Power grid defect diagnosis method, device, computer equipment and storage medium

Abstract

The embodiments of this specification provide a power grid defect diagnosis method, device, computer device, and storage medium, which relate to the field of image processing technology. The method includes: obtaining defect aggregation rules, multiple inspection images, and the position information corresponding to the inspection images; the multiple inspection images are images taken for multiple specified areas of at least one pole tower in the power grid; inputting the inspection images into a defect recognition model to extract device component features, and based on the extracted device component features, performing defect prediction to obtain the first defect information corresponding to the inspection images; according to the position information corresponding to the inspection images, updating the first defect information corresponding to the inspection images to obtain the second defect information corresponding to the inspection images; according to the defect aggregation rules and the second defect information corresponding to each of the multiple inspection images, determining the defect point set of any pole tower in the power grid; outputting the defect point set of any pole tower according to the pole tower dimension. In this way, the diagnostic accuracy of power grid defects can be improved.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of image processing technology, and more particularly to a power grid defect diagnosis method, device, computer device, and storage medium. Background Art

[0002] The overhead lines in the power grid are transmission lines that are erected above the ground and fix the transmission conductors on multiple poles standing upright on the ground to transmit electric energy. With the rapid development of the power industry, the scale and complexity of the power grid are increasing day by day, and the overhead lines in the power grid are also vulnerable to the influence of meteorology and environment (such as strong wind, lightning strike, pollution, ice and snow, etc.), which may cause defects and thus result in failures.

[0003] In the related art, when diagnosing defects and performing operation and maintenance on the overhead lines in the power grid, it usually relies on manual inspection and experience judgment, with relatively low defect diagnosis accuracy and great operation and maintenance difficulty, which is difficult to meet the current requirements of power grid operation.

[0004] Therefore, there is an urgent need to provide a power grid defect diagnosis method to improve the accuracy of power grid defect diagnosis. Summary of the Invention

[0005] In view of this, multiple embodiments of this specification are committed to providing a power grid defect diagnosis method, device, computer device, and storage medium to improve the accuracy of power grid defect diagnosis.

[0006] An embodiment of this specification provides a power grid defect diagnosis method. The power grid includes multiple poles in the overhead lines. The method includes: obtaining a defect aggregation rule, multiple inspection images, and position information corresponding to the inspection images; wherein, the multiple inspection images are images taken for multiple specified areas of at least one of the poles, and the defect aggregation rule is used to aggregate and count defect points in the multiple specified areas of the poles; inputting the inspection images into a pre-trained defect recognition model to extract device element features, and performing defect prediction based on the extracted device element features to obtain first defect information corresponding to the inspection images; wherein, the first defect information includes the type and quantity of defect points in the specified areas corresponding to the inspection images; updating the first defect information corresponding to the inspection images according to the position information corresponding to the inspection images to obtain second defect information corresponding to the inspection images; wherein, the second defect information includes the type, quantity, and position of defect points in the specified areas corresponding to the inspection images; determining a defect point set of any one of the poles in the power grid according to the defect aggregation rule and the second defect information corresponding to each of the multiple inspection images; wherein, the defect point set includes the second defect information corresponding to each of the multiple specified areas in the pole; outputting the defect point set of any one of the poles according to the pole dimension.

[0007] In some embodiments, the specified area is any one of the tower body area, the large-side channel, and the small-side channel; the tower body area is any one of the whole tower, the tower top, the tower head, the tower body, the left cross arm, and the right cross arm; the inspection image corresponding to the large-side channel is the large-side channel inspection image, the inspection image corresponding to the small-side channel is the small-side channel inspection image, and the inspection image corresponding to the tower body area is the tower body inspection image; determining the set of defect points of any tower in the power grid according to the defect aggregation rule and the second defect information corresponding to each of the multiple inspection images includes: for any tower, merging and de-duplicating the second defect information corresponding to each of the multiple tower body inspection images in the tower to obtain the tower body defect point set of the tower; merging the second defect information corresponding to the large-side channel inspection image and the small-side channel inspection image of the tower, and the tower body defect point set of the tower to obtain the defect point set of the tower.

[0008] In some embodiments, before inputting the inspection image into the defect recognition model, the method further includes: preprocessing the inspection image to obtain the preprocessed inspection image; wherein, the preprocessing includes at least one of image denoising, image enhancement, and image segmentation.

[0009] In some embodiments, the method further includes: outputting the defect point set of any tower according to the type dimension of the defect points in the defect point set of the tower.

[0010] In some embodiments, the types of the defect points include at least one of the following: insulator contamination, insulator discharge marks, tree and bamboo obstacles, tower nests, unbundled conductors, and damaged conductors.

[0011] In some embodiments, the defect recognition model is obtained through the following training method: constructing a training image sample set; wherein, the training image sample set includes multiple training image samples, and the label of the training image sample is the type of the defect point; training the initial model according to the training image sample set and the label to obtain the defect recognition model; wherein, the initial model is built based on any one of an autoencoder, a Transformer, and a convolutional neural network model.

[0012] An embodiment of this specification provides a power grid defect diagnosis device. The power grid includes multiple poles and towers in overhead lines. The device includes: a data acquisition module, configured to acquire defect aggregation rules, multiple inspection images, and position information corresponding to the inspection images. Among them, the multiple inspection images are images taken for multiple specified areas of at least one of the poles and towers, and the defect aggregation rules are used to aggregate and count defect points in the multiple specified areas of the poles and towers; a defect prediction module, configured to input the inspection images into a pre-trained defect recognition model to extract device component features, and perform defect prediction based on the extracted device component features to obtain first defect information corresponding to the inspection images. Among them, the first defect information includes the type and quantity of defect points in the specified area corresponding to the inspection image; a defect positioning module, configured to update the first defect information corresponding to the inspection image according to the position information corresponding to the inspection image to obtain second defect information corresponding to the inspection image. Among them, the second defect information includes the type, quantity, and position of defect points in the specified area corresponding to the inspection image; a defect aggregation module, configured to determine a set of defect points of any one of the poles and towers in the power grid according to the defect aggregation rules and the second defect information corresponding to each of the multiple inspection images. Among them, the set of defect points includes the second defect information corresponding to each of the multiple specified areas in the pole and tower; a defect output module, configured to output the set of defect points of any one of the poles and towers according to the pole and tower dimension.

[0013] In some embodiments, the specified area is any one of the pole and tower body area, the large-side channel, and the small-side channel; the pole and tower body area is any one of the whole tower, the tower top, the tower head, the tower body, the left cross arm, and the right cross arm; the inspection image corresponding to the large-side channel is a large-side channel inspection image, the inspection image corresponding to the small-side channel is a small-side channel inspection image, and the inspection image corresponding to the pole and tower body area is a pole and tower body inspection image. The defect aggregation module is further configured to: for any one of the poles and towers, merge and de-duplicate the second defect information corresponding to each of the multiple pole and tower body inspection images in the pole and tower to obtain a set of pole and tower body defect points of the pole and tower; merge the second defect information corresponding to the large-side channel inspection image and the small-side channel inspection image of the pole and tower, and the set of pole and tower body defect points of the pole and tower to obtain the set of defect points of the pole and tower.

[0014] An embodiment of this specification provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the power grid defect diagnosis method described in any of the above embodiments.

[0015] An embodiment of this specification provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the power grid defect diagnosis method described in any of the above embodiments is implemented.

[0016] In multiple embodiments provided in this specification, by taking pictures of at least one pole tower in the overhead line of the power grid, multiple inspection images and their corresponding position information are obtained, and a defect collection rule for collecting and counting defect points in multiple specified areas of the pole tower is obtained. Then, the inspection images are input into a pre-trained defect recognition model to extract device component features, and defect prediction is performed based on the extracted device component features to obtain first defect information corresponding to the inspection images. The first defect information includes the type and quantity of defect points in the specified area corresponding to the inspection image. Then, according to the position information corresponding to the inspection image, the first defect information corresponding to the inspection image is updated to obtain second defect information corresponding to the inspection image. The second defect information includes the type, quantity, and position of defect points in the specified area corresponding to the inspection image. Then, according to the defect collection rule and the second defect information corresponding to each of the multiple inspection images of any pole tower in the power grid, a set of defect points of the pole tower is determined. The set of defect points of the pole tower includes the second defect information corresponding to each of the multiple specified areas in the pole tower. Then, after determining the set of defect points corresponding to at least one pole tower in the power grid, according to the pole tower dimension, the set of defect points of any pole tower is output. In this way, the diagnostic accuracy of power grid defects can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the power grid defect diagnosis system provided by the embodiment of this specification;

[0018] Figure 2 It is a schematic flowchart of the power grid defect diagnosis method provided by the embodiment of this specification;

[0019] Figure 3 It is a schematic diagram of the power grid defect diagnosis device provided by the embodiment of this specification;

[0020] Figure 4 It is a schematic diagram of the computer device provided by the embodiment of this specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In order to enable those skilled in the art to better understand the solutions of this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this specification.

[0022] The embodiments of this specification provide a scenario example of a power grid defect diagnosis system. Please refer to Figure 1 , Figure 1 which is a schematic diagram of the power grid defect diagnosis system provided by the embodiments of this specification. The power grid defect diagnosis system 100 can be used to diagnose defects of one or more poles 110 of overhead lines in the power grid. The power grid defect diagnosis system 100 can include a data acquisition device 120, a defect diagnosis device 130, and a defect output device 140.

[0023] The data acquisition device 120 can be used to collect images of the overhead lines of the power grid or one or more poles 110 in the overhead lines during the inspection process, and obtain inspection images of one or more poles 110 in the overhead lines. Exemplarily, the data acquisition device 120 can be an image acquisition device or a vision device that can collect inspection images or inspection videos in real time during the inspection process. For example, it can be a drone with a camera. After the data acquisition device 120 collects the inspection images or inspection videos, it can transmit the collected inspection images or inspection videos to the defect diagnosis device 130.

[0024] The defect diagnosis device 130 can be a client or a server. The client can be an electronic device with network access capabilities. For example, the client can be a desktop computer, a tablet computer, a laptop computer, a smart phone, a digital assistant, etc. Or, the client can also be software that can run on the electronic device. The server can be an electronic device with certain computing and processing capabilities. It can have a network communication module, a processor, a memory, etc. Exemplarily, the server can also be a distributed server, which can be a system with multiple processors, memories, network communication modules, etc. operating in cooperation. Or, the server can also be a server cluster formed by several servers. Or, with the development of science and technology, the server can also be a new technical means that can implement the corresponding functions of the embodiments of the specification. For example, it can be a new form of "server" based on quantum computing.

[0025] Exemplarily, the defect diagnosis device 130 can perform defect diagnosis based on the inspection images after receiving them. Or it can, after receiving the inspection video, obtain inspection images based on the inspection video and perform defect diagnosis based on the inspection images. Exemplarily, after the defect diagnosis device 130 performs defect diagnosis to determine the defect points or defect information, it can be output through the defect output device 140.

[0026] The defect output device 140 can include an image output device. For example, the defect output device can include a display or other display terminals. As another example, the defect output device 140 can also include a voice output device such as a microphone to assist in voice output or alarm.

[0027] An embodiment of this specification provides a power grid defect diagnosis method. Please refer to Figure 2 , Figure 2 which is a schematic flowchart of a power grid defect diagnosis method provided by this embodiment. This embodiment provides method operation steps as shown in the flowchart, but based on routine or non-creative labor, it may include more or fewer operation steps. The order of steps listed in the embodiment is only one execution manner among the execution orders of numerous steps, and does not represent the only execution order. When the actual system or server product executes, it can be executed in the method order shown in the embodiment or in parallel (for example, in an environment of parallel processors or multi-threaded processing). This power grid defect diagnosis method can be applied to defect diagnosis devices in a power grid defect diagnosis system, specifically as Figure 2 shown, and this power grid defect diagnosis method may include the following steps.

[0028] Step S210: Obtain defect aggregation rules, multiple inspection images, and position information corresponding to the inspection images; wherein, the multiple inspection images are images taken for multiple specified areas of at least one tower, and the defect aggregation rules are used to aggregate and count defect points in the multiple specified areas of the tower.

[0029] In some cases, the video acquisition device can inspect one or more towers in the power grid and take pictures of multiple specified areas of one or more towers in the power grid during the inspection process to obtain multiple inspection images. That is to say, the multiple inspection images may include inspection images corresponding to multiple specified areas of one tower respectively, or may include inspection images corresponding to multiple specified areas of multiple towers respectively.

[0030] In still other cases, the video acquisition device can record the corresponding position information when taking pictures of any tower. That is, the inspection images collected by the video acquisition device carry the corresponding position information.

[0031] Among them, the multiple specified areas of the tower may refer to multiple shooting angles of the data acquisition device during the inspection of the tower.

[0032] Specifically, the defect diagnosis device can receive multiple inspection images and the position information carried by each inspection image from the video acquisition device.

[0033] Exemplarily, the defect diagnosis device can obtain the defect aggregation rules or may pre-store the defect aggregation rules to facilitate subsequent aggregation and counting of defect points in the multiple specified areas of the tower.

[0034] Step S220: Input the inspection image into a pre-trained defect recognition model to extract device component features, and perform defect prediction based on the extracted device component features to obtain first defect information corresponding to the inspection image; wherein, the first defect information includes the types and quantities of defect points in a specified area corresponding to the inspection image.

[0035] Specifically, the defect diagnosis device may be deployed with a pre-trained defect recognition model. After inputting the inspection image into the defect recognition model, the defect recognition model can extract the device component features in the inspection image and perform defect prediction based on the extracted device component features to obtain the first defect information of the inspection image.

[0036] Among them, the device component features may be the features of the region of interest in the inspection image. Exemplarily, the region of interest in the inspection image may refer to the device component region, for example, the insulator region, the tower region, the conductor region, the cross-arm region, etc. Exemplarily, the region of interest in the inspection image may also be the tree and bamboo obstacle region.

[0037] Exemplarily, since the inspection image of any tower corresponds to a specified area of that tower, that is to say, the first defect information of any inspection image corresponds to the specified area corresponding to that inspection image. The first defect information includes the types and quantities of defect points in the corresponding specified area.

[0038] As an example, the types of defect points may include at least one of the following: insulator contamination, insulator discharge marks, tree and bamboo obstacles, tower bird nests, unbundled conductors, and damaged conductors. It can be understood that the types of defect points may also include other types, such as damaged insulators, conductor icing, etc.

[0039] Step S230: Update the first defect information corresponding to the inspection image according to the position information corresponding to the inspection image to obtain second defect information corresponding to the inspection image; wherein, the second defect information includes the types, quantities, and positions of defect points in a specified area corresponding to the inspection image.

[0040] Specifically, after determining the first defect information of the inspection image, the first defect information of the inspection image can be updated based on the position information corresponding to the inspection image to obtain the second defect information of the inspection image. Exemplarily, after determining the types and quantities of defect points in the specified area corresponding to the inspection image, the positions of the defect points in the specified area corresponding to the inspection image can be located according to the position information corresponding to the inspection image, that is, the positions of the defect points in the specified area corresponding to the inspection image are determined. In this way, the types, quantities, and positions of defect points in the specified area corresponding to any inspection image can be determined.

[0041] Step S240: Determine the set of defect points of any pole tower in the power grid according to the defect aggregation rule and the second defect information corresponding to each of the multiple inspection images; wherein, the set of defect points includes the second defect information corresponding to multiple specified areas in the pole tower.

[0042] Among them, the defect aggregation rule may refer to statistically determining the second defect information corresponding to each pole tower in the overhead line of the power grid according to the pole tower dimension, that is, statistically determining the second defect information corresponding to multiple specified areas of each pole tower in the overhead line according to the pole tower dimension. The defect aggregation rule can describe the statistical logic of the set of defect points corresponding to any pole tower to reduce the duplication and missed detection of defect points.

[0043] Step S250: Output the set of defect points of any pole tower according to the pole tower dimension.

[0044] Specifically, the set of defect points of at least one pole tower in the power grid can be sent to the defect output device so that the defect output device outputs the set of defect points of any pole tower according to the pole tower dimension.

[0045] In the above embodiment, for multiple specified areas corresponding to multiple pole towers in the overhead line of the power grid, the corresponding inspection images and their position information are obtained. Then, the inspection images are input into a pre-trained defect recognition model to extract device component features, so as to perform defect prediction based on the extracted device component features to obtain the first defect information of the inspection images. Then, according to the position information corresponding to the inspection images, the first defect information corresponding to the inspection images is updated to obtain the second defect information corresponding to the inspection images. Then, according to the defect aggregation rule and the second defect information corresponding to each of the multiple inspection images, the set of defect points of any pole tower in the power grid is determined, and the set of defect points of any pole tower is output according to the pole tower dimension. In this way, the efficiency of power grid defect inspection can be improved while the diagnostic accuracy is improved.

[0046] In some embodiments, the specified area may be any one of the pole tower body area, the large-side channel, and the small-side channel.

[0047] Among them, the large-side channel refers to the space or path of the pole tower facing the power receiving end, and the small-side channel refers to the space or path of the pole tower facing the power sending end.

[0048] The pole tower body area may be any one of the whole tower, the tower top, the tower head, the tower body, the left cross arm, and the right cross arm.

[0049] Correspondingly, the inspection image corresponding to the large-side channel is the large-side channel inspection image, the inspection image corresponding to the small-side channel is the small-side channel inspection image, and the inspection image corresponding to the pole tower body area is the pole tower body inspection image.

[0050] In this embodiment, according to the defect aggregation rule and the second defect information corresponding to each of multiple inspection images, determining the defect point set of any pole tower in the power grid may include the following steps S310 - S320.

[0051] Step S310: For any pole tower, merge and deduplicate the second defect information corresponding to each of the multiple pole tower body inspection images in the pole tower to obtain the pole tower body defect point set of the pole tower.

[0052] Specifically, during the inspection of the pole tower by the data acquisition device, the perspectives and information captured at multiple shooting angles are different. The perspectives and information captured at different shooting angles may be different. For example, the perspectives captured for pole tower body areas such as the whole tower, tower top, tower head, tower body, left cross arm, and right cross arm may have repetitions or omissions. Therefore, for the defect information corresponding to multiple pole tower body inspection images, it can be merged and deduplicated according to the defect aggregation rule, thereby improving the accuracy of defect diagnosis and reducing the false detection rate of defect diagnosis.

[0053] Step S320: Merge the second defect information corresponding to the large - side channel inspection image and the small - side channel inspection image of the pole tower, and the pole tower body defect point set of the pole tower to obtain the defect point set of the pole tower.

[0054] Specifically, the perspectives captured for the large - side channel, small - side channel, and pole tower body areas usually do not overlap or have a low overlap rate. Therefore, for the defect information corresponding to the pole tower body inspection image, large - side channel inspection image, and small - side channel inspection image respectively, they can be merged, and in this way, the defect point set of the pole tower can be obtained.

[0055] In some embodiments, the defect aggregation rule can be used for the statistical logic between the second defect information corresponding to each of multiple inspection images of any pole tower, among defect points of the same type, so as to determine the defect point set of the same type of the pole tower.

[0056] Taking the type of defect point as tree and bamboo obstacles as an example for illustration below. For any pole tower, merge and deduplicate the tree and bamboo obstacle defect points in the second defect information corresponding to each of the multiple pole tower body inspection images, or in other words, the inspection images of pole tower body areas such as the whole tower, tower top, tower head, tower body, left cross arm, and right cross arm, to obtain the tree and bamboo obstacle defect points in the pole tower body area of the pole tower. Then, determine the tree and bamboo obstacle defect points of the large - side channel according to the second defect information corresponding to the large - side channel inspection image, and determine the tree and bamboo obstacle defect points of the small - side channel according to the second defect information corresponding to the small - side channel inspection image. Then, sum up the tree and bamboo obstacle defect points in the pole tower body area, the tree and bamboo obstacle defect points of the large - side channel, and the tree and bamboo obstacle defect points of the small - side channel to obtain the tree and bamboo obstacle defect points of the pole tower.

[0057] As an example, continuing with the example of the tree and bamboo barrier defect points, the defect aggregation rule can be expressed as , where represents the defect points after merging and deduplicating the tree and bamboo barrier defect points in multiple tower body areas, B represents the tree and bamboo barrier defect points in the large-size side channel, and C represents the tree and bamboo barrier defect points in the small-size side channel.

[0058] In some embodiments, the power grid defect diagnosis method may further include: for the defect point set of any tower, outputting the defect point set of the tower according to the type dimension of the defect points in the defect point set.

[0059] In some embodiments, before inputting the inspection image into the defect recognition model, the power grid diagnosis method may further include: preprocessing the inspection image to obtain a preprocessed inspection image; wherein, the preprocessing includes at least one of image denoising, image enhancement, and image segmentation.

[0060] In some embodiments, the defect recognition model can be obtained through the training method of the following steps S410 - S420.

[0061] Step S410: Construct a training image sample set; wherein, the training image sample set includes multiple training image samples, and the label of the training image sample is the type of the defect point.

[0062] Among them, the training image sample can be collected by the data acquisition device at a historical moment.

[0063] Step S420: Perform model training on the initial model according to the training image sample set and the label to obtain a defect recognition model; wherein, the initial model is built based on any one of an autoencoder, a Transformer, and a convolutional neural network model.

[0064] Exemplarily, the convolutional neural network (CNN) model can adopt any one of LeNet, VGG, ResNet, and DenseNet.

[0065] The embodiments of this specification provide a power grid defect diagnosis device. This power grid defect diagnosis device can be applied to the defect diagnosis device in the power grid defect diagnosis system. Please refer to Figure 3 , this power grid defect diagnosis device may include a data acquisition module 510, a defect prediction module 520, a defect location module 530, a defect aggregation module 540, and a defect output module 550.

[0066] A data acquisition module 510, configured to acquire defect aggregation rules, multiple inspection images, and location information corresponding to the inspection images; wherein, the multiple inspection images are images taken for multiple specified areas of at least one tower, and the defect aggregation rules are used to aggregate and count defect points in the multiple specified areas of the tower;

[0067] A defect prediction module 520, configured to input the inspection images into a pre-trained defect recognition model to extract device component features, and perform defect prediction based on the extracted device component features to obtain first defect information corresponding to the inspection images; wherein, the first defect information includes the type and quantity of defect points in the specified area corresponding to the inspection images;

[0068] A defect location module 530, configured to update the first defect information corresponding to the inspection images according to the location information corresponding to the inspection images to obtain second defect information corresponding to the inspection images; wherein, the second defect information includes the type, quantity, and location of defect points in the specified area corresponding to the inspection images;

[0069] A defect aggregation module 540, configured to determine a set of defect points of any tower in the power grid according to the defect aggregation rules and the second defect information corresponding to each of the multiple inspection images; wherein, the set of defect points includes the second defect information corresponding to each of the multiple specified areas in the tower;

[0070] A defect output module 550, configured to output the set of defect points of any tower according to the tower dimension.

[0071] In some embodiments, the specified area is any one of the tower body area, the large-side channel, and the small-side channel; the tower body area is any one of the whole tower, the tower top, the tower head, the tower body, the left cross arm, and the right cross arm; the inspection image corresponding to the large-side channel is the large-side channel inspection image, the inspection image corresponding to the small-side channel is the small-side channel inspection image, and the inspection image corresponding to the tower body area is the tower body inspection image.

[0072] In this embodiment, the defect aggregation module 540 can also be used to: for any tower, merge and deduplicate the second defect information corresponding to each of the multiple tower body inspection images in the tower to obtain the tower body defect point set of the tower; merge the second defect information corresponding to the large-side channel inspection image and the small-side channel inspection image of the tower, and the tower body defect point set of the tower to obtain the defect point set of the tower.

[0073] Regarding the specific functions and effects achieved by the power grid defect diagnosis device, reference may be made to other embodiments of this specification for comparison and explanation, which will not be elaborated herein. Each module in the power grid defect diagnosis device can be implemented in whole or in part by software, hardware, and their combinations. Each module can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0074] An embodiment of this specification also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by the computer, the computer executes the power grid defect diagnosis method in any of the above embodiments.

[0075] An embodiment of this specification also provides a computer program product containing instructions. When the instructions are executed by the computer, the computer executes the power grid defect diagnosis method in any of the above embodiments.

[0076] An embodiment of this specification also provides a computer device, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the power grid defect diagnosis method in the above embodiment is implemented.

[0077] In some embodiments, please refer to Figure 4 , this computer device may be a terminal, and its internal structure diagram may be as Figure 4 shown. This computer device includes a processor, a memory, and a communication interface connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of this computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, the power grid defect diagnosis method is implemented.

[0078] It can be understood that the specific examples in this article are only to help those skilled in the art better understand the embodiments of this specification, rather than limiting the scope of the present invention.

[0079] It can be understood that in various embodiments of this specification, the magnitudes of the sequence numbers of each process do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this specification.

[0080] It is understood that the various embodiments described in this specification can be implemented alone or in combination, and the embodiments of this specification do not limit this.

[0081] Unless otherwise specified, all technical and scientific terms used in the embodiments of this specification have the same meaning as commonly understood by those skilled in the technical field of this specification. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the scope of this specification. The term "and / or" used in the embodiments of this specification includes any and all combinations of one or more of the related listed items. The singular forms "a", "the above", and "the" used in the embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0082] It is understood that the processor in the embodiments of this specification can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method embodiments can be completed by the integrated logic circuit in the hardware of the processor or by instructions in the form of software. The above-mentioned processor can be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this specification. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of this specification can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method.

[0083] It can be understood that the memory in the embodiments of this specification can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0084] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this specification.

[0085] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0086] As described above, only the specific embodiments of this specification are provided, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this specification, and all of them should be covered by the protection scope of this specification. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for diagnosing power grid defects, characterized in that: The power grid includes a plurality of towers in an overhead line; the method includes: Acquire configured defect collection rules, multiple inspection images, and location information corresponding to the inspection images; wherein the multiple inspection images are images taken for multiple designated areas of at least one of the towers; the defect collection rules are used to collect and count defect points in multiple designated areas of the towers, and are used for statistical logic between defect points of the same type in the second defect information of the towers, so as to determine a set of defect points of the same type of the towers; Input the inspection image into a pre-trained defect recognition model to extract equipment component features, so as to perform defect prediction based on the extracted equipment component features, and obtain first defect information corresponding to the inspection image; wherein the first defect information includes the type and number of defect points in the specified area corresponding to the inspection image; According to the position information corresponding to the inspection image, the first defect information corresponding to the inspection image is updated to obtain the second defect information corresponding to the inspection image; wherein the second defect information includes the type, quantity and position of the defect points in the specified area corresponding to the inspection image; According to the defect collection rule and the second defect information corresponding to each of the multiple inspection images, a defect point set of any of the pole towers in the power grid is determined; wherein the defect point set includes the second defect information corresponding to each of the multiple designated areas in the pole tower; the multiple designated areas include a pole tower body area, a large side channel and a small side channel, the pole tower body area includes at least one of the entire tower, the tower top, the tower head, the tower body, the left cross arm and the right cross arm, the large side channel is the path of the pole tower facing the power receiving end, and the small side channel is the path of the pole tower facing the power transmitting end; the types of defect points in the defect point set include tree and bamboo barrier defect points, and the defect collection rule corresponding to the tree and bamboo barrier defect points is A max +B+C,A max A represents the tree and bamboo barrier defect points in the tower body area after merging and deduplicating the tree and bamboo barrier defect points in the tower body area, B represents the tree and bamboo barrier defect points in the large side channel, and C represents the tree and bamboo barrier defect points in the small side channel; Outputting a defect point set of any of the towers according to the tower dimensions and the defect point type dimensions; The defect recognition model is obtained by the following training method: constructing a training image sample set; wherein the training image sample set includes multiple training image samples, the labels of the training image samples are the types of defect points, and the labels include tree and bamboo barrier defect points; the initial model is trained according to the training image sample set and the labels to obtain the defect recognition model; wherein the initial model is built based on any one of the autoencoder, Transformer, and convolutional neural network models.

2. The power grid defect diagnosis method according to claim 1, characterized in that: The designated area is any one of the tower body area, the large side channel and the small side channel; the tower body area is any one of the entire tower, the tower top, the tower head, the tower body, the left cross arm and the right cross arm; the inspection image corresponding to the large side channel is the large side channel inspection image, the inspection image corresponding to the small side channel is the small side channel inspection image, and the inspection image corresponding to the tower body area is the tower body inspection image; Determining a defect point set of any of the towers in the power grid according to the defect aggregation rule and the second defect information corresponding to each of the plurality of inspection images includes: For any of the pole towers, merging and removing duplicates of the second defect information corresponding to a plurality of the pole tower body inspection images in the pole tower to obtain a pole tower body defect point set of the pole tower; The second defect information corresponding to the large side channel inspection image and the small side channel inspection image of the tower and the tower body defect point set of the tower are merged to obtain the defect point set of the tower.

3. The power grid defect diagnosis method according to claim 1, characterized in that: Before inputting the inspection image into the defect recognition model, the method further includes: Preprocessing is performed on the inspection image to obtain the preprocessed inspection image; wherein the preprocessing includes at least one of image denoising, image enhancement and image segmentation.

4. The power grid defect diagnosis method according to claim 1, characterized in that: The types of defect points include at least one of the following: dirty insulators, discharge marks on insulators, tree and bamboo barriers, bird nests on poles and towers, untied conductors, and damaged conductors.

5. A power grid defect diagnosis device, characterized in that: The power grid includes a plurality of pole towers in an overhead line; the device includes: A data acquisition module, used to acquire a configured defect collection rule, a plurality of inspection images and position information corresponding to the inspection images; wherein the plurality of inspection images are images taken for a plurality of designated areas of at least one of the towers; the defect collection rule is used to collect and count defect points in a plurality of designated areas of the towers, and is used for statistical logic between defect points of the same type in the second defect information of the towers, so as to determine a set of defect points of the same type of the towers; A defect prediction module, used for inputting the inspection image into a pre-trained defect recognition model to extract equipment component features, so as to perform defect prediction based on the extracted equipment component features, and obtain first defect information corresponding to the inspection image; wherein the first defect information includes the type and number of defect points in the specified area corresponding to the inspection image; A defect location module, used to update the first defect information corresponding to the inspection image according to the position information corresponding to the inspection image, and obtain the second defect information corresponding to the inspection image; wherein the second defect information includes the type, quantity and position of the defect points in the specified area corresponding to the inspection image; A defect collection module is used to determine a defect point set of any of the towers in the power grid according to the defect collection rule and the second defect information corresponding to each of the multiple inspection images; wherein the defect point set includes the second defect information corresponding to each of the multiple designated areas in the tower; the multiple designated areas include a tower body area, a large side channel and a small side channel, the tower body area includes at least one of the entire tower, the tower top, the tower head, the tower body, the left cross arm and the right cross arm, the large side channel is the path of the tower facing the power receiving end, and the small side channel is the path of the tower facing the power transmitting end; the types of defect points in the defect point set include tree and bamboo barrier defect points, and the defect collection rule corresponding to the tree and bamboo barrier defect points is A max +B+C,A max A represents the tree and bamboo barrier defect points in the tower body area after merging and deduplicating the tree and bamboo barrier defect points in the tower body area, B represents the tree and bamboo barrier defect points in the large side channel, and C represents the tree and bamboo barrier defect points in the small side channel; A defect output module, used for outputting a defect point set of any of the towers according to the tower dimension and the defect point type dimension; The defect recognition model is obtained by the following training method: constructing a training image sample set; wherein the training image sample set includes multiple training image samples, the labels of the training image samples are the types of defect points, and the labels include tree and bamboo barrier defect points; the initial model is trained according to the training image sample set and the labels to obtain the defect recognition model; wherein the initial model is built based on any one of the autoencoder, Transformer, and convolutional neural network models.

6. The power grid defect diagnosis device according to claim 5, characterized in that: The designated area is any one of the tower body area, the large side channel and the small side channel; the tower body area is any one of the entire tower, the tower top, the tower head, the tower body, the left cross arm and the right cross arm; the inspection image corresponding to the large side channel is the large side channel inspection image, the inspection image corresponding to the small side channel is the small side channel inspection image, and the inspection image corresponding to the tower body area is the tower body inspection image; The defect collection module is also used to: for any of the towers, merge and remove duplicate second defect information corresponding to each of the multiple tower body inspection images in the tower to obtain a tower body defect point set of the tower; merge the second defect information corresponding to each of the large side channel inspection image and the small side channel inspection image of the tower with the tower body defect point set of the tower to obtain a defect point set of the tower.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the power grid defect diagnosis method according to any one of claims 1 to 4 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the power grid defect diagnosis method according to any one of claims 1 to 4 is implemented.

Images