Electrical equipment voltage heating type defect identification method and system based on thermal field change

Abstract

The invention belongs to the technical field of anomaly detection of electrical equipment, and particularly relates to a method and a system for identifying voltage-induced heat type defects of electrical equipment based on thermal field change. The method comprises the following steps: identifying electrical equipment in an infrared image based on a Faster RCNN algorithm, and determining a three-phase region of the electrical equipment; identifying each porcelain bushing in the infrared image based on a # imgabs0 # algorithm model; associating the porcelain bushings contained in the three-phase region into similar porcelain bushings; based on the infrared image temperature matrix, determining each temperature matrix of the same kind of porcelain bushings; respectively determining the maximum value of each porcelain bushing temperature matrix in the same kind of porcelain bushings; and based on the maximum value of each porcelain bushing temperature matrix, calculating a temperature difference value of the same kind of porcelain bushings, and based on a comparison relationship between the temperature difference value of the same kind of porcelain bushings and a preset standard, judging whether the current electrical equipment has voltage-induced heat type defect abnormity or not. The method can effectively and automatically identify the voltage-induced heat type defect from the infrared image.

Description

Technical Field

[0001] The present invention belongs to the technical field of abnormality detection of electric equipment, and in particular relates to a method and system for identifying voltage-induced thermal defects of electric equipment based on thermal field changes. Background Art

[0002] Thermal defects of electrical equipment are the main factors affecting the safe and stable operation of the power grid system, and will affect the entire temperature control work. Voltage-induced thermal defects of electrical equipment are mainly caused by the deterioration of the internal insulating medium or the heat caused by surface contamination. The temperature rise of the fault itself is small, and it is difficult to accurately analyze it in infrared detection. When voltage-induced thermal equipment shows temperature defects, the degree of defects is often more serious, which may cause equipment damage or even threaten personal safety. Therefore, it is urgent to propose an intelligent identification method for voltage-induced thermal defects of electrical equipment to improve the efficiency and quality of thermal defect diagnosis, thereby ensuring the safe and stable operation of power equipment and power grids.

[0003] At present, the main status and problems of thermal defect identification technology for electrical equipment at home and abroad are summarized as follows: 1) A structured positioning method based on the CenterNet model uses CenterNet combined with a structured positioning algorithm model. By collecting on-site infrared image data samples, training, and verifying the calculation of the algorithm model, different substation equipment and its components can be identified and located from complex infrared images with a high accuracy. According to the surface temperature range of equipment components and the type of substation equipment identified and located, the infrared image defect detection of power equipment is realized in combination with relevant temperature specifications. This method improves the accuracy of infrared image defect detection of power equipment, and completes the thermal type defect detection of electric heating devices in combination with temperature specifications. However, the defect recognition rate of this method is low.

[0004] 2) Infrared thermal image recognition method for substation equipment based on image enhancement and deep learning. Aiming at the characteristics of low contrast, blurred edges and complex background of substation infrared thermal image, a parameter self-adjusting Retinex infrared thermal image enhancement method combined with fast guided filtering is proposed to provide conditions for accurate recognition of infrared thermal image of substation equipment, and improve the feature extraction network and loss function of YOLOv3 convolutional neural network, solving the problem of low recognition accuracy of infrared thermal image of substation equipment. However, this method has the problem of high false alarm rate of defects.

[0005] 3) Infrared defect recognition method for substation equipment based on deep learning, using Faster RCNN algorithm to detect seven types of substation equipment, including transformers, bushings, circuit breakers, disconnectors, voltage transformers, current transformers, and lightning arresters, to achieve accurate positioning and identification of the equipment, and then using the temperature threshold discrimination method to identify thermal defects in the equipment area to determine whether the equipment has defects and the severity. The experiment verified its feasibility and accuracy. However, this method fails to consider the impact of low contrast of infrared images on the recognition results, and the two-stage model detection such as Faster RCNN takes a long time and cannot meet the real-time requirements.

[0006] At present, the technology for identifying thermal defects in electrical equipment is difficult to apply to scenarios where the electrical equipment background is complex, multiple devices have similar appearances, and defects manifest in various forms. In addition, the current infrared image defect diagnosis basically identifies and diagnoses defects of single devices, making it difficult to detect voltage-induced thermal defects with smaller temperature rise. Summary of the invention

[0007] In view of the above problems, in order to realize intelligent diagnosis of voltage-induced thermal defects, the present invention proposes a method and system for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, which improves the accuracy of equipment part area selection under complex background of electrical equipment, automatically associates and matches similar equipment based on three-phase equipment and logical relationships, and realizes voltage overheating defect diagnosis based on the temperature difference threshold method of similar equipment.

[0008] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, comprising: Acquire infrared images of target areas of electrical equipment to be inspected; Based on the pre-trained Faster RCNN algorithm model, the electrical equipment in the infrared image is identified, and the three-phase area of ​​the electrical equipment is determined; based on the pre-trained The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; Based on the infrared image, determining an infrared image temperature matrix; based on the infrared image temperature matrix, determining temperature matrices of the same type of porcelain sleeves; The maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeve is determined respectively; the temperature difference value of the same type of porcelain sleeve is calculated based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeve and the preset standard, it is judged whether the current electrical equipment has a voltage-induced thermal defect abnormality.

[0009] Further, in the step of respectively determining the center point of each porcelain sleeve, judging the inclusion relationship between each porcelain sleeve and the three-phase area based on the center point of each porcelain sleeve, and associating each porcelain sleeve contained in the three-phase area as a porcelain sleeve of the same type: When the center point of the porcelain bushing is included in the three-phase area, it is determined that the porcelain bushing is included in the three-phase area; and the porcelain bushings included in the three-phase area are associated as the same type of porcelain bushings.

[0010] Furthermore, based on the pre-trained In the step of the algorithm model identifying each porcelain sleeve in the infrared image, the The algorithm model is trained as follows: Acquire infrared image data of electrical equipment for model training, annotate the porcelain sleeve in the infrared image data of the electrical equipment by using a rotation annotation method, and obtain infrared image data of the electrical equipment with a rotation annotation frame; The infrared image data of the electrical equipment with the rotating annotation frame is used as a training set. The algorithm model is trained, and the trained Algorithmic model.

[0011] Furthermore, the porcelain sleeve in the infrared image data of the electrical equipment is annotated by a rotation annotation method, including: In the infrared image data of the electrical equipment, marking the center point of the porcelain sleeve; Construct a plane coordinate system, take the center point of the porcelain sleeve as the rotation center point, and rotate the porcelain sleeve based on the rotation center point in the plane coordinate system. Generate a bounding box of any rotation direction and obtain the rotated and annotated porcelain sleeve; where (x, y) is the rotation center point of the bounding box. indicates that the bounding box is rotated to an acute angle parallel to the x-axis, The value range is [- / 2,0], w is the length of the side of the bounding box that forms an acute angle with the x-axis, and h is the length of the other side of the bounding box.

[0012] Furthermore, in the step of identifying the electrical equipment in the infrared image based on the pre-trained Faster RCNN algorithm model, the Faster RCNN algorithm model is trained as follows: Acquire infrared image data of electrical equipment for model training, and annotate the electrical equipment and three-phase areas in the infrared image data of the electrical equipment by using a non-rotational annotation method to obtain infrared image data of the electrical equipment with an annotation frame; The infrared image data of the electrical equipment with the marked boxes is used as a training set to train the Faster RCNN algorithm model, and a trained Faster RCNN algorithm model is obtained after the model converges.

[0013] Further, based on the infrared image, determining an infrared image temperature matrix includes: Extracting temperature information from the infrared image; wherein the temperature information includes a byte-type temperature matrix and environmental parameters; Decode the byte temperature matrix and convert it into integer data. Perform bitwise operations on the integer data to obtain the matrix raw. The temperature matrix of the entire infrared image is determined based on the environmental parameters and the matrix raw.

[0014] Furthermore, the temperature difference of similar porcelain sleeves is calculated based on the maximum value of each porcelain sleeve temperature matrix, and based on the comparison between the temperature difference of similar porcelain sleeves and the preset standard, it is judged whether the current electrical equipment has a voltage-induced thermal defect abnormality, including: Determine the maximum value of each porcelain sleeve temperature matrix , … , n Indicates the number of porcelain sleeves; Calculate the temperature difference of the same type of porcelain sleeves , … -min( , … ); The temperature difference Compared with the preset standard, when the temperature difference When it exceeds the preset standard, it is determined that the current electrical equipment has a voltage-induced thermal defect anomaly.

[0015] In a second aspect, the present invention provides a device for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, comprising: A data acquisition module, used to acquire an infrared image of a target area of ​​the electrical equipment to be inspected; The recognition module is used to recognize the electrical equipment in the infrared image based on the pre-trained Faster RCNN algorithm model and determine the three-phase area of ​​the electrical equipment; based on the pre-trained Faster RCNN algorithm model The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; A matrix determination module, used to determine the infrared image temperature matrix based on the infrared image; based on the infrared image temperature matrix, determine each temperature matrix of the same type of porcelain sleeve; The abnormality identification module is used to determine the maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeves; calculate the temperature difference value of the same type of porcelain sleeves based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeves and the preset standard, determine whether the current electrical equipment has a voltage-induced thermal defect abnormality.

[0016] According to a third aspect of the present invention, there is provided an electronic device comprising a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the above-mentioned method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes.

[0017] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes as described above is implemented.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: In order to cope with the problem of difficulty in finding voltage-induced thermal defects with small temperature rise in scenarios where the background of electrical equipment is complex, multiple equipment has similar appearances, and defects manifest in various forms, the method of the present invention is based on three-phase equipment and logical relationships, associates and matches similar equipment, and realizes voltage overheating defect diagnosis based on the temperature difference threshold method of similar equipment. It can effectively and automatically identify voltage-induced thermal defects from infrared images, laying the foundation for realizing intelligent infrared diagnosis of machine inspections, and has the advantages of no power outage, no contact, and high sensitivity. By automatically associating similar equipment contained in the three-phase detection frame to achieve similar comparison, this method is easier to identify voltage overheating defects with small temperature rise than the method of only performing target detection defects on single equipment. The electrical equipment voltage-induced thermal defect identification device, electronic device, and computer-readable storage medium provided by the present invention based on thermal field changes also solve the problems raised in the background technology section.

[0019] The method of the present invention utilizes Combined with the rotation annotation training set, the identification and positioning of the porcelain sleeve of electrical equipment is realized. The porcelain sleeve is prone to voltage-induced thermal defects such as increased dielectric loss and cracks. Accurate selection can further improve the accuracy of voltage overheating defect identification. At the same time, the detection accuracy of the target area of ​​the equipment is improved, and the influence of the interference area is eliminated to the greatest extent. It is more suitable for the positioning of the target area of ​​electrical equipment based on thermal field changes with complex backgrounds. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 It is a schematic diagram of the implementation principle of a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to an embodiment of the present invention; Figure 2 This is a flow chart of a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a rotation mark in an embodiment of the present invention; Figure 4 It is a structural diagram of FRM in an embodiment of the present invention; Figure 5 Schematic diagram of DCN in an embodiment of the present invention; Figure 6 This is a structural block diagram of a device for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to an embodiment of the present invention; Figure 7 The present invention is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0022] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention.

[0023] Example 1 The embodiment of the present invention provides a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, which uses the Faster RCNN model to identify and locate electrical equipment and three-phase areas in infrared images. The model identifies the porcelain sleeve of electrical equipment, automatically associates similar equipment contained in the three-phase area, calculates the temperature difference of similar equipment, and uses the temperature difference threshold method to diagnose voltage-induced thermal defects, which can timely and effectively discover thermal defects of electrical equipment.

[0024] like Figure 1 and Figure 2 As shown, a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes includes: S1. Acquire an infrared image of a target area of ​​an electrical device to be inspected.

[0025] S2, based on the pre-trained Faster RCNN algorithm model, identify the electrical equipment in the infrared image and determine the three-phase area of ​​the electrical equipment; based on the pre-trained Faster RCNN algorithm model, The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, judges the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as the same type of porcelain sleeve.

[0026] Specifically, in step S2, when the center point of the porcelain sleeve is included in the three-phase area, it is determined that the porcelain sleeve is included in the three-phase area; and the porcelain sleeves included in the three-phase area are associated as the same type of porcelain sleeves.

[0027] Specifically, in step S2, the The algorithm model is trained as follows: Acquire infrared image data of electrical equipment for model training, annotate the porcelain sleeve in the infrared image data of the electrical equipment by using a rotation annotation method, and obtain infrared image data of the electrical equipment with a rotation annotation frame; The infrared image data of the electrical equipment with the rotating annotation frame is used as a training set. The algorithm model is trained, and the trained Algorithmic model.

[0028] Specifically, the porcelain sleeve in the infrared image data of the electrical device is annotated by a rotation annotation method, including: annotating the center point of the porcelain sleeve in the infrared image data of the electrical device; constructing a plane coordinate system, taking the center point of the porcelain sleeve as a rotation center point, rotating the porcelain sleeve based on the rotation center point in the plane coordinate system, and using Generate a bounding box of any rotation direction and obtain the rotated and annotated porcelain sleeve; where (x, y) is the rotation center point of the bounding box. indicates that the bounding box is rotated to an acute angle parallel to the x-axis, The value range is [- / 2,0], w is the length of the side of the bounding box that forms an acute angle with the x-axis, and h is the length of the other side of the bounding box.

[0029] Specifically, in step S2, the Faster RCNN algorithm model is trained as follows: The infrared image data of electrical equipment used for model training is obtained, and the electrical equipment and three-phase areas in the infrared image data of the electrical equipment are labeled by non-rotating labeling to obtain the infrared image data of the electrical equipment with a labeling box; the infrared image data of the electrical equipment with the labeling box is used as a training set to train the Faster RCNN algorithm model, and a trained Faster RCNN algorithm model is obtained after the model converges.

[0030] In the above model training process, the data set is labeled. In this solution, the labeling categories are divided into 3 categories: Category 1 is labeled as three-phase area, Category 2 is labeled as electrical equipment, and Category 3 is labeled as porcelain sleeve. The three-phase area and electrical equipment targets are labeled without rotation, and the Faster RCNN model is used for target detection. The porcelain sleeve is rotated and labeled using the improved The model is used for target detection. Since the porcelain bushing needs to be compared with the temperature of similar equipment to determine whether the equipment has defects, the accuracy of the target detection area recognition directly determines the accuracy of the voltage overheating defect recognition. The infrared image equipment target presents a rectangular area with uncertain direction. The use of rotating target detection can eliminate background interference to the greatest extent; the three-phase area and equipment target detection is only to determine the equipment type and automatically associate the three-phase relationship. The requirements for the accuracy of the target detection area recognition are not high, so the three-phase and electrical equipment areas are marked without a rotating label box.

[0031] To detect rotating targets of electrical equipment porcelain sleeves, the algorithm uses Represents a bounding box in any direction, where (x, y) is the center point of the bounding box. indicates that the bounding box is rotated to an acute angle parallel to the x-axis, The value range is [- / 2,0], w is the length of the side that forms an acute angle with the x-axis, and h is the length of the other side of the rectangle, such as Figure 3 shown.

[0032] In a preferred embodiment, Combine the two forms of anchor frames to locate the target area and rotate the anchor frame (i.e. Not taking the boundary value of the value range) can play a better role in scenes with dense objects, and the horizontal anchor box (i.e. Taking the boundary value of the value range) can achieve a higher recall rate with fewer numbers, In the first stage, horizontal anchor boxes are used to increase speed and generate more candidate boxes, and then rotated anchor boxes are used in the refinement stage (FRM) to adapt to scenes with dense targets. After the algorithm obtains the predicted bounding box of the first stage through the pyramid network (FPN), it continues to correct the predicted bounding box through FRM, and re-predicts the final result through the target regression and classification subnet. FPM filters the predicted bounding box to increase the inference speed. Considering that the predicted bounding box is displaced relative to the anchor box boundary, FRM reconstructs the feature map using the bilinear interpolation method according to the position of the predicted bounding box. The structure diagram of FRM is shown in the figure. Figure 4 shown.

[0033] In order to adapt to the different sizes, postures, and perspectives of electrical equipment in images, this solution The algorithm is improved. The deformable convolutional network (DCN) is introduced in the algorithm. In DCN, there is an additional convolution layer for learning the offset of the sampling point. Then the offset and the feature map are used as the input of DCN. The DCN layer operates the sampling point after the offset and then performs convolution. By replacing the convolution layer responsible for input of ResNet-101 with the DCN layer, the algorithm can adapt to changes in the position and size of electrical equipment, so as to more effectively complete the extraction of image features, thereby achieving better electrical equipment target detection effect. The schematic diagram of DCN is shown below: Figure 5 The method of the present invention replaces the convolution layer responsible for input of ResNet-10 with a deformable convolution layer, and the algorithm can adapt to changes in the position and size of electrical equipment, thereby more effectively completing the extraction of image features, thereby achieving a better target recognition effect of thermal defects in electrical equipment.

[0034] In this scheme, automatic association of similar equipment is one of the key points of the intelligent similar comparison and discrimination method of the voltage overheating type defect diagnosis method. The three-phase area, electrical equipment and porcelain sleeve in the infrared image are identified, and similar porcelain sleeves are automatically associated. Based on the inclusion logical relationship, similar equipment is automatically associated, and the center point of the porcelain sleeve in the infrared image is calculated respectively. If the center point of the porcelain sleeve is contained in the three-phase area, it is considered to be a porcelain sleeve contained in the three-phase area, and the porcelain sleeve contained in the three-phase area is automatically associated as similar equipment.

[0035] In an optional embodiment, in addition to detecting electrical equipment in the traditional technology, the present invention also adds detection of the three-phase area and the porcelain sleeve of the equipment. The three-phase detection frame is added to make the same type of equipment have mutual reference and can be compared with each other. By associating the equipment whose center point is included in the three-phase area, the same type of equipment can be automatically associated and matched; the porcelain sleeve part is detected because the porcelain sleeve part is prone to voltage-induced heating defects such as increased dielectric loss and cracks. The present invention marks the electrical equipment target without rotation and marks the porcelain sleeve of the electrical equipment with rotation.

[0036] In an optional embodiment, the present invention automatically associates similar devices (porcelain sleeves) based on the inclusion logical relationship, calculates the center point of the device (porcelain sleeve) in the infrared image, and if the center point is included in the three-phase area, it is considered to be a device included in the three-phase area, and automatically associates the same type of devices included in the three-phase area as similar devices, and then calculates the temperature difference value of similar devices according to the infrared image temperature matrix. This solution mainly determines whether the electrical equipment to which the porcelain sleeve belongs is abnormal through the porcelain sleeve.

[0037] In an optional embodiment, the improved algorithm, Using a refined single-stage target detector, a feature refinement module (FRM) is designed and introduced on the basis of the Retina Net target detection algorithm to address the feature alignment problem. The algorithm first extracts the features of the infrared image of electrical equipment through the ResNet-101 network, and then constructs a multi-scale feature pyramid through the feature pyramid network (FPN) with a top-down path and lateral connections, thereby enhancing the algorithm's ability to recognize electrical equipment targets of different sizes. The output of each layer of FPN is connected to the classification and regression subnets respectively to identify the target category and position.

[0038] In an alternative embodiment, Figure 4 and Figure 5 Two key improvements in the network structure are demonstrated: one is to optimize the prediction bounding box through FRM, and the other is to enhance the feature extraction capability through DCN. The two modules work together to make the algorithm more accurate and effective in identifying electrical equipment, especially those porcelain sleeves that are prone to voltage-induced heating defects. Figure 4 The structure of the Feature Refinement Module (FRM) is shown, and Figure 5 The schematic diagram of the Deformable Convolutional Network (DCN) is shown. These two parts are important improvements in the entire network structure. The role of the refined feature module (FRM) is to further correct the position of these bounding boxes after the initial prediction of the bounding boxes to improve the accuracy of the final detection results. The structure of FRM is as follows Figure 4 As shown in the figure. First, the predicted bounding box is corrected. After obtaining the preliminary predicted bounding box through the pyramid network (FPN), FRM corrects these bounding boxes. Considering that the predicted bounding box is displaced relative to the anchor box boundary, FRM uses a bilinear interpolation method to reconstruct the feature map. The corrected bounding box is then finally predicted by the classification subnet and the regression subnet. The introduction of DCN is to enable the model to better adapt to the different sizes, postures and perspective changes of electrical equipment in the image and improve the effect of image feature extraction. An additional convolutional layer is added to DCN to learn the offset of the sampling point. The learned offset is combined with the feature map as the input of DCN, so that the sampling point can be offset before convolution. By replacing the convolutional layer responsible for input in ResNet-101 with the DCN layer, the model can better adapt to the position and size changes of electrical equipment. First, after extracting image features through a backbone network such as ResNet-101, FPN constructs a multi-scale feature pyramid. Feature maps of different levels are generated, which gradually increase from the top to the bottom, and each layer of feature maps represents target information of different scales. In the top-down path, the feature maps of each level are merged with the feature maps from deeper levels through 1x1 and 5x1 convolutions. This connection method can be understood as a fusion of information, ensuring that the feature maps of each level contain all the information from the original input image to that level. After the algorithm obtains the predicted bounding box of the first stage through the pyramid network (FPN), it continues to correct the predicted bounding box through FRM, and re-predicts the final result through the target regression and classification subnet. FPM filters the predicted bounding box to increase the inference speed. Considering that the predicted bounding box is displaced relative to the anchor box boundary, FRM uses bilinear interpolation to reconstruct the feature map according to the position of the predicted bounding box. The structure diagram of FRM is shown below. Figure 4 In order to adapt to the different sizes, postures, and perspectives of electrical equipment in images, this paper The algorithm is improved. The deformable convolutional network (DCN) is introduced in the algorithm. In DCN, there is an additional convolution layer for learning the offset of the sampling point. Then the offset and the feature map are used as the input of DCN. The DCN layer operates the sampling point after the offset and then performs convolution. By replacing the convolution layer responsible for input of ResNet-101 with the DCN layer, the algorithm can adapt to changes in the position and size of electrical equipment, so as to more effectively complete the extraction of image features, thereby achieving better electrical equipment target detection effect. The schematic diagram of DCN is shown below: Figure 5As shown in the figure. With these improvements, the model can perform better in the task of identifying electrical equipment targets, especially in areas such as porcelain sleeves that are prone to voltage-induced heating defects, and can provide more accurate identification and positioning. These improvements help improve the effect of electrical equipment thermal defect target recognition, especially in complex background conditions.

[0039] S3. Based on the infrared image, determine an infrared image temperature matrix; based on the infrared image temperature matrix, determine each temperature matrix of the same type of porcelain sleeves.

[0040] Specifically, in step S3, based on the infrared image, determining the infrared image temperature matrix includes: Extracting temperature information from the infrared image; wherein the temperature information includes a byte-type temperature matrix and environmental parameters; Decode the byte temperature matrix and convert it into integer data, and perform bitwise operations on the integer data to obtain the matrix raw; The temperature matrix of the entire infrared image is determined based on the environmental parameters and the matrix raw.

[0041] In an optional embodiment, for the captured infrared image, the temperature information of the infrared image is first determined, and the temperature information includes the original thermal sensor byte-type temperature matrix and environmental parameters; the environmental parameters include the emissivity E, the reflected temperature , distance D, atmospheric temperature , External optical temperature , External optical transmission and relative humidity Then, the extracted byte temperature matrix information is decoded into integer data, and the matrix raw is obtained by bit operation. Finally, based on the relationship between the environmental parameters and the matrix raw, the temperature matrix of the entire infrared image is obtained. .

[0042] Optionally, an infrared camera is used to shoot infrared images. The infrared camera outputs temperature data in the form of bytes, and these bytes are decoded into integer data that is easier to process; the decoded integer data is further converted into the original temperature matrix (matrix raw) of the infrared image through bit operations (such as displacement, bitwise AND, bitwise OR, etc.). Next, the temperature data of the infrared image is corrected using the collected environmental parameters (such as emissivity, reflected temperature, distance, atmospheric temperature, external optical temperature, external optical transmission, and relative humidity). Using the corrected data, the actual temperature value of each pixel in the entire infrared image can be calculated to form the final temperature matrix.

[0043] S4. Determine the maximum value of each temperature matrix of the same type of porcelain sleeves respectively; calculate the temperature difference of the same type of porcelain sleeves based on the maximum value of each temperature matrix of the same type of porcelain sleeves, and judge whether the current electrical equipment has a voltage-induced thermal defect abnormality based on the comparison relationship between the temperature difference of the same type of porcelain sleeves and the preset standard.

[0044] Specifically, in step S4, the temperature difference of similar porcelain sleeves is calculated based on the maximum value of each porcelain sleeve temperature matrix, and based on the comparison between the temperature difference of similar porcelain sleeves and the preset standard, it is determined whether the current electrical equipment has a voltage-induced thermal defect abnormality, including: Determine the maximum value of each porcelain sleeve temperature matrix , … , n Indicates the number of porcelain sleeves; Calculate the temperature difference of the same type of porcelain sleeves , … -min( , … ); The temperature difference Compared with the preset standard, when the temperature difference When it exceeds the preset standard, it is determined that the current electrical equipment has a voltage-induced thermal defect anomaly.

[0045] In an optional embodiment, the porcelain sleeve area of ​​the electrical equipment with higher temperature is marked with a red frame.

[0046] As an example, based on the acquired infrared image temperature matrix , respectively obtain the temperature of the comparable porcelain sleeve, calculate the maximum value of the temperature matrix in the associated porcelain sleeve, taking three porcelain sleeves as an example, respectively , , , then the temperature difference -min( , , ), compare the temperature difference value with the threshold value to determine whether a voltage-induced thermal defect occurs, thereby realizing the identification of voltage-induced thermal defects of electrical equipment based on temperature field changes.

[0047] In an extended embodiment, the temperature difference values ​​of the porcelain sleeves of different electrical devices may also be compared by temperature matrix, thereby realizing abnormality detection of multiple electrical devices.

[0048] Example 2 like Figure 6 As shown, based on the same inventive concept as the above embodiment, the present invention also provides a device for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, comprising: A data acquisition module, used to acquire an infrared image of a target area of ​​the electrical equipment to be inspected; The recognition module is used to recognize the electrical equipment in the infrared image based on the pre-trained Faster RCNN algorithm model and determine the three-phase area of ​​the electrical equipment; based on the pre-trained Faster RCNN algorithm model The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; A matrix determination module, used to determine the infrared image temperature matrix based on the infrared image; based on the infrared image temperature matrix, determine each temperature matrix of the same type of porcelain sleeve; The abnormality identification module is used to determine the maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeves; calculate the temperature difference value of the same type of porcelain sleeves based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeves and the preset standard, determine whether the current electrical equipment has a voltage-induced thermal defect abnormality.

[0049] Example 3 like Figure 7 As shown, the present invention also provides an electronic device 100 for implementing the above-mentioned embodiment of a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes; The electronic device 100 includes a memory 101 , at least one processor 102 , a computer program 103 stored in the memory 101 and executable on the at least one processor 102 , and at least one communication bus 104 .

[0050] The memory 101 can be used to store a computer program 103. The processor 102 implements the steps of a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes in Example 1 by running or executing the computer program stored in the memory 101 and calling data stored in the memory 101.

[0051] The memory 101 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data (such as audio data) created according to the use of the electronic device 100, etc. In addition, the memory 101 may include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0052] At least one processor 102 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and uses various interfaces and lines to connect various parts of the entire electronic device 100.

[0053] The memory 101 in the electronic device 100 stores a plurality of instructions to implement a method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, and the processor 102 can execute the plurality of instructions to implement: Acquire infrared images of target areas of electrical equipment to be inspected; Based on the pre-trained Faster RCNN algorithm model, the electrical equipment in the infrared image is identified, and the three-phase area of ​​the electrical equipment is determined; based on the pre-trained The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; Based on the infrared image, determining an infrared image temperature matrix; based on the infrared image temperature matrix, determining temperature matrices of the same type of porcelain sleeves; Determine the maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeve respectively; calculate the temperature difference value of the same type of porcelain sleeve based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeve and the preset standard, judge whether the current electrical equipment has a voltage-induced thermal defect anomaly.

[0054] Example 4 If the module / unit integrated in the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0055] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0056] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0057] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0058] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0059] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, characterized in that: include: Acquire infrared images of target areas of electrical equipment to be inspected; Based on the pre-trained Faster RCNN algorithm model, the electrical equipment in the infrared image is identified, and the three-phase area of ​​the electrical equipment is determined; based on the pre-trained The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; Based on the infrared image, determining an infrared image temperature matrix; based on the infrared image temperature matrix, determining temperature matrices of the same type of porcelain sleeves; Determine the maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeve respectively; calculate the temperature difference value of the same type of porcelain sleeve based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeve and the preset standard, judge whether the current electrical equipment has a voltage-induced thermal defect anomaly.

2. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 1, characterized in that: In the step of respectively determining the center point of each porcelain bushing, judging the inclusion relationship between each porcelain bushing and the three-phase region based on the center point of each porcelain bushing, and associating each porcelain bushing contained in the three-phase region as a same type of porcelain bushing: When the center point of the porcelain bushing is included in the three-phase area, it is determined that the porcelain bushing is included in the three-phase area; and the porcelain bushings included in the three-phase area are associated as the same type of porcelain bushings.

3. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 1, characterized in that: Based on pre-trained In the step of the algorithm model identifying each porcelain sleeve in the infrared image, the The algorithm model is trained as follows: Acquire infrared image data of electrical equipment for model training, annotate the porcelain sleeve in the infrared image data of the electrical equipment by using a rotation annotation method, and obtain infrared image data of the electrical equipment with a rotation annotation frame; The infrared image data of the electrical equipment with the rotating annotation frame is used as a training set. The algorithm model is trained, and the trained Algorithmic model.

4. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 3 is characterized in that: The porcelain sleeve in the infrared image data of electrical equipment is marked by rotating marking, including: In the infrared image data of the electrical equipment, marking the center point of the porcelain sleeve; Construct a plane coordinate system, take the center point of the porcelain sleeve as the rotation center point, and rotate the porcelain sleeve based on the rotation center point in the plane coordinate system. Generate a bounding box of any rotation direction and obtain the rotated and annotated porcelain sleeve; where (x, y) is the rotation center point of the bounding box. indicates that the bounding box is rotated to an acute angle parallel to the x-axis, The value range is [- / 2,0], w is the length of the side of the bounding box that forms an acute angle with the x-axis, and h is the length of the other side of the bounding box.

5. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 1, characterized in that: In the step of identifying the electrical equipment in the infrared image based on the pre-trained Faster RCNN algorithm model, the Faster RCNN algorithm model is trained as follows: Acquire infrared image data of electrical equipment for model training, and annotate the electrical equipment and three-phase areas in the infrared image data of the electrical equipment by using a non-rotational annotation method to obtain infrared image data of the electrical equipment with an annotation frame; The infrared image data of the electrical equipment with the marked boxes is used as a training set to train the Faster RCNN algorithm model, and a trained Faster RCNN algorithm model is obtained after the model converges.

6. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 1, characterized in that: Based on the infrared image, determining an infrared image temperature matrix includes: Extracting temperature information from the infrared image; wherein the temperature information includes a byte-type temperature matrix and environmental parameters; Decode the byte temperature matrix and convert it into integer data. Perform bitwise operations on the integer data to obtain the matrix raw. The temperature matrix of the entire infrared image is determined based on the environmental parameters and the matrix raw.

7. The method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes according to claim 1, characterized in that: Based on the maximum value of each porcelain sleeve temperature matrix, the temperature difference of the same type of porcelain sleeve is calculated. Based on the comparison between the temperature difference of the same type of porcelain sleeve and the preset standard, it is judged whether the current electrical equipment has a voltage-induced thermal defect abnormality, including: Determine the maximum value of each porcelain sleeve temperature matrix , … , n Indicates the number of porcelain sleeves; Calculate the temperature difference of the same type of porcelain sleeves , … -min( , … ); The temperature difference Compared with the preset standard, when the temperature difference When it exceeds the preset standard, it is determined that the current electrical equipment has a voltage-induced thermal defect anomaly.

8. A device for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes, characterized in that: include: A data acquisition module, used to acquire an infrared image of a target area of ​​the electrical equipment to be inspected; The recognition module is used to recognize the electrical equipment in the infrared image based on the pre-trained Faster RCNN algorithm model and determine the three-phase area of ​​the electrical equipment; based on the pre-trained Faster RCNN algorithm model The algorithm model identifies each porcelain sleeve in the infrared image; determines the center point of each porcelain sleeve respectively, and based on the center point of each porcelain sleeve, determines the inclusion relationship between each porcelain sleeve and the three-phase area, and associates each porcelain sleeve contained in the three-phase area as a similar porcelain sleeve; A matrix determination module, used to determine the infrared image temperature matrix based on the infrared image; based on the infrared image temperature matrix, determine each temperature matrix of the same type of porcelain sleeve; The abnormality identification module is used to determine the maximum value of each porcelain sleeve temperature matrix in the same type of porcelain sleeves; calculate the temperature difference value of the same type of porcelain sleeves based on the maximum value of each porcelain sleeve temperature matrix; based on the comparison relationship between the temperature difference value of the same type of porcelain sleeves and the preset standard, determine whether the current electrical equipment has a voltage-induced thermal defect abnormality.

9. An electronic device, characterized in that: It comprises a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, the method for identifying voltage-induced thermal defects of electrical equipment based on thermal field changes as described in any one of claims 1 to 7 is implemented.

Images