CN114429421A - Method for detecting state of disconnecting link facing scheduling service scene - Google Patents
Method for detecting state of disconnecting link facing scheduling service scene Download PDFInfo
- Publication number
- CN114429421A CN114429421A CN202011097719.3A CN202011097719A CN114429421A CN 114429421 A CN114429421 A CN 114429421A CN 202011097719 A CN202011097719 A CN 202011097719A CN 114429421 A CN114429421 A CN 114429421A
- Authority
- CN
- China
- Prior art keywords
- state
- disconnecting link
- data set
- switch
- training
- 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
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000007689 inspection Methods 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 claims description 31
- 238000012549 training Methods 0.000 claims description 28
- 238000012360 testing method Methods 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000002372 labelling Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 238000013434 data augmentation Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 abstract description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4046—Scaling of whole images or parts thereof, e.g. expanding or contracting using neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/327—Testing of circuit interrupters, switches or circuit-breakers
-
- 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
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/60—Rotation of whole images or parts thereof
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for detecting the state of a switch facing a scheduling service scene, which comprises the following steps: and reading the state of the disconnecting link by a sensor, identifying the state of the disconnecting link by an image, comparing the states and the like. The method for detecting the state of the disconnecting link facing to the dispatching service scene has high correct recognition rate, relatively low calculation complexity and high operation efficiency, so that the method can be rapidly deployed in the power industry in large quantities, and the intelligent level of power inspection work can be effectively improved.
Description
Technical Field
The invention relates to a method for detecting the state of a disconnecting link facing a scheduling service scene, and belongs to the technical field of power equipment.
Background
With the continuous promotion of electric power construction work, the scale of the power grid in China is continuously enlarged. The method also brings huge pressure to the operation and maintenance work of the power transformation equipment while promoting the economic construction process of China. On the other hand, as robotics and cameras have grown mature, remote camcorder systems based on robots and cameras have been successfully applied to power patrol work. However, due to the lack of a matched intelligent processing technology, the video images obtained by inspection still need to be manually analyzed, the method is not only low in efficiency, but also low in accuracy, easily causes missing inspection and false inspection, and cannot deal with increasingly severe operation and maintenance pressure.
At present, most transformer substations acquire the equipment state by acquiring the sensor information of the equipment state, and the method has certain false detection and easily causes electric power hidden trouble. The method for analyzing the equipment state of the transformer substation power equipment images shot by the camera by adopting an artificial intelligence analysis method is developed in some transformer substations, but the analysis result is not comprehensively distinguished from the analysis result of the intelligent power grid dispatching control system D5000.
Disclosure of Invention
The invention provides a switch state detection method for a scheduling service scene, which is used for carrying out double judgment and analysis on the state of routing inspection equipment and improving the accuracy of equipment state judgment.
In order to solve the technical problems, the technical scheme provided by the invention is as follows: a method for detecting the state of a switch facing a scheduling service scene comprises the following steps:
step S1, at the station end, the sensor corresponding to the disconnecting link transmits the switch state corresponding to the current disconnecting link to the intelligent power grid dispatching control system at the main station end through the network;
step S2, the station side acquires a disconnecting link image through a camera, then analyzes the current disconnecting link state through a method based on a deep neural network and deployed on an intelligent analysis device, and uploads the disconnecting link image and the current disconnecting link state obtained through analysis to a unified video monitoring platform of the main station side through a network;
step S3, comparing the switch state of the disconnecting link obtained by the intelligent power grid dispatching control system with the current disconnecting link state obtained by the unified video monitoring platform, if the two states are consistent, returning the current state to the intelligent power grid dispatching control system and displaying; and if the two states are not consistent, manually checking the disconnecting link image, confirming the state of the disconnecting link, and returning the state of the disconnecting link to the intelligent power grid dispatching control system for displaying.
The scheme is further improved in that: the method for using the deep neural network in the step S2 includes the following steps:
step 1, collecting a power inspection video image of power equipment; the power equipment comprises a single-arm folding type isolating switch, a double-column rotary type isolating switch and a three-column type isolating switch; a power patrol video image at least comprises a power device;
step 2, labeling the power patrol video image, wherein the labeling information comprises a target frame and a category label;
step 3, preparing a training data set and a testing data set before training the target detection network model; randomly dividing a training data set and a testing data set;
step 4, preprocessing the training data set; normalizing the size, and acquiring target images at different angles by adopting a data augmentation mode;
step 7, training a neural network to obtain a target detection network model;
step 8, testing the detection performance of the target detection network model through the test data set, and if the detection performance requirement is not met, continuing training on the training data set until the requirement is met; the detection performance takes the mAP value as a measurement index, and takes 0.6 as a qualified standard.
The scheme is further improved in that: the target detection network model is a YOLOV3 target detection network model.
The scheme is further improved in that: the size normalization is a uniform normalization of the images in the training dataset to 300 x 300 pixels.
The scheme is further improved in that: the data augmentation is by rotating the images in the training dataset 90 °, 180 ° and 270 ° clockwise.
The scheme is further improved in that: the intelligent power grid dispatching control system is a D5000 system.
The method for detecting the state of the disconnecting link facing the scheduling service scene, provided by the invention, is used for detecting and identifying typical power equipment in a power inspection video image by applying a deep learning theory and utilizing a deep convolutional neural network with better performance in the industry at present, and can effectively solve the problems of low detection performance, low operation efficiency and the like of other target detection methods; the state of the disconnecting link is comprehensively analyzed by adopting the disconnecting link state detection based on video analysis and the disconnecting link state detection based on sensor information, so that the state of the disconnecting link is dualized and judged in a scheduling service scene; the method has the advantages of high correct recognition rate, relatively low calculation complexity, high operation efficiency and low requirement on computer hardware, so that the method can be rapidly and massively deployed in the power industry, and the intelligent level of power inspection work can be effectively improved.
Drawings
The invention is further described below with reference to the accompanying drawings.
Fig. 1 is a schematic view of a preferred embodiment of the present invention.
Detailed Description
Examples
As shown in fig. 1, the method for detecting a state of a switch facing a scheduling service scenario, provided by this embodiment, includes the following steps:
step S1, at the station end, the sensor corresponding to the disconnecting link transmits the switch state corresponding to the current disconnecting link to the intelligent power grid dispatching control system at the main station end, namely a D5000 system, through the network;
step S2, the station side acquires a disconnecting link image through a camera, then analyzes the current disconnecting link state through a method based on a deep neural network and deployed on an intelligent analysis device, and uploads the disconnecting link image and the current disconnecting link state obtained through analysis to a unified video monitoring platform of the main station side through a network;
step S3, comparing the switch state of the disconnecting link obtained by the intelligent power grid dispatching control system with the current disconnecting link state obtained by the unified video monitoring platform, if the two states are consistent, returning the current state to the intelligent power grid dispatching control system and displaying; and if the two states are not consistent, manually checking the disconnecting link image, confirming the state of the disconnecting link, and returning the state of the disconnecting link to the intelligent power grid dispatching control system for displaying.
Further, the method using the deep neural network in step S2 includes the following steps:
step 1, collecting a power inspection video image of power equipment; when the power patrol video image is collected, the image is kept clear and the power equipment is highlighted; the power equipment comprises a single-arm folding type isolating switch, a double-column rotary type isolating switch and a three-column type isolating switch; a power patrol video image at least comprises a power device;
step 2, labeling the power patrol video image, wherein the labeling information comprises a target frame and a category label;
step 3, preparing a training data set and a testing data set before training a YOLOV3 target detection network model; randomly dividing a training data set and a testing data set; the images contained in the two images are independent of each other and the quantity ratio is 17: 3;
step 4, preprocessing the training data set; size normalization, wherein the images are uniformly normalized to 300 × 300 pixels; data augmentation, including generating copies of the original image after clockwise rotation by 90 °, 180 ° and 270 °;
step 7, training a neural network to obtain a target detection network model; the Yolov3 target detection network model is obtained based on an electric power map database established by screening electric power inspection video images and a Yolov3 neural network training, and specifically comprises the following steps:
firstly, initializing each parameter of a YOLOV3 target detection network;
thirdly, training the network by adopting a channel sparse regularization method on the BN layer;
then, cutting the training network channel through a small-size factor;
and finally, fine-tuning the cut network to obtain a trained network model. Repeating the above steps for multiple times until the trained network model reaches a set compression threshold value;
step 8, testing the detection performance of the target detection network model through the test data set, and if the detection performance requirement is not met, continuing training on the training data set until the requirement is met; the detection performance takes the mAP value as a measurement index, and takes 0.6 as a qualified standard. That is, a value of 0.6 or more is acceptable.
In conclusion, the invention aims at the intelligent identification problem of the routing inspection video image in the power scene, establishes a high-quality power picture database, and trains and obtains a target detection model capable of identifying a typical power device in an input picture by taking a YOLOV3 network as a core on the basis. The experimental result on the test set shows that the method has higher recognition rate for the typical power device, can be applied to the field of power inspection video analysis, and has strong feasibility and practicability.
The present invention is not limited to the above embodiments, and any technical solutions formed by equivalent substitutions fall within the scope of the claims of the present invention.
Claims (6)
1. A method for detecting the state of a switch facing a scheduling service scene is characterized by comprising the following steps:
step S1, at the station end, the sensor corresponding to the disconnecting link transmits the switch state corresponding to the current disconnecting link to the intelligent power grid dispatching control system at the main station end through the network;
step S2, the station side acquires a disconnecting link image through a camera, then analyzes the current disconnecting link state through a method based on a deep neural network and deployed on an intelligent analysis device, and uploads the disconnecting link image and the current disconnecting link state obtained through analysis to a unified video monitoring platform of the main station side through a network;
step S3, comparing the switch state of the disconnecting link obtained by the intelligent power grid dispatching control system with the current disconnecting link state obtained by the unified video monitoring platform, if the two states are consistent, returning the current state to the intelligent power grid dispatching control system and displaying; and if the two states are not consistent, manually checking the disconnecting link image, confirming the state of the disconnecting link, and returning the state of the disconnecting link to the intelligent power grid dispatching control system for displaying.
2. The method for detecting the state of the switch facing the scheduling service scenario according to claim 1, wherein: the method for using the deep neural network in the step S2 includes the following steps:
step 1, collecting a power inspection video image of power equipment; the power equipment comprises a single-arm folding type isolating switch, a double-column rotary type isolating switch and a three-column type isolating switch; a power patrol video image at least comprises a power device;
step 2, labeling the power patrol video image, wherein the labeling information comprises a target frame and a category label;
step 3, preparing a training data set and a testing data set before training the target detection network model; randomly dividing a training data set and a testing data set;
step 4, preprocessing the training data set; normalizing the size, and acquiring target images at different angles by adopting a data augmentation mode;
step 7, training a neural network to obtain a target detection network model;
step 8, testing the detection performance of the target detection network model through the test data set, and if the detection performance requirement is not met, continuing training on the training data set until the requirement is met; the detection performance takes the mAP value as a measurement index, and takes 0.6 as a qualified standard.
3. The method for detecting the state of the switch facing the scheduling service scenario according to claim 2, wherein: the target detection network model is a YOLOV3 target detection network model.
4. The method for detecting the state of the switch facing the scheduling service scenario according to claim 2, wherein: the size normalization is a uniform normalization of the images in the training dataset to 300 x 300 pixels.
5. The method for detecting the state of the switch facing the scheduling service scenario according to claim 2, wherein: the data augmentation is by rotating the images in the training dataset 90 °, 180 ° and 270 ° clockwise.
6. The method for detecting the state of the switch facing the scheduling service scenario according to claim 1, wherein: the intelligent power grid dispatching control system is a D5000 system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011097719.3A CN114429421A (en) | 2020-10-14 | 2020-10-14 | Method for detecting state of disconnecting link facing scheduling service scene |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011097719.3A CN114429421A (en) | 2020-10-14 | 2020-10-14 | Method for detecting state of disconnecting link facing scheduling service scene |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114429421A true CN114429421A (en) | 2022-05-03 |
Family
ID=81309235
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011097719.3A Pending CN114429421A (en) | 2020-10-14 | 2020-10-14 | Method for detecting state of disconnecting link facing scheduling service scene |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114429421A (en) |
-
2020
- 2020-10-14 CN CN202011097719.3A patent/CN114429421A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110826538B (en) | Abnormal off-duty identification system for electric power business hall | |
CN108416294B (en) | Fan blade fault intelligent identification method based on deep learning | |
CN111179249A (en) | Power equipment detection method and device based on deep convolutional neural network | |
CN110738127A (en) | Helmet identification method based on unsupervised deep learning neural network algorithm | |
CN108680833B (en) | Composite insulator defect detection system based on unmanned aerial vehicle | |
CN113205039B (en) | Power equipment fault image recognition disaster investigation system and method based on multiple DCNN networks | |
CN107133592B (en) | Human body target feature detection algorithm for power substation by fusing infrared thermal imaging and visible light imaging technologies | |
CN108711148A (en) | A kind of wheel tyre defect intelligent detecting method based on deep learning | |
CN110096945B (en) | Indoor monitoring video key frame real-time extraction method based on machine learning | |
CN110705414A (en) | Power transmission line construction machinery hidden danger detection method based on deep learning | |
CN113343998A (en) | Reading monitoring system and method for electric power mechanical meter, computer equipment and application | |
CN112508911A (en) | Rail joint touch net suspension support component crack detection system based on inspection robot and detection method thereof | |
CN112614130A (en) | Unmanned aerial vehicle power transmission line insulator fault detection method based on 5G transmission and YOLOv3 | |
CN115147591A (en) | Transformer equipment infrared image voltage heating type defect diagnosis method and system | |
CN117114420B (en) | Image recognition-based industrial and trade safety accident risk management and control system and method | |
Manninen et al. | Multi-stage deep learning networks for automated assessment of electricity transmission infrastructure using fly-by images | |
CN114694130A (en) | Method and device for detecting telegraph poles and pole numbers along railway based on deep learning | |
CN112260402B (en) | Monitoring method for state of intelligent substation inspection robot based on video monitoring | |
CN106355187A (en) | Application of visual information to electrical equipment monitoring | |
CN110969081B (en) | Power transmission line external force damage detection method based on KL divergence of multi-module division | |
CN117252459A (en) | Fruit quality evaluation system based on deep learning | |
CN108573233B (en) | Power grid ceramic insulator identification method based on image processing technology | |
CN113724247B (en) | Intelligent substation inspection method based on image discrimination technology | |
CN114429421A (en) | Method for detecting state of disconnecting link facing scheduling service scene | |
CN111028289B (en) | Method for positioning foreign matters in equipment in transformer substation based on template matching |
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 |