CN115311539A - Overhead transmission line defect identification method, device, equipment and storage medium - Google Patents
Overhead transmission line defect identification method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN115311539A CN115311539A CN202210775132.6A CN202210775132A CN115311539A CN 115311539 A CN115311539 A CN 115311539A CN 202210775132 A CN202210775132 A CN 202210775132A CN 115311539 A CN115311539 A CN 115311539A
- Authority
- CN
- China
- Prior art keywords
- detection image
- defect
- transmission line
- overhead transmission
- defect identification
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 208
- 230000005540 biological transmission Effects 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000001514 detection method Methods 0.000 claims abstract description 159
- 238000007689 inspection Methods 0.000 claims abstract description 47
- 238000004458 analytical method Methods 0.000 claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 16
- 230000010354 integration Effects 0.000 claims abstract description 10
- 238000012360 testing method Methods 0.000 claims description 33
- 238000012549 training Methods 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 12
- 230000009466 transformation Effects 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 11
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000741 silica gel Substances 0.000 description 1
- 229910002027 silica gel Inorganic materials 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a method for identifying defects of an overhead transmission line, which comprises the following steps: receiving a first detection image generated by real-time inspection and photographing when the overhead transmission line robot enters a preset inspection area; inputting the first detection image into a target defect identification model based on a Faster-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of defect identification of the first detection image; and sending the judgment result to the intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report. Receiving a first detection image generated by real-time inspection and photographing of the overhead transmission line robot; and inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the first detection image. The target defect identification model based on the Faster-RCNN algorithm can effectively improve the defect identification of the detected image, improve the efficiency and the working efficiency and reduce errors in the defect identification.
Description
Technical Field
The invention relates to the field of artificial intelligence, in particular to a method, a device, equipment and a storage medium for identifying defects of an overhead transmission line.
Background
At present, the novel power line inspection modes at home and abroad mainly comprise: the unmanned helicopter is used for routing inspection, the unmanned plane is used for routing inspection, and the line robot is used for routing inspection. The electric equipment of the overhead transmission line has defects due to the reasons of construction technology, operation environment, external factors and the like in the operation process, and the main intelligent detection means aiming at the defects in the overhead transmission line at present is to acquire data through visible light, infrared and ultraviolet imagers and analyze and judge the data on the basis. The equipment is detected by using an infrared thermal imager, an ultraviolet imager and the like, and the characteristics of sound, temperature abnormality, abnormal discharge and the like generated when the equipment in the overhead transmission line breaks down are mainly utilized. However, infrared and ultraviolet images cannot effectively show the surface details of the equipment, and are not suitable for surface defects without light-emitting and heating characteristics, such as rusting of a cabinet door abnormal switch and a transformer in an overhead transmission line, opening and closing of a circuit breaker, breakage of an indicating dial plate, absence of a silica gel insulator umbrella skirt and the like, so that a visible light camera is still required to be used for surface defect detection.
The traditional method is to look up pictures of the power equipment one by one on a computer after the pictures are taken, and the method has the advantages of low working efficiency, large workload, easy fatigue generation of people after long working time, easy defect omission and incapability of meeting the requirements of safe and efficient routing inspection of modern power grids.
Disclosure of Invention
The invention mainly aims to provide a method, a device, equipment and a storage medium for identifying defects of an overhead transmission line, and aims to solve the technical problem that detection images of some defects are easy to omit due to the fact that the detection images of electrical equipment are shot based on manual inspection.
The invention provides a method for identifying defects of an overhead transmission line, which comprises the following steps:
receiving a first detection image generated when the overhead transmission line robot enters a preset inspection area to perform real-time inspection and photographing;
inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of defect identification of the first detection image;
and sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
Further, after receiving the step that overhead transmission line robot got into to predetermine and patrols and examines the region and patrol and examine the first detection image that the photograph generated in real time, include:
carrying out data enhancement on the first detection image to obtain a detection image sample set;
dividing the detection image sample set into a training set and a test set according to a preset proportion;
inputting the training set into a defect recognition model for training to obtain a trained defect recognition model;
inputting the test set into the trained defect recognition model for testing, judging that the test result of the trained defect recognition model meets the requirement, and taking the trained defect recognition model as the target defect recognition model.
Further, the step of obtaining a detection image sample set after performing data enhancement on the first detection image includes:
carrying out homogeneous data enhancement on the first detection image to obtain a homogeneous data enhanced first detection image;
based on a mixed data enhancement algorithm, fusing the homogeneous data enhanced first detection images according to a preset proportion to obtain fused first detection images;
and classifying the fused first detection image and detecting a target to obtain the detection image sample set.
Further, the step of performing homogeneous data enhancement on the first detection image to obtain a homogeneous data enhanced first detection image includes:
and horizontally turning or vertically turning or rotating the first detection image, and extracting the homogeneous data to enhance the first detection image.
Further, the step of inputting the test set into the trained defect recognition model for testing, and if it is determined that the test result of the trained defect recognition model meets the requirement, taking the trained defect recognition model as the target defect recognition model further includes:
obtaining a loss function value of the trained defect recognition model, and comparing the loss function value with an optimized value of the trained defect recognition model;
and if the loss function value and the optimized value are within a preset difference value range, judging that the test result of the trained defect recognition model meets the requirement.
Further, the formula for calculating the loss function value of the trained defect recognition model is as follows:
where n is the number of the detected image sample set, y is the true value of the detected image sample set, and f (x) i ) Is the weight of the trained defect recognition model for the ith iteration.
Further, after the step of obtaining a detection image sample set after performing data enhancement on the first detection image, the method includes:
and acquiring the detection image sample set, and carrying out pixel brightness transformation and/or geometric change and/or gray level transformation on the detection image sample set.
Further, an overhead transmission line defect recognition device, the device includes:
the receiving and photographing module is used for receiving a first detection image generated by real-time routing inspection and photographing when the overhead transmission line robot enters a preset routing inspection area;
the analysis result module is used for inputting the first detection image into a target defect identification model based on a Faster-RCNN algorithm, analyzing the first detection image and obtaining a judgment result of the defect identification of the first detection image;
and the generating module is used for sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
Further, a computer device comprising a memory storing a computer program and a processor implementing the method of any of the above when the processor executes the computer program.
Further, a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the above-mentioned method.
Compared with the prior art, the invention has the beneficial effects that: receiving a first detection image generated by real-time routing inspection and photographing of the overhead transmission line robot; inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of defect identification of the first detection image; and sending the judgment result to an intelligent defect analysis platform and generating a defect report. The target defect identification model based on the Faster-RCNN algorithm can effectively improve the defect identification of the detected image, improve the efficiency and the working efficiency and reduce errors in the defect identification.
Drawings
Fig. 1 is a schematic step diagram of a method for identifying defects of an overhead transmission line according to an embodiment of the present invention;
fig. 2 is a schematic step diagram of a method for identifying defects of an overhead transmission line according to an embodiment of the present invention;
fig. 3 is a schematic step diagram of a method for identifying defects of an overhead transmission line according to an embodiment of the present invention;
fig. 4 is a schematic step diagram of a method for identifying defects of an overhead transmission line according to an embodiment of the present invention;
fig. 5 is a block diagram schematically illustrating the structure of an overhead transmission line defect identifying apparatus according to an embodiment of the present invention;
FIG. 6 is a block diagram of a computer device according to an embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, the application provides a method for identifying defects of an overhead transmission line, which includes the following steps S100-S300:
s100: receiving a first detection image generated when the overhead transmission line robot enters a preset inspection area to perform real-time inspection and photographing;
s200: inputting the first detection image into a target defect identification model based on a Faster-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of the defect identification of the first detection image;
s300: and sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the first judgment result to obtain a second judgment result, and generating a defect report based on the second judgment result.
For S100, the transformer substation of the high-voltage transmission line has the tasks of changing voltage grades, adjusting voltage, collecting current and distributing electric energy in the power system, so that the guarantee of normal stability of the high-voltage transmission line has important significance for stable operation of the power system. Traditional power equipment tours mainly adopts the mode of artifical patrolling and examining, but along with the constantly development of social economy, the power consumption increases day by day, and power equipment's quantity increases by a wide margin, appears patrolling and examining the problem that personnel shortage is not enough. The overhead transmission line robot is used for power equipment inspection, and the improvement of inspection efficiency becomes an important means. In the embodiment, the inspection area to be detected is obtained, the map on the intelligent terminal is obtained, the number of the inspection area on the map is determined by clicking, and the inspection area is input into the overhead transmission line robot to be operated on the map. The overhead transmission line robot is driven into the inspection area to shoot a first detection image, and shooting is performed for a preset time period, for example, 30 minutes. It is to be noted that the first detection image in the present embodiment is not only one image, but may be two, three, five hundred, and so on. The overhead transmission line robot can return to the workstation after finishing shooting for a preset time, and a first detection image generated by shooting is stored in the database, and meanwhile, the generated first detection image is also stored in the cloud server for backup, so that the detection image in the database is prevented from being lost and being acquired from the cloud server.
For S200, the main means of intelligent detection of the routing inspection defects of the power equipment aiming at the overhead transmission line at present is to acquire data through visible light, infrared and ultraviolet imagers and analyze and judge the data on the basis. The infrared thermal imager, the ultraviolet imaging and other devices mainly utilize a series of characteristics of sound, abnormal temperature or abnormal discharge and the like of the power equipment when the power equipment fails at present. The imager may not effectively display the surface details of the device and may sometimes not effectively detect surface defect detection of the inspection image. In recent years, artificial intelligence algorithms are widely applied, and are based on the Faster-RCNN algorithm to solve the problem that defects are difficult to accurately locate and identify. The fast-RCNN is a Faster regional convolutional neural network (fast-Region with CNN feature), the fast-RCNN algorithm extracts candidate frames and classifies defects into one network, the tedious process of manually designing a target frame is omitted, the detection efficiency is greatly improved, and the fast-RCNN model is superior to other algorithms in the aspects of identification accuracy and positioning accuracy through verification of a defect data set of the power equipment. And inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the input first detection image to obtain a judgment result of defect identification of the first detection image. The evaluation result comprises the defect type belonging to the first detection image and the like.
For S300, obtaining a judgment result of defect identification of the first detection image, storing the judgment result into a text for storage, and sending the stored text to the intelligent defect analysis platform. And extracting similar defect types from the intelligent defect analysis platform database based on the obtained judgment result, integrating and processing the similar defect types, and generating a defect report. The defect report comprises the proportion of the defects in the first detection image and the types of the defects, and the proportion of the defects in the whole intelligent defect analysis platform database, wherein the defects are obtained from the intelligent defect analysis platform database and are the same as the defects in the first detection image. And the shooting area of the defect in the first detection image is convenient to position the defect area. The defect report is sent to the intelligent terminal, so that the defect report can be conveniently inquired and watched.
Referring to fig. 2, in an embodiment of the application, after the step of receiving a first detection image generated by real-time inspection photographing of an overhead transmission line robot entering a preset inspection area, the method includes:
s110: performing data enhancement on the first detection image to obtain a detection image sample set;
s120: dividing the detection image sample set into a training set and a test set according to a preset proportion;
s130: inputting the training set into a defect recognition model for training to obtain a trained defect recognition model;
s140: inputting the test set into the trained defect recognition model for testing, and taking the trained defect recognition model as the target defect recognition model when the test result of the trained defect recognition model meets the requirements.
For S110, a large number of sample sets are often needed for training the model, and when the training data is limited, the accuracy of the trained model cannot be improved. And obtaining a detection image sample set based on data enhancement of the first detection image. The data enhancement method is a process of generating incremental data according to the existing data samples and rules on the premise of not collecting more data on the basis of the existing limited data.
For S120, the detection image sample set is divided into a training set and a test set according to a preset ratio, for example, the ratio of the training set to the test set may be 7. The training set is used for training the defect recognition model, and the testing set is used for testing the defect recognition model.
For S130, inputting the training set into the defect recognition model for training to obtain a trained defect recognition model, where the defect recognition model is an initial defect recognition model.
And S140, acquiring the trained defect recognition model, and testing the trained defect recognition model. Inputting the test set into the trained defect recognition model, and if the test accuracy of the test set output is within a preset range, the trained defect recognition model meets the requirements, and the trained defect recognition model can be used as a target defect recognition model. The target defect identification model can be used for identifying defects of detection images generated by real-time inspection and photographing of the overhead transmission line robot in the follow-up process.
Referring to fig. 3, in an embodiment of the present application, the step of obtaining a sample set of detection images after performing data enhancement on the first detection image includes:
s111: carrying out homogeneous data enhancement on the first detection image to obtain a homogeneous data enhanced first detection image;
s112: based on a mixed data enhancement algorithm, fusing the homogeneous data enhanced first detection images according to a preset fixed proportion to obtain fused first detection images;
s113: and classifying the fused first detection image and detecting a target to obtain the detection image sample set.
And S111, performing homogeneous data enhancement on the first detection image, wherein the homogeneous data enhancement method is to increase data sets belonging to the same class to obtain homogeneous data enhanced first detection images and enhance the generalization capability of the defect identification model.
And S112, acquiring a homogeneous data enhanced first detection image, fusing the homogeneous data enhanced first detection image according to a preset proportion based on a mixed data enhancement algorithm, fusing different types of detection images in the homogeneous data enhanced first detection image, wherein the fusion proportion is a random real number between [0,1], and obtaining a fused first detection image. The mixed data enhancement algorithm is used for carrying out linearization processing on homogeneous data to enhance the area between the first detection image and the homogeneous data to enhance the area between the first detection images, seeking for enhancing the adaptability of the model to predict the data set except the training sample, and further improving the robustness of the defect identification model to the sample identification and the judgment capability of the error category.
For S113, the fused first detection image is used as a detection image sample set, and target detection and classification are performed on the fused first detection image, for example, the acquisition target is a ground line and a category of the ground line, the acquisition target is a tower bolt and a category of the tower bolt, and the like.
In an embodiment of the application, the step of performing homogeneous data enhancement on the first detected image to obtain a homogeneous data enhanced first detected image includes:
s115: and carrying out horizontal turning or vertical turning or rotation processing on the first detection image, and extracting the homogeneous data to enhance the first detection image.
For S115, a first detection image is acquired, and the first detection image is horizontally flipped or vertically flipped or rotated, where the rotation may be any angle, such as 30 °, 45 °, 55 °, or counterclockwise, or clockwise. Homogeneous data enhancement is performed based on rotation of any angle, so that homogeneous data enhanced first detection images are obtained. And extracting homogeneous data enhanced first detection images, and taking the homogeneous data enhanced first detection images as a sample set of a subsequent defect identification model.
Referring to fig. 4, in an embodiment of the present application, the step of inputting the test set into the trained defect recognition model for testing, and if it is determined that the test result of the trained defect recognition model meets the requirement, taking the trained defect recognition model as the target defect recognition model further includes:
s141: obtaining a loss function value of the trained defect identification model, and comparing the loss function value with an optimized value of the trained defect identification model;
s142: and if the loss function value and the optimized value are within a preset difference range, judging that the test result of the trained defect recognition model meets the requirement.
For step S141, a defect recognition model after training is obtained, and a loss function value of the defect recognition model after training is further calculated, where the loss function is used to express a difference degree between prediction and actual data. And calculating to obtain a loss function value of the trained defect recognition model, and comparing the calculated loss function value with an optimized value of the trained defect recognition model.
And S142, judging the loss function value and the optimized value of the trained defect recognition model, if the loss function value and the optimized value are within a preset difference value range, judging that the trained defect recognition model meets the requirement, judging that the test result of the trained defect recognition model meets the requirement, and taking the trained defect recognition model as the target defect recognition model.
In an embodiment of the present application, the calculation formula of the loss function value of the trained defect recognition model is as follows:
where n is the number of the detected image sample set, y is the true value of the detected image sample set, and f (x) i ) Is the weight of the trained defect recognition model for the ith iteration. Through the upper partThe calculation of the formula can obtain the loss function value of the trained defect identification model.
In an embodiment of the application, after the step of obtaining a sample set of inspection images by performing data enhancement on the first inspection image, the method includes:
s114: and acquiring the detection image sample set, and carrying out pixel brightness transformation and/or geometric change and/or gray level transformation on the detection image sample set.
For S114, when data are collected, the types of the power equipment are numerous, factors such as shooting angles, illumination and weather influence the characteristics of the collected images of the power equipment, and therefore the image recognition effect is interfered. For the pixel brightness conversion processing of the detected image sample set, the image quality can be effectively improved, the image is clearer, and the generalization capability under different weather conditions can be improved. For the geometric change processing of the detection image sample set, the problem of significant size difference existing in the images of the power equipment is solved, the extraction of multi-scale features can be ensured, the effectiveness of large-scale feature maps is enhanced, and a default frame generation strategy and a training strategy are optimized to be more suitable for the construction data set of the text. For gray level transformation processing of a detection image sample set, contrast enhancement is realized, and thus, an image is more effectively identified.
Referring to fig. 5, in an embodiment of the present application, an overhead transmission line defect identifying apparatus includes:
the receiving and photographing module 100 is used for receiving a first detection image generated by real-time polling and photographing when the overhead transmission line robot enters a preset polling area;
an analysis result module 200, configured to input the first detection image to a target defect identification model based on a fast-RCNN algorithm, and analyze the first detection image to obtain a judgment result of defect identification of the first detection image;
and a generating module 300, configured to send the evaluation result to an intelligent defect analysis platform, call an intelligent defect analysis platform database for integration processing based on the evaluation result, and generate a defect report.
For the receiving and photographing module 100, the transformer substation of the high-voltage transmission line has the tasks of converting voltage grades, adjusting voltage, collecting current and distributing electric energy in the power system, so that the guarantee of normal stability of the high-voltage transmission line has important significance for stable operation of the power system. Traditional power equipment patrols and examines the mode that mainly adopts the manual work to patrol and examine, nevertheless along with the continuous development of socioeconomic, the power consumption increases with each day, and the quantity of power equipment increases by a wide margin, appears patrolling and examining personnel shortage not enough problem. The overhead transmission line robot is used for power equipment inspection, and the improvement of inspection efficiency becomes an important means. In the embodiment, the inspection area to be detected is obtained, the map on the intelligent terminal is obtained, the number of the inspection area on the map is determined by clicking, and the inspection area on the map is input into the overhead transmission line robot to be operated. And driving the overhead transmission line robot into the inspection area to shoot a first detection image, and shooting for a preset time period, for example, 30 minutes. It is to be noted that the first detection image in the present embodiment is not only one image, but may be two, three, five hundred, and so on. The overhead transmission line robot can return to the workstation after finishing shooting for a preset time, and a first detection image generated by shooting is stored in the database, and meanwhile, the generated first detection image is also stored in the cloud server for backup, so that the detection image in the database is prevented from being lost and being acquired from the cloud server.
For the analysis result module 200, the main means of intelligent detection of the routing inspection defects of the power equipment aiming at the overhead transmission line at present is to acquire data through visible light, infrared and ultraviolet imagers and analyze and judge on the basis. The infrared thermal imager, the ultraviolet imaging and other devices mainly utilize a series of characteristics of sound, abnormal temperature or abnormal discharge and the like of the power equipment when the power equipment fails at present. The imager may not effectively display the surface details of the device and may sometimes not effectively detect surface defect detection of the inspection image. In recent years, artificial intelligence algorithms are widely applied, and are based on the Faster-RCNN algorithm to solve the problem that defects are difficult to accurately locate and identify. The fast-RCNN is a Faster regional convolutional neural network (fast-Region with CNN feature), the fast-RCNN algorithm extracts candidate frames and classifies defects into one network, the complex process of manually designing a target frame is omitted, the detection efficiency is greatly improved, and through verification of a defect data set of the power equipment, the fast-RCNN model is superior to other algorithms in recognition accuracy and positioning accuracy. And inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the input first detection image to obtain a judgment result of defect identification of the first detection image. Wherein, the judgment result comprises the defect type belonging to the first detection image and the like.
For the generating module 300, a judgment result of the defect identification of the first detection image is obtained, the judgment result is stored in a text for storage, and the stored text is sent to the intelligent defect analysis platform. And based on the obtained evaluation result, extracting the similar defect types from the intelligent defect analysis platform database for integration treatment, and generating a defect report. The defect report comprises the proportion of the defects in the first detection image and the types of the defects, and the proportion of the defects in the whole intelligent defect analysis platform database, which are the same as the defects in the first detection image, is obtained from the intelligent defect analysis platform database. And the shooting area of the defect in the first detection image is convenient to position the area of the defect. The defect report is sent to the intelligent terminal, so that the defect report can be conveniently inquired and watched.
In an embodiment of the present application, a computer device is further provided, where the computer device may be a server, and an internal structure of the computer device may be as shown in fig. 6. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the computer designed processor is used to provide computational and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a method for identifying the defects of the overhead transmission line, and the method comprises the following steps: receiving a first detection image generated when the overhead transmission line robot enters a preset inspection area to perform real-time inspection and photographing; inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of defect identification of the first detection image; and sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements a method for identifying defects of an overhead transmission line, the method including the steps of: receiving a first detection image generated by real-time inspection and photographing when the overhead transmission line robot enters a preset inspection area; inputting the first detection image into a target defect identification model based on a fast-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of defect identification of the first detection image; and sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above may be implemented by hardware instructions of a computer program, which may be stored in a non-volatile computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium provided herein and used in the examples may include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRDRAM), enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), rambus (Rambus) direct RAM (RDRAM), direct bused dynamic RAM (DRDRAM), and bused dynamic RAM (RDRAM).
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of another identical element in a process, apparatus, article, or method comprising the element.
The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are also included in the scope of the present application.
Claims (10)
1. A method for identifying defects of overhead transmission lines is characterized by comprising the following steps:
receiving a first detection image generated when the overhead transmission line robot enters a preset inspection area to perform real-time inspection and photographing;
inputting the first detection image into a target defect identification model based on a Faster-RCNN algorithm, and analyzing the first detection image to obtain a judgment result of the defect identification of the first detection image;
and sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
2. The overhead transmission line defect identification method according to claim 1, wherein after the step of receiving the first detection image generated by real-time inspection photographing of the overhead transmission line robot entering the preset inspection area, the method comprises:
carrying out data enhancement on the first detection image to obtain a detection image sample set;
dividing the detection image sample set into a training set and a test set according to a preset proportion;
inputting the training set into a defect recognition model for training to obtain a trained defect recognition model;
inputting the test set into the trained defect recognition model for testing, judging that the test result of the trained defect recognition model meets the requirement, and taking the trained defect recognition model as the target defect recognition model.
3. The overhead transmission line defect identification method according to claim 1, wherein the step of obtaining a detection image sample set after data enhancement of the first detection image comprises:
carrying out homogeneous data enhancement on the first detection image to obtain a homogeneous data enhanced first detection image;
based on a mixed data enhancement algorithm, fusing the homogeneous data enhanced first detection images according to a preset fixed proportion to obtain fused first detection images;
and classifying the fused first detection image and detecting a target to obtain the detection image sample set.
4. The overhead transmission line defect identification method according to claim 3, wherein the step of performing homogeneous data enhancement on the first detection image to obtain a homogeneous data enhanced first detection image comprises:
and carrying out horizontal turning or vertical turning or rotation processing on the first detection image, and extracting the homogeneous data to enhance the first detection image.
5. The overhead transmission line defect identification method according to claim 2, wherein the step of inputting the test set into the trained defect identification model for testing, and taking the trained defect identification model as the target defect identification model if the test result of the trained defect identification model is judged to meet the requirements, further comprises:
obtaining a loss function value of the trained defect identification model, and comparing the loss function value with an optimized value of the trained defect identification model;
and if the loss function value and the optimized value are within a preset difference value range, judging that the test result of the trained defect recognition model meets the requirement.
6. The overhead transmission line defect identification method according to claim 5, wherein a calculation formula of a loss function value of the trained defect identification model is:
where n is the number of the detected image sample set, y is the true value of the detected image sample set, and f (x) i ) Is the weight of the trained defect recognition model for the ith iteration.
7. The overhead transmission line defect identification method according to claim 3, wherein after the step of obtaining a detection image sample set by performing data enhancement on the first detection image, the method comprises:
and acquiring the detection image sample set, and carrying out pixel brightness transformation and/or geometric change and/or gray level transformation on the detection image sample set.
8. An overhead transmission line defect recognition device, characterized in that, the device includes:
the receiving and photographing module is used for receiving a first detection image generated by real-time routing inspection and photographing when the overhead transmission line robot enters a preset routing inspection area;
the analysis result module is used for inputting the first detection image into a target defect identification model based on a Faster-RCNN algorithm, analyzing the first detection image and obtaining a judgment result of the defect identification of the first detection image;
and the generating module is used for sending the judgment result to an intelligent defect analysis platform, calling an intelligent defect analysis platform database for integration processing based on the judgment result, and generating a defect report.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method of any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210775132.6A CN115311539A (en) | 2022-07-01 | 2022-07-01 | Overhead transmission line defect identification method, device, equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210775132.6A CN115311539A (en) | 2022-07-01 | 2022-07-01 | Overhead transmission line defect identification method, device, equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115311539A true CN115311539A (en) | 2022-11-08 |
Family
ID=83856915
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210775132.6A Pending CN115311539A (en) | 2022-07-01 | 2022-07-01 | Overhead transmission line defect identification method, device, equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115311539A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116054414A (en) * | 2023-03-06 | 2023-05-02 | 广东电网有限责任公司佛山供电局 | Line defect hidden danger monitoring method, device, computer equipment and storage medium |
-
2022
- 2022-07-01 CN CN202210775132.6A patent/CN115311539A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116054414A (en) * | 2023-03-06 | 2023-05-02 | 广东电网有限责任公司佛山供电局 | Line defect hidden danger monitoring method, device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112734692B (en) | Defect identification method and device for power transformation equipment | |
CN110598736B (en) | Power equipment infrared image fault positioning, identifying and predicting method | |
CN108765373A (en) | A kind of insulator exception automatic testing method based on integrated classifier on-line study | |
CN108734143A (en) | A kind of transmission line of electricity online test method based on binocular vision of crusing robot | |
CN110703800A (en) | Unmanned aerial vehicle-based intelligent identification method and system for electric power facilities | |
CN110567964B (en) | Method and device for detecting defects of power transformation equipment and storage medium | |
CN108537154A (en) | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning | |
CN112766103A (en) | Machine room inspection method and device | |
CN112697798B (en) | Infrared image-oriented diagnosis method and device for current-induced thermal defects of power transformation equipment | |
CN112613454A (en) | Electric power infrastructure construction site violation identification method and system | |
CN116191680B (en) | Monitoring management system applied to photovoltaic power generation | |
CN109389146A (en) | Equipment state feedback method and system based on neural network image sorting algorithm | |
CN108229524A (en) | A kind of chimney and condensing tower detection method based on remote sensing images | |
CN112508911A (en) | Rail joint touch net suspension support component crack detection system based on inspection robot and detection method thereof | |
CN115311539A (en) | Overhead transmission line defect identification method, device, equipment and storage medium | |
CN116228670A (en) | Yolov 5-based photovoltaic defect detection method, device, computer equipment and medium | |
CN114049620A (en) | Image data feature extraction and defect identification method, device and system | |
CN117809297B (en) | Three-dimensional reconstruction-based intelligent identification method for dangerous source of power transmission line | |
CN115171045A (en) | YOLO-based power grid operation field violation identification method and terminal | |
CN117252840B (en) | Photovoltaic array defect elimination evaluation method and device and computer equipment | |
CN113778091A (en) | Method for inspecting equipment of wind power plant booster station | |
CN109712018A (en) | A kind of power transmission line comprehensive monitoring system | |
CN117636129A (en) | Identification method of power construction supervision image | |
CN113284103B (en) | Substation equipment defect online detection method based on space transformation fast R-CNN model | |
CN116486307A (en) | Chemical industry park safety early warning method and system based on video analysis technology |
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 |