CN118134933B - SF6 leakage quantitative detection method and system for sensing state of power transformation equipment - Google Patents
SF6 leakage quantitative detection method and system for sensing state of power transformation equipment Download PDFInfo
- Publication number
- CN118134933B CN118134933B CN202410559119.6A CN202410559119A CN118134933B CN 118134933 B CN118134933 B CN 118134933B CN 202410559119 A CN202410559119 A CN 202410559119A CN 118134933 B CN118134933 B CN 118134933B
- Authority
- CN
- China
- Prior art keywords
- leakage
- image
- gas cloud
- quantitative detection
- frame
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 62
- 230000009466 transformation Effects 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000002372 labelling Methods 0.000 claims abstract description 26
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 11
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims description 22
- 238000003709 image segmentation Methods 0.000 claims description 19
- 230000011218 segmentation Effects 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 6
- 238000009792 diffusion process Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 5
- 230000004927 fusion Effects 0.000 claims description 5
- 230000007246 mechanism Effects 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 4
- 238000000926 separation method Methods 0.000 claims description 4
- 229910018503 SF6 Inorganic materials 0.000 description 128
- SFZCNBIFKDRMGX-UHFFFAOYSA-N sulfur hexafluoride Chemical compound FS(F)(F)(F)(F)F SFZCNBIFKDRMGX-UHFFFAOYSA-N 0.000 description 128
- 239000007789 gas Substances 0.000 description 103
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000004445 quantitative analysis Methods 0.000 description 4
- 230000001502 supplementing effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000000243 solution Substances 0.000 description 3
- 229960000909 sulfur hexafluoride Drugs 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000003331 infrared imaging Methods 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- 238000001931 thermography Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 240000004282 Grewia occidentalis Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 239000005431 greenhouse gas Substances 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009965 odorless effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000007789 sealing Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- 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
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Examining Or Testing Airtightness (AREA)
Abstract
The invention discloses a quantitative SF6 leakage detection method and a quantitative SF6 leakage detection system for sensing the state of power transformation equipment, wherein the method comprises the steps of acquiring infrared video images of the power transformation equipment and performing multi-frame synthesis pretreatment; SF6 leakage gas cloud cluster detection and rectangular labeling are carried out on the preprocessed infrared video image; dividing the rectangular labeling area image to separate SF6 leakage gas cloud cluster images; and determining SF6 leakage points, a gas cloud center and cloud edges according to the SF6 leakage gas cloud image, measuring SF6 concentration values of the SF6 leakage points, the gas cloud center and the cloud edges and on-site environment data, and further calculating to obtain SF6 leakage rate. The SF6 leakage point detection method not only can automatically detect the SF6 leakage point position, but also combines the SF6 leakage gas cloud cluster shape with the gas concentration at the leakage position, quantitatively analyzes the SF6 leakage rate, and provides quantitative reference for the state sensing of the follow-up power transformation equipment.
Description
Technical Field
The invention belongs to the technical field of SF6 gas leakage detection of power equipment, and particularly relates to a quantitative SF6 leakage detection method and system for sensing the state of power transformation equipment.
Background
SF6 is a colorless and odorless gas that is widely used in the power field due to its excellent insulation and arc extinguishing properties. While SF6 leakage due to sealing problems occurs during operation of the electrical equipment. When SF6 pressure is reduced, the insulation level of the equipment is reduced, the arc extinguishing performance of the switch equipment is reduced, and the operation safety of the power equipment is seriously affected. In addition, since the SF6 gas has a high density, the leaked SF6 in the indoor environment is accumulated toward the bottom, which may be harmful to the human body. SF6 gas is also a very damaging greenhouse gas, and environmental problems caused by SF6 leakage are not small. Therefore, when the SF6 leakage problem occurs in the power equipment, the leakage point is timely and accurately found, and corresponding treatment measures are adopted according to the gas leakage rate, so that the method has great significance in guaranteeing the safety of the equipment and personnel.
In recent years, due to the advantages of far distance, non-contact, dynamic intuitiveness and the like of an infrared thermal imaging technology, SF6 leakage detection based on the infrared thermal imaging technology is widely applied in the electric field, but various challenges still exist, such as a scheme for SF6 leakage detection in an infrared imaging mode, an image comparison method is mostly adopted, SF6 leakage points are marked through changes among different video frames, the SF6 leakage points are analyzed through changes among the video frames, a certain effect is achieved when the gas leakage amount is large and the injection speed is high, the method is not suitable for the situation of slow gas leakage, and for a handheld infrared imager, the video frame comparison algorithm effect is poor due to hand shake; the image enhancement algorithm is adopted, so that SF6 gas is more obvious in the image, and an operator can find leakage points conveniently, namely the currently widely adopted SF6 image enhancement algorithm still needs to analyze and judge the infrared video image manually, and the operation of the operator cannot be reduced effectively. In addition, the infrared imaging technology can only perform qualitative analysis on SF6 gas leakage, and cannot quantitatively analyze the gas leakage condition, so that the accurate quantification basis is lacking when the equipment leakage defect is treated. In the aspect of quantitative detection of SF6 gas leakage, the currently widely used handheld leak detectors can only measure the concentration of SF6 at the leakage position, and cannot accurately analyze the gas leakage rate. According to industry regulations, the annual leakage rate of sulfur hexafluoride of electrical equipment must not exceed 0.5%, and the leakage rate must be correspondingly treated once exceeding a standard value. The traditional quantitative detection method comprises a buckling cover method, a bottle hanging method, a local binding method, a pressure drop method and the like, and the methods basically need tens of hours, so that time and labor are wasted, and part of the positions can be detected only by power failure. Therefore, how the leakage rate of sulfur hexafluoride is judged through the interval time of two air supplementing, and the interval time of two air supplementing is generally less than half a year, so that the equipment is evaluated as an abnormal state, corresponding maintenance is needed, and the leaked sulfur hexafluoride in the interval time of two air supplementing can possibly damage the environment and personnel. Therefore, an immediate quantitative detection method is needed to make an overhaul decision and assist in time.
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides the quantitative detection method and the quantitative detection system for SF6 leakage for sensing the state of the power transformation equipment, which not only can automatically detect the position of the SF6 leakage point, but also combine the shape of SF6 leakage gas cloud cluster with the gas concentration at the leakage position, quantitatively analyze the SF6 leakage rate and provide quantitative reference for the state sensing of the follow-up power transformation equipment.
The invention adopts the following technical scheme.
An SF6 leakage quantitative detection method for state sensing of power transformation equipment, comprising:
step 1, acquiring an infrared video image of power transformation equipment and performing multi-frame synthesis pretreatment;
step 2, SF6 leakage gas cloud cluster detection and rectangular labeling are carried out on the preprocessed infrared video image;
Step3, dividing the rectangular labeling area image, and separating SF6 leakage gas cloud cluster images;
And 4, determining SF6 leakage points, a gas cloud center and cloud edges according to the SF6 leakage gas cloud image, measuring SF6 concentration values of the SF6 leakage points, the gas cloud center and the cloud edges and on-site environment data, and further calculating to obtain SF6 leakage rate.
Preferably, in step 1, the manner of performing the multi-frame synthesis preprocessing is as follows: extracting continuous three-frame images from infrared video of power transformation equipment, taking the continuous three-frame images as a front frame, a reference frame and a rear frame, and calculating inter-frame SAD; and calculating the weight of each frame of data according to the inter-frame SAD, and carrying out multi-frame image weighted fusion to generate an image infrared video image after noise reduction, namely the preprocessed infrared video image.
Preferably, in step2, SF6 leakage gas cloud detection is performed on the preprocessed infrared video image, and a rectangular label of the gas cloud is obtained on the infrared video image, which is specifically as follows:
Carrying out repeated downsampling on the preprocessed infrared video image by adopting Darknet-53 networks with a plurality of stacked residual modules;
constructing an FPN feature pyramid, and carrying out enhanced feature extraction on the gas cloud cluster in the image after multiple downsampling to obtain a gas cloud cluster boundary box;
And obtaining a prediction result of the gas cloud image by utilizing the YOLO Head, taking a gas cloud bounding box as an image segmentation limit of the gas cloud rectangular label, and carrying out the gas cloud rectangular label on the prediction result.
Preferably, step 3 specifically includes: and (3) carrying out image pixel area calculation and segmentation condition judgment on the rectangular labeling area image, and segmenting the rectangular labeling area image by adopting a pre-trained U2-Net image segmentation model if the segmentation condition is met, so as to separate SF6 leakage gas cloud cluster images, otherwise, returning to the step (1).
Preferably, the image pixel area calculation formula is: ; wherein m and n are the number of pixels in the length direction and the width direction of the rectangular labeling area image respectively;
The segmentation conditions are as follows: ; wherein the method comprises the steps of Is a threshold value.
Preferably, a pre-trained U2-Net image segmentation model is adopted to segment the rectangular labeling area image, and the specific steps of separating SF6 leakage gas cloud cluster images are as follows:
The pre-trained U2-Net image segmentation model adopts an encoder to extract image characteristics of a rectangular marked gas cloud cluster region in an infrared video image;
The decoder is adopted to map the extracted image characteristics back to the rectangular marked gas cloud cluster area;
and capturing the details and the edge information of the original image by using a self-attention mechanism, and separating the gas cloud clusters marked by the rectangles in the infrared video image.
Preferably, in the step 4, the highest point of the SF6 concentration value in the gas cloud cluster image is taken as an SF6 leakage point;
The gas cloud center was determined according to the following formula ：
、，
Wherein,The horizontal and vertical coordinates of the center of the gas cloud;
n is the total number of pixel points in the gas cloud;
x i、yi is the horizontal and vertical coordinates of the ith pixel point;
w i is the weight of the ith pixel point.
Preferably, the method comprises the steps of,，
，
，
Wherein,Is the equilibrium coefficient; Is a brightness weight; Is the brightness value of the pixel point; Is an adjustment coefficient; Is the spatial density weight; is the number of adjacent pixels within the set radius, Is a threshold.
Preferably, in step 4, measuring SF6 concentration values of SF6 leakage points, gas cloud center and cloud edges by using a negative corona discharge sensor, wherein the SF6 concentration values of the cloud edges are determined by taking the average value of a plurality of edge point concentration values; the field environment data is wind speed.
Preferably, in step 4, the SF6 leakage rate is calculated as:
，
Wherein, (L= L, C, E) represents the SF6 concentration value; q is SF6 leakage rate; σ y and σ z are diffusion parameters; u is wind speed; x, y, z are the horizontal, vertical and vertical coordinates of the video image acquisition point relative to the leakage point; h is the height of the leak;
and substituting SF6 concentration values C L,CC,CE of the SF6 leakage point, the gas cloud center and the cloud edge into the calculation formula respectively to obtain SF6 leakage rates Q, sigma y and sigma z, wherein Q is the SF6 leakage rate obtained through quantitative detection.
An SF6 leak quantitative detection system for power transformation equipment status awareness, comprising:
The image acquisition module is used for acquiring the infrared video image of the power transformation equipment and carrying out multi-frame synthesis pretreatment;
the gas cloud cluster labeling module is used for carrying out SF6 leakage gas cloud cluster detection and rectangular labeling on the preprocessed infrared video image;
the gas cloud cluster separation module is used for dividing the rectangular marked area image and separating SF6 leakage gas cloud cluster images;
And the SF6 leakage rate calculation module is used for determining SF6 leakage points, gas cloud centers and cloud edges according to the SF6 leakage gas cloud cluster images, measuring SF6 concentration values of the SF6 leakage points, the gas cloud centers and the cloud edges and on-site environment data, and further calculating to obtain the SF6 leakage rate.
A terminal comprising a processor and a storage medium; the storage medium is used for storing instructions; the processor is configured to operate in accordance with the instructions to perform the steps of the method.
A computer readable storage medium having stored thereon a computer program which when executed by a processor realizes the steps of the method.
The invention has the beneficial effects that compared with the prior art:
1. in the aspect of quantitative analysis of SF6 leakage, the invention provides accurate quantitative analysis of SF6 leakage rate for the first time, so as to make overhaul decision assistance in time;
2. The invention realizes the extraction of key information of multi-frame images based on the inter-frame SAD, automatically eliminates noise points, synthesizes the final infrared video image by utilizing the multi-frame images, effectively improves the image quality and can effectively reduce the subsequent calculation consumption;
3. According to the invention, the infrared video image is detected in real time, the leaked gas cloud cluster is automatically marked in the image in a frame selection way, and the SF6 leakage position is dynamically and intuitively marked on the image, so that an operator can conveniently and rapidly and accurately find the SF6 leakage point, the workload of the operator is reduced, and the accuracy and the detection efficiency of SF6 leakage point detection are improved;
4. according to the invention, through image segmentation of the SF6 gas cloud, and combining with a calculation formula of the center of the gas cloud and the SF6 leakage rate, quantitative calculation analysis of the SF6 gas leakage rate is realized through multipoint SF6 concentration values and environmental data at the leakage points measured by the handheld negative corona discharge sensor, and data support is provided for subsequent maintenance. The self-attention mechanism is introduced into the segmentation model, so that the model has higher capability of capturing details and edge information of an original image, thereby realizing accurate segmentation of SF6 leakage gas cloud cluster images, and calculating SF6 leakage rate according to SF6 leakage points, the centers of the gas cloud clusters, SF6 concentration values of cloud cluster edges and on-site environment data.
Drawings
Fig. 1 is a flowchart of a quantitative detection method for SF6 leakage for sensing the state of power transformation equipment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. The described embodiments of the application are only some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art without making any inventive effort, are within the scope of the present application.
As shown in fig. 1, embodiment 1 of the present invention provides a quantitative detection method for SF6 leakage for sensing the state of power transformation equipment, which includes the following steps:
step 1, acquiring an infrared video image of power transformation equipment and performing multi-frame synthesis pretreatment;
And shooting positions such as bolts, flanges, air supplementing interfaces and the like, which possibly have SF6 leakage, by adopting a fixed SF6 gas detection infrared imager or a handheld SF6 gas detection infrared imager arranged in the transformer substation, acquiring infrared video images, carrying out multi-frame synthesis pretreatment, and storing the pretreated infrared video images of equipment at the positions.
Considering that the original video has higher frame rate, the target detection calculation amount is large directly, the original video is processed by multi-frame synthesis, and the image quality is improved. Specifically, the noise reduction algorithm can be adopted to reduce the original video frame rate to 1 frame of image per second, and the processing steps are as follows:
Step 1.1, extracting continuous three-frame images from infrared video of power transformation equipment, taking the continuous three-frame images as a front frame, a reference frame and a rear frame, and calculating inter-frame SAD;
the inter SAD calculation method comprises the following steps: and comparing gray value differences of corresponding pixel points between the continuous frames, and calculating the sum of absolute values of the gray value differences to obtain an inter-frame SAD value. The SAD value can effectively reflect the variation degree between frames and provide key data for the weight determination of the subsequent frames;
gradient analysis can also be performed on reference frames, preceding frames, following frames: by calculating the horizontal and vertical gradients of pixels in each frame and averaging, the dynamic changes of SF6 gas leakage are accurately identified and tracked. The process utilizes the gradient result to obviously highlight the motion trail of the gas cloud cluster, and provides basis for quantitative detection.
And 1.2, calculating the weight of each frame of data according to the inter-frame SAD, and carrying out multi-frame image weighted fusion to generate an image infrared video image after noise reduction, namely the preprocessed infrared video image.
Determining frame weights from the inter-frame SAD: weights are assigned based on the SAD value between each frame and the reference frame. Frames with lower SAD values, i.e. frames with less image variation and more stable information, will be given higher weights.
Multi-frame image composition application: and carrying out weighted fusion on the multi-frame images by using the calculated frame weights to generate the noise-reduced infrared video images. The method improves the accuracy of leakage detection and the image quality by optimizing frames with large information quantity and small noise.
Step 2, SF6 leakage gas cloud cluster detection and rectangular labeling are carried out on the preprocessed infrared video image;
SF6 leakage gas cloud cluster detection is carried out on the preprocessed infrared video image by adopting a Yolo v-based target detection model, and gas cloud cluster rectangular labels are obtained on the infrared video image picture;
yolo v3 is a commonly used image target detection algorithm, and the target detection model based on Yolo v is trained by using data containing the leaked gas cloud cluster marks in the model training process, so that the model can obtain the capability of detecting the leaked gas cloud cluster. And inputting each frame of image in the video into a trained model, automatically marking the leaked gas cloud cluster (in a rectangular frame form) in the image based on a Yolo v target detection model, and outputting the model as four-corner vertex coordinates of a marked rectangular frame.
Further, SF6 leakage gas cloud cluster detection is carried out on the preprocessed infrared video image by adopting a target detection model based on Yolo v, and gas cloud cluster rectangular labels are obtained on the infrared video image, and the method comprises the following specific steps of:
Step 2.1, adopting Darknet-53 networks with a plurality of residual modules stacked, and downsampling the preprocessed infrared video image for a plurality of times;
2.2, constructing an FPN feature pyramid, and carrying out enhanced feature extraction on the gas cloud cluster in the image after multiple downsampling to obtain a gas cloud cluster boundary box;
and 2.3, obtaining a prediction result of the gas cloud image by utilizing the YOLO Head, taking a gas cloud bounding box as an image segmentation limit of the gas cloud rectangular label, and carrying out the gas cloud rectangular label on the prediction result.
Step3, dividing the rectangular labeling area image, and separating SF6 leakage gas cloud cluster images;
Taking the rectangular area image obtained in the step 2 as input, and separating the leaked gas cloud cluster from the image by using a U2-Net image segmentation algorithm, wherein the method specifically comprises the following steps of:
step 3.1, carrying out image pixel area calculation and segmentation condition judgment on the rectangular labeling area image, if the segmentation condition is met, entering step 3.2, otherwise, returning to step 1;
Before image segmentation, the pixel ratio of the target rectangle in the whole picture is ensured to exceed a certain threshold, so that the gas cloud cluster to be segmented is ensured to completely appear and occupy the main body position of the picture, and the picture segmentation effect and the accuracy of subsequent calculation are ensured. Specific:
The image pixel area calculation formula is: ; wherein m and n are the number of pixels in the length direction and the width direction of the rectangular labeling area image respectively;
The segmentation conditions are as follows: ; wherein the method comprises the steps of Is a threshold value.
That is, the number of pixels in the length direction of the image in the rectangular labeling area is m, the number of pixels in the width direction is n, and the threshold value isWhen thenAnd (3) if not, prompting the user to move the near infrared imager to shoot again.
And 3.2, segmenting the rectangular labeling area image by adopting a pre-trained U2-Net image segmentation model, and separating out SF6 leakage gas cloud cluster images.
Inputting the rectangular marked area image meeting the segmentation condition into a pre-trained U2-Net image segmentation model (namely the pre-trained U2-Net image segmentation model adopted in the process is subjected to gas cloud cluster segmentation task), wherein the output result of the U2-Net image segmentation model is a binary mask matrix with the same length and width pixel number as that of the original imageWherein the pixels corresponding to 1 in the mask matrix are the gas cloud image, and the pixels corresponding to 0 are the background image. The specific separation steps are as follows:
The pre-trained U2-Net image segmentation model adopts an encoder to extract image characteristics of a rectangular marked gas cloud cluster region in an infrared video image;
The decoder is adopted to map the extracted image characteristics back to the rectangular marked gas cloud cluster area;
and capturing the details and the edge information of the original image by using a self-attention mechanism, and separating the gas cloud clusters marked by the rectangles in the infrared video image.
And 4, determining SF6 leakage points, a gas cloud center and cloud edges according to the SF6 leakage gas cloud image, measuring SF6 concentration values of the SF6 leakage points, the gas cloud center and the cloud edges and on-site environment data, and further calculating to obtain SF6 leakage rate.
Taking the point with the highest SF6 concentration value (namely the point with the deepest color) in the gas cloud cluster image as an SF6 leakage point;
determining the center coordinates of the gas cloud according to the following formula ：
、
Wherein,The horizontal and vertical coordinates of the center of the gas cloud;
n is the total number of pixel points in the gas cloud;
x i、yi is the horizontal and vertical coordinates of the ith pixel point;
w i is the weight of the ith pixel point, and is determined by the brightness and density of the pixel.
Weighting of pixel pointsThe calculation may be performed as follows:
Wherein, the method comprises the steps of, wherein, Is a coefficient for balancing luminance weight and spatial density weight;
For brightness weight, converting brightness value of pixel point into weight:
，
Wherein the method comprises the steps of Is the luminance value of the pixel point,Is an adjustment coefficient; for spatial density weights, the number of neighboring pixels around a pixel point is calculated:
，
Wherein the method comprises the steps of Is the number of adjacent pixels within the set radius,Is a threshold. Based on the analysis, the SF6 gas leakage rate is calculated, and the method specifically comprises the following steps.
And 4.1, respectively measuring SF6 concentration values of the leakage points, the center of the SF6 gas cloud and the edge of the cloud by using a handheld negative corona discharge sensor. The method comprises the following steps: collecting SF6 concentration value (C L) at SF6 leakage point (L) with unit of ppm; SF6 concentration values (C C and C E) were collected at the center (C) and edge (E) of the gas cloud, respectively. The SF6 concentration value C E at the edge of the cloud cluster can be determined by taking the average value of the concentration value of any point at the edge or the concentration values of a plurality of edge points according to the actual situation.
Wind speed is measured in meters per second (m/s) using the relevant instrument.
Step 4.2, SF6 leakage in the transformer substation can be approximately regarded as continuous point source type diffusion in an open and flat area and under uniform atmospheric conditions, and SF6 leakage rate modeling is conducted on gas leakage conditions. The image after image segmentation is approximately regarded as an equal concentration plane, and the gas diffusion flow speed can be reversely deduced by the wind speed and the multi-point SF6 concentration value.
The SF6 leak rate calculation model is:
，
Wherein, (L= L, C, E) represents the SF6 concentration value; q is unknown SF6 leakage rate; σ y and σ z are diffusion parameters, related to environmental conditions; u is wind speed; x, y, z are the coordinates of the video image acquisition point relative to the leak point; h is the height of the leak.
And substituting the measured SF6 concentration value C L,CC,CE into the model respectively, and solving to obtain SF6 leakage rates Q, sigma y and sigma z, wherein Q is the SF6 leakage rate obtained by quantitative detection.
The embodiment 2 of the invention provides an SF6 leakage quantitative detection system for sensing the state of power transformation equipment, which comprises the following components:
The image acquisition module is used for acquiring the infrared video image of the power transformation equipment and carrying out multi-frame synthesis pretreatment;
the gas cloud cluster labeling module is used for carrying out SF6 leakage gas cloud cluster detection and rectangular labeling on the preprocessed infrared video image;
the gas cloud cluster separation module is used for dividing the rectangular marked area image and separating SF6 leakage gas cloud cluster images;
And the SF6 leakage rate calculation module is used for determining SF6 leakage points, gas cloud centers and cloud edges according to the SF6 leakage gas cloud cluster images, measuring SF6 concentration values of the SF6 leakage points, the gas cloud centers and the cloud edges and on-site environment data, and further calculating to obtain the SF6 leakage rate.
A terminal comprising a processor and a storage medium; the storage medium is used for storing instructions; the processor is operative to perform steps according to the method in accordance with the instructions.
A computer readable storage medium having stored thereon a computer program which when executed by a processor realizes the steps of the method.
The invention has the beneficial effects that compared with the prior art:
1. in the aspect of quantitative analysis of SF6 leakage, the invention provides accurate quantitative analysis of SF6 leakage rate for the first time, so as to make overhaul decision assistance in time;
2. The invention realizes the extraction of key information of multi-frame images based on the inter-frame SAD, automatically eliminates noise points, synthesizes the final infrared video image by utilizing the multi-frame images, effectively improves the image quality and can effectively reduce the subsequent calculation consumption;
3. According to the invention, the infrared video image is detected in real time, the leaked gas cloud cluster is automatically marked in the image in a frame selection way, and the SF6 leakage position is dynamically and intuitively marked on the image, so that an operator can conveniently and rapidly and accurately find the SF6 leakage point, the workload of the operator is reduced, and the accuracy and the detection efficiency of SF6 leakage point detection are improved;
4. according to the invention, through image segmentation of the SF6 gas cloud, and combining with a calculation formula of the center of the gas cloud and the SF6 leakage rate, quantitative calculation analysis of the SF6 gas leakage rate is realized through multipoint SF6 concentration values and environmental data at the leakage points measured by the handheld negative corona discharge sensor, and data support is provided for subsequent maintenance. The self-attention mechanism is introduced into the segmentation model, so that the model has higher capability of capturing details and edge information of an original image, thereby realizing accurate segmentation of SF6 leakage gas cloud cluster images, and calculating SF6 leakage rate according to SF6 leakage points, the centers of the gas cloud clusters, SF6 concentration values of cloud cluster edges and on-site environment data.
The present disclosure may be a system, method, and/or computer program product. The computer program product may include a computer readable storage medium having computer readable program instructions embodied thereon for causing a processor to implement aspects of the present disclosure.
The computer readable storage medium may be a tangible device that can hold and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: portable computer disks, hard disks, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static Random Access Memory (SRAM), portable compact disk read-only memory (CD-ROM), digital Versatile Disks (DVD), memory sticks, floppy disks, mechanical coding devices, punch cards or in-groove structures such as punch cards or grooves having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media, as used herein, are not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., optical pulses through fiber optic cables), or electrical signals transmitted through wires.
The computer readable program instructions described herein may be downloaded from a computer readable storage medium to a respective computing/processing device or to an external computer or external storage device over a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers and/or edge servers. The network interface card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium in the respective computing/processing device.
The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as SMALLTALK, C ++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer readable program instructions may be executed entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, aspects of the present disclosure are implemented by personalizing electronic circuitry, such as programmable logic circuitry, field Programmable Gate Arrays (FPGAs), or Programmable Logic Arrays (PLAs), with state information of computer readable program instructions, which can execute the computer readable program instructions.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that: modifications and equivalents may be made to the specific embodiments of the invention without departing from the spirit and scope of the invention, which is intended to be covered by the claims.
Claims (11)
1. A quantitative detection method for SF6 leakage for sensing the state of a power transformation device, the method comprising:
Step 1, acquiring infrared video images of power transformation equipment, calculating the weight of each frame of data according to the inter-frame SAD, and carrying out multi-frame image weighted fusion to realize multi-frame synthesis pretreatment;
step 2, comprehensively utilizing Darknet-53 networks, FPN characteristic pyramids and YOLO Head to detect SF6 leakage gas cloud clusters and label rectangles on the preprocessed infrared video images;
Step 3, calculating the image pixel area and judging the segmentation condition of the rectangular labeling area image, if the segmentation condition is met, segmenting the rectangular labeling area image, separating SF6 leakage gas cloud cluster images, otherwise, returning to the step 1;
Step 4, determining SF6 leakage points, a gas cloud center and cloud edges according to SF6 leakage gas cloud images, measuring SF6 concentration values of the SF6 leakage points, the gas cloud center and the cloud edges and on-site environment data, and further calculating to obtain SF6 leakage rate;
The gas cloud center was determined according to the following formula ：
Wherein,The horizontal and vertical coordinates of the center of the gas cloud;
n is the total number of pixel points in the gas cloud;
x i、yi is the horizontal and vertical coordinates of the ith pixel point;
w i is the weight of the ith pixel point;
Wherein, Is the equilibrium coefficient; Is a brightness weight; Is the brightness value of the pixel point; Is an adjustment coefficient; Is the spatial density weight; Is the number of adjacent pixels within the set radius, Is a threshold;
The SF6 leak rate is calculated as:
Wherein, (L= L, C, E) represents the SF6 concentration value; q is SF6 leakage rate; And Is a diffusion parameter; u is wind speed; x, y, z are the horizontal, vertical and vertical coordinates of the video image acquisition point relative to the leakage point; h is the height of the leak;
Substituting SF6 concentration values C L,CC,CE of SF6 leakage points, the centers of gas cloud clusters and the edges of the cloud clusters into the calculation formula respectively, and solving to obtain SF6 leakage rate Q, AndWherein Q is the SF6 leakage rate obtained by quantitative detection.
2. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
in step 1, the mode of performing multi-frame synthesis pretreatment is as follows: extracting continuous three-frame images from infrared video of power transformation equipment, taking the continuous three-frame images as a front frame, a reference frame and a rear frame, and calculating inter-frame SAD; and calculating the weight of each frame of data according to the inter-frame SAD, and carrying out multi-frame image weighted fusion to generate an image infrared video image after noise reduction, namely the preprocessed infrared video image.
3. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
in the step 2, SF6 leakage gas cloud cluster detection is carried out on the preprocessed infrared video image, and rectangular labels of the gas cloud clusters are obtained on the infrared video image, wherein the method comprises the following steps of:
Carrying out repeated downsampling on the preprocessed infrared video image by adopting Darknet-53 networks with a plurality of stacked residual modules;
constructing an FPN feature pyramid, and carrying out enhanced feature extraction on the gas cloud cluster in the image after multiple downsampling to obtain a gas cloud cluster boundary box;
And obtaining a prediction result of the gas cloud image by utilizing the YOLO Head, taking a gas cloud bounding box as an image segmentation limit of the gas cloud rectangular label, and carrying out the gas cloud rectangular label on the prediction result.
4. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
The step 3 specifically comprises the following steps: and (3) carrying out image pixel area calculation and segmentation condition judgment on the rectangular labeling area image, and segmenting the rectangular labeling area image by adopting a pre-trained U2-Net image segmentation model if the segmentation condition is met, so as to separate SF6 leakage gas cloud cluster images, otherwise, returning to the step (1).
5. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 4, wherein the quantitative detection method comprises the following steps:
The image pixel area calculation formula is: ; wherein m and n are the number of pixels in the length direction and the width direction of the rectangular labeling area image respectively;
The segmentation conditions are as follows: ; wherein the method comprises the steps of Is a threshold value.
6. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
The method for separating SF6 leakage gas cloud cluster images by segmenting rectangular labeling area images by using a pre-trained U2-Net image segmentation model comprises the following specific steps:
The pre-trained U2-Net image segmentation model adopts an encoder to extract image characteristics of a rectangular marked gas cloud cluster region in an infrared video image;
The decoder is adopted to map the extracted image characteristics back to the rectangular marked gas cloud cluster area;
and capturing the details and the edge information of the original image by using a self-attention mechanism, and separating the gas cloud clusters marked by the rectangles in the infrared video image.
7. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
And 4, taking the highest point of the SF6 concentration value in the gas cloud cluster image as an SF6 leakage point.
8. The quantitative detection method for SF6 leakage for sensing the state of power transformation equipment according to claim 1, wherein the quantitative detection method comprises the following steps:
In the step 4, SF6 concentration values of SF6 leakage points, gas cloud cluster centers and cloud cluster edges are measured through a negative corona discharge sensor, wherein the SF6 concentration values of the cloud cluster edges are determined in a mode of taking the average value of a plurality of edge point concentration values; the field environment data is wind speed.
9. A quantitative detection system for SF6 leakage for sensing the status of a power transformation device, using the method of any of claims 1-8, characterized in that: the system comprises:
The image acquisition module is used for acquiring the infrared video image of the power transformation equipment and carrying out multi-frame synthesis pretreatment;
the gas cloud cluster labeling module is used for carrying out SF6 leakage gas cloud cluster detection and rectangular labeling on the preprocessed infrared video image;
the gas cloud cluster separation module is used for dividing the rectangular marked area image and separating SF6 leakage gas cloud cluster images;
And the SF6 leakage rate calculation module is used for determining SF6 leakage points, gas cloud centers and cloud edges according to the SF6 leakage gas cloud cluster images, measuring SF6 concentration values of the SF6 leakage points, the gas cloud centers and the cloud edges and on-site environment data, and further calculating to obtain the SF6 leakage rate.
10. A terminal comprising a processor and a storage medium; the method is characterized in that: the storage medium is used for storing instructions; the processor being operative according to the instructions to perform the steps of the method according to any one of claims 1-8.
11. Computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410559119.6A CN118134933B (en) | 2024-05-08 | 2024-05-08 | SF6 leakage quantitative detection method and system for sensing state of power transformation equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410559119.6A CN118134933B (en) | 2024-05-08 | 2024-05-08 | SF6 leakage quantitative detection method and system for sensing state of power transformation equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118134933A CN118134933A (en) | 2024-06-04 |
| CN118134933B true CN118134933B (en) | 2024-08-30 |
Family
ID=91238176
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410559119.6A Active CN118134933B (en) | 2024-05-08 | 2024-05-08 | SF6 leakage quantitative detection method and system for sensing state of power transformation equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118134933B (en) |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102609906A (en) * | 2012-01-12 | 2012-07-25 | 北京理工大学 | Gas infrared image enhancing method based on anisotropic diffusion |
| WO2018231735A1 (en) * | 2017-06-12 | 2018-12-20 | Flir Systems Ab | System and method for quantifying a gas leak |
| EP3788331A4 (en) * | 2018-05-03 | 2022-02-23 | Quantum IR Technologies, LLC | Infrared imaging systems and methods for gas leak detection |
| CN110672278B (en) * | 2019-10-16 | 2021-04-30 | 北京工业大学 | Method for quantitatively measuring VOCs leakage of production device based on infrared imaging |
| CN111537157A (en) * | 2020-05-11 | 2020-08-14 | 嘉兴极光物联网科技有限公司 | Method and system for efficiently detecting gas leakage by visual qualitative and accurate quantitative cooperation |
| CN114112972B (en) * | 2021-12-02 | 2023-09-26 | 国网安徽省电力有限公司电力科学研究院 | Closed space SF 6 Gas leakage infrared telemetry device and imaging positioning method thereof |
| CN115909180B (en) * | 2023-01-04 | 2024-01-26 | 中环碳和(北京)科技有限公司 | Sulfur hexafluoride measuring method |
-
2024
- 2024-05-08 CN CN202410559119.6A patent/CN118134933B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| "被动式红外成像气体目标智能检测算法及量化研究进展";王琦等;《控制与决策》;20230831;第38卷(第8期);第2265-2282页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118134933A (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107256225B (en) | Method and device for generating heat map based on video analysis | |
| CN111583198A (en) | Insulator picture defect detection method combining FasterR-CNN + ResNet101+ FPN | |
| CN109712127B (en) | Power transmission line fault detection method for machine inspection video stream | |
| CN103605983B (en) | Remnant detection and tracking method | |
| CN113516124A (en) | Electric energy meter electricity consumption information identification algorithm based on computer vision technology | |
| CN113344475A (en) | Transformer bushing defect identification method and system based on sequence modal decomposition | |
| CN112989995B (en) | Text detection method and device and electronic equipment | |
| CN116229052B (en) | Method for detecting state change of substation equipment based on twin network | |
| CN109389569A (en) | Based on the real-time defogging method of monitor video for improving DehazeNet | |
| CN101996307A (en) | Intelligent video human body identification method | |
| CN114220009A (en) | Infrared image-based wire windage yaw identification method and system | |
| CN111145222A (en) | Fire detection method combining smoke movement trend and textural features | |
| CN115797846A (en) | Wind power generation blade block defect comparison method and device and electronic equipment | |
| CN117333776A (en) | VOCs gas leakage detection method, device and storage medium | |
| Yi et al. | Insulator and defect detection model based on improved YOLO-S | |
| CN111507398A (en) | Transformer substation metal instrument corrosion identification method based on target detection | |
| CN118134933B (en) | SF6 leakage quantitative detection method and system for sensing state of power transformation equipment | |
| CN117576632B (en) | Multi-mode AI large model-based power grid monitoring fire early warning system and method | |
| CN113191469A (en) | Logistics management method, system, server and storage medium based on two-dimension code | |
| CN118411724A (en) | Icing form segmentation method for overhead transmission line | |
| CN117333815A (en) | Method and system for identifying operators on marketing site | |
| CN113689399B (en) | Remote sensing image processing method and system for power grid identification | |
| Cui et al. | An exploratory framework to identify dust on photovoltaic panels in offshore floating solar power stations | |
| CN113705494A (en) | Dynamic detection method, system and device for braking performance of freight train | |
| CN118608876B (en) | GIS partial discharge mode recognition model and training method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |