample expansion processing on the marked image to finally obtain a preprocessed image dataset.
3. The method for detecting the external damage based on the inspection data of the monocular distribution network machine according to claim 1, wherein in model training, the performance of the model is evaluated according to the accuracy, the recall and the F1 score.
4. The method for detecting the external damage based on the inspection data of the monocular distribution network machine according to claim 1, wherein the image or video data obtained by inspection of the unmanned aerial vehicle equipment is detected and identified by using the identification model, and specifically comprises the following steps:
acquiring images or video data obtained by patrol of unmanned aerial vehicle equipment;
if the file name is the image data, traversing all file names, judging whether the file names contain the extension names of the image files, if so, calling the identification model to identify the image files, outputting an image identification result and labeling the identification result.
5. An external damage detection system based on inspection data of a monocular distribution network machine, which is characterized by comprising:
the image acquisition unit is used for storing an image data set of the power line and the surrounding environment obtained by carrying out image acquisition along the power line by using unmanned aerial vehicle equipment;
a preprocessing unit for preprocessing the image dataset;
the model training unit is used for selecting a model version of the YOLO algorithm, and performing model training by using the preprocessed image data set to obtain an identification model meeting the set condition;
and the identification unit is used for detecting and identifying the image or video data obtained by the inspection of the unmanned aerial vehicle equipment by using the identification model to obtain hidden danger of the power line and surrounding environment thereof.
6. The external damage detection system based on inspection data of a monocular distribution network machine according to claim 5, wherein the preprocessing unit performs preprocessing on the image dataset, and specifically includes:
performing data cleaning on the image data set to obtain an image meeting the set requirement;
marking the hidden danger positions of the images after data cleaning by using a marking tool, and giving corresponding labels;
and performing sample expansion processing on the marked image to finally obtain a preprocessed image dataset.
7. The system for detecting the external damage based on the inspection data of the monocular distribution network machine according to claim 5, wherein in the model training unit, the performance of the model is evaluated according to the accuracy, the recall and the F1 score.
8. The external damage detection system based on the inspection data of the monocular distribution network machine according to claim 5, wherein in the identification unit, the image or video data obtained by inspection of the unmanned aerial vehicle device is detected and identified by using the identification model, and specifically comprises:
acquiring images or video data obtained by patrol of unmanned aerial vehicle equipment;
if the file name is the image data, traversing all file names, judging whether the file names contain the extension names of the image files, if so, calling the identification model to identify the image files, outputting an image identification result and labeling the identification result.
9. The utility model provides a broken check out test set outward based on monocular distribution network machine patrols data which characterized in that, equipment includes treater and memory:
the memory is used for storing a computer program and sending instructions of the computer program to the processor;
the processor executes the external damage detection method based on the inspection data of the monocular distribution network machine according to the instructions of the computer program as set forth in any one of claims 1 to 4.
10. A computer storage medium, wherein a computer program is stored on the computer storage medium, and when the computer program is executed by a processor, the method for detecting external damage based on inspection data of a monocular distribution network machine is realized according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311678629.7A CN117635589A (en) | 2023-12-07 | 2023-12-07 | External damage detection method and system based on inspection data of monocular distribution network machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311678629.7A CN117635589A (en) | 2023-12-07 | 2023-12-07 | External damage detection method and system based on inspection data of monocular distribution network machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117635589A true CN117635589A (en) | 2024-03-01 |
Family
ID=90016190
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311678629.7A Pending CN117635589A (en) | 2023-12-07 | 2023-12-07 | External damage detection method and system based on inspection data of monocular distribution network machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117635589A (en) |
-
2023
- 2023-12-07 CN CN202311678629.7A patent/CN117635589A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108037133B (en) | Intelligent electric power equipment defect identification method and system based on unmanned aerial vehicle inspection image | |
| US11361423B2 (en) | Artificial intelligence-based process and system for visual inspection of infrastructure | |
| CN112613569B (en) | Image recognition method, training method and device for image classification model | |
| CN110910360B (en) | Positioning method of power grid image and training method of image positioning model | |
| CN116363125B (en) | Deep learning-based battery module appearance defect detection method and system | |
| CN113723189B (en) | Intelligent power equipment fault diagnosis method based on single-order infrared image target detection | |
| CN111044149A (en) | Method and device for detecting temperature abnormal point of voltage transformer and readable storage medium | |
| CN113515655A (en) | Fault identification method and device based on image classification | |
| CN113205039A (en) | Power equipment fault image identification and disaster investigation system and method based on multiple DCNNs | |
| CN115830407A (en) | Cable pipeline fault discrimination algorithm based on YOLOV4 target detection model | |
| CN112949457A (en) | Maintenance method, device and system based on augmented reality technology | |
| CN109145752A (en) | For assessing the method, apparatus, equipment and medium of object detection and track algorithm | |
| CN115239646A (en) | Defect detection method and device for power transmission line, electronic equipment and storage medium | |
| CN114724140A (en) | Strawberry maturity detection method and device based on YOLO V3 | |
| CN117114420B (en) | Image recognition-based industrial and trade safety accident risk management and control system and method | |
| CN114694130A (en) | Method and device for detecting telegraph poles and pole numbers along railway based on deep learning | |
| CN110618129A (en) | Automatic power grid wire clamp detection and defect identification method and device | |
| CN111931721B (en) | Method and device for detecting color and number of annual inspection label and electronic equipment | |
| CN114120086A (en) | Pavement disease recognition method, image processing model training method, device and electronic equipment | |
| CN111738312B (en) | Power transmission line state monitoring method and device based on GIS and virtual reality fusion and computer readable storage medium | |
| CN117635589A (en) | External damage detection method and system based on inspection data of monocular distribution network machine | |
| CN115438945A (en) | Risk identification method, device, equipment and medium based on power equipment inspection | |
| CN115187880A (en) | Communication optical cable defect detection method and system based on image recognition and storage medium | |
| CN112529881A (en) | Method and device for identifying cable abnormity of electric control cabinet | |
| CN112070730A (en) | Anti-vibration hammer falling detection method based on power transmission line inspection image |
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 |