Insulator fault detection method, device, medium and electronic equipment

Abstract

The embodiment of the present application discloses an insulator fault detection method, device, medium and electronic device. The method comprises: extracting feature data of insulators in a target image using a preset feature extraction algorithm as features to be matched; matching the features to be matched with fault features in a faulty insulator sample database; wherein the faulty insulator sample database is pre-constructed based on a convolutional neural network model; if the match is successful, marking the insulator in the target image as a faulty insulator. Executing the technical solution of the present application can improve the efficiency and accuracy of insulator fault detection.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of insulator fault detection, and in particular to an insulator fault detection method, device, medium and electronic equipment. Background Art

[0002] Power insulators are important power components that support and insulate power transmission lines. Once an insulator fails, it will seriously affect the stable transmission of electric energy. In order to ensure the stable operation of the power grid, it is particularly important to detect faulty insulators in transmission lines.

[0003] At present, insulator fault detection is mainly completed by patrol and maintenance personnel manually sorting the insulator images collected by patrol equipment. This manual insulator fault detection method is limited by the energy and experience of patrol and maintenance personnel, and generally has the problems of low detection efficiency and low detection accuracy. Summary of the invention

[0004] The embodiments of the present application provide an insulator fault detection method, device, medium and electronic equipment, which realize the automation of insulator fault detection and achieve the purpose of improving the efficiency and accuracy of insulator fault detection.

[0005] In a first aspect, an embodiment of the present application provides an insulator fault detection method, the method comprising:

[0006] The feature data of the insulator in the target image is extracted using a preset feature extraction algorithm as the feature to be matched;

[0007] Matching the to-be-matched feature with the fault feature in a fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model;

[0008] If the matching is successful, the insulator in the target image is marked as a faulty insulator.

[0009] In a second aspect, an embodiment of the present application provides an insulator fault detection device, the device comprising:

[0010] A feature extraction module to be matched is used to extract feature data of insulators in the target image using a preset feature extraction algorithm as features to be matched;

[0011] A feature matching module, used to match the to-be-matched feature with the fault feature in the fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model;

[0012] The faulty insulator marking module is used to mark the insulator in the target image as a faulty insulator if the matching is successful.

[0013] In a third aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the insulator fault detection method as described in the embodiment of the present application.

[0014] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the insulator fault detection method as described in the embodiment of the present application when executing the computer program.

[0015] The technical solution provided in the embodiment of the present application improves the detection speed of insulator faults by pre-building a faulty insulator sample database and matching the feature data of the insulators in the target image with the fault features in the faulty insulator sample database. The faulty insulator sample database constructed based on the convolutional neural network model provides accurate and reliable data reference for insulator fault detection, ensuring the accuracy of insulator fault detection. The embodiment of the present application realizes the automation of insulator fault detection, reduces labor costs, effectively improves the work efficiency of power inspections, and improves the quality of power supply services. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flow chart of an insulator fault detection method provided in Embodiment 1 of the present application;

[0017] Figure 2 is a flow chart of another insulator fault detection method provided in Embodiment 2 of the present application;

[0018] Figure 3 is a flow chart of another insulator fault detection method provided in Embodiment 3 of the present application;

[0019] Figure 4 It is a structural schematic diagram of an insulator fault detection device provided in Embodiment 4 of the present application;

[0020] Figure 5 This is a structural diagram of an electronic device provided in Example 6 of the present application. DETAILED DESCRIPTION

[0021] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only the parts related to the present application, rather than all structures, are shown in the accompanying drawings.

[0022] It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the steps as sequential processes, many of the steps therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the steps can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0023] Embodiment 1

[0024] Figure 1 This is a flow chart of an insulator fault detection method provided in the first embodiment of the present application. This embodiment can be applied to the situation where power grid operation and maintenance personnel detect faulty insulators in transmission lines when patrolling high-voltage lines. The method can be executed by the insulator fault detection device provided in the embodiment of the present application, which can be implemented by software and / or hardware, and can be integrated into the electronic equipment running this system.

[0025] like Figure 1 As shown, the insulator fault detection method includes:

[0026] S110, extracting feature data of insulators in the target image using a preset feature extraction algorithm as features to be matched.

[0027] The target image is an image of an insulator to be detected, and the target image should include at least one insulator body. The insulator to be detected is an insulator that needs to be fault-detected, and optionally, the insulator to be detected is an insulator installed on a power transmission line.

[0028] In an optional embodiment, the target image is collected by a camera configured on an unmanned aircraft.

[0029] The unmanned aircraft may be a multi-rotor aircraft, which has the advantages of high flexibility, simple control and not easy to be damaged. The multi-rotor aircraft can reach various high-altitude power equipment and can get close to the insulator. Optionally, the camera configured on the unmanned aircraft is a high-definition camera with adjustable focal length, and the high-definition camera can be mounted on the unmanned aircraft through a turntable that can rotate freely 360 degrees.

[0030] Optionally, the unmanned aircraft is also equipped with a 5G communication module to support communication between the unmanned aircraft and the cloud server. Specifically, a multi-rotor aircraft equipped with a high-definition camera with adjustable focal length approaches the insulator at a close distance to capture images of the insulator, and transmits the captured images to the cloud server in real time through the 5G communication module. The cloud server completes the detection of insulator faults based on the insulator fault detection method provided in the embodiment of the present application. Optionally, the target image is collected in real time by the unmanned aircraft and transmitted to the cloud server in real time.

[0031] Since insulators are installed on outdoor power transmission lines and are affected by weather and other conditions, there is a risk of degradation in the images taken by the camera. Optionally, before the cloud server performs fault detection on the insulator, the original image received from the drone is pre-processed. For example, the original image is denoised or dehazed to obtain a target image that meets the feature extraction requirements.

[0032] The cloud server uses a preset feature extraction algorithm to extract feature data of the insulator in the target image as features to be matched. The preset feature extraction algorithm is an algorithm for extracting effective information from the image area where the insulator is located in the target image, wherein effective information refers to information that is helpful in determining whether the insulator is faulty. Since there are differences between faulty insulators and normal insulators in terms of color, texture, and shape, optionally, the color features, texture features, and shape features of the image area where the insulator is located in the target image are extracted using a preset feature extraction algorithm. The preset feature extraction algorithm is not the focus of the application embodiment, and the preset feature extraction algorithm can be any image feature extraction algorithm that can extract the above-mentioned feature data.

[0033] The feature data of the insulator in the target image extracted by the preset feature extraction algorithm is the feature to be matched. There is a corresponding relationship between the feature to be matched and the target image. For the convenience of description, the insulator in the target image is called the target insulator, the feature to be matched corresponds to the target insulator, and the feature to be matched is the feature data of the target insulator.

[0034] S120. Match the feature to be matched with the fault feature in a faulty insulator sample database; wherein the faulty insulator sample database is pre-constructed based on a convolutional neural network model.

[0035] The faulty insulator sample database includes a large number of fault features. The fault features included in the faulty insulator sample database almost cover the common fault types that affect the insulation performance of insulators. The faulty insulator sample database is constructed based on a convolutional neural network. Specifically, the faulty insulators are screened out through a convolutional neural network, and the feature data of the faulty insulators are used as the fault features in the faulty insulator sample database. It is worth noting that there are no two identical fault features in the faulty insulator sample database.

[0036] The fault feature refers to the feature data of the faulty insulator. The fault feature is extracted from the image including the faulty insulator. The fault feature is used as the reference data for insulator fault detection. The cloud server matches the feature to be matched with the fault feature, and determines whether the target insulator corresponding to the feature to be matched has a fault based on the matching result. The matching result includes a successful match and a failed match.

[0037] S130: If the matching is successful, mark the insulator in the target image as a faulty insulator.

[0038] If the feature to be matched successfully matches the fault feature, it means that the feature to be matched is highly similar to the fault feature, the target insulator corresponding to the feature to be matched has the fault feature, and the target insulator is likely to be a faulty insulator. At this time, the cloud server marks the insulator in the target image, that is, the target insulator, as a faulty insulator.

[0039] On the contrary, if the matching feature fails to match the fault feature, it means that the similarity between the matching feature and the fault feature is low, and the target insulator corresponding to the matching feature does not have the fault feature, and the target insulator is likely to be a normal insulator. In order to save computing resources, there is no need to mark the target insulator at this time.

[0040] Insulators are important power components that support and insulate transmission lines. Once an insulator fails, it will affect the stable transmission of electricity. For example, an insulator failure may cause a grounding fault in the transmission line, causing a large-scale power outage in the power line, affecting residents' production and living electricity. In order to reduce unnecessary complaints and shorten the duration of power outages, after detecting a faulty insulator, the operation and maintenance team can be instructed to promptly repair the faulty insulator.

[0041] In an optional embodiment, after marking the insulator in the target image as a faulty insulator, the method further includes: obtaining geographic location information of the target image as the target location; and generating insulator fault alarm information based on the target location to enable power operation and maintenance personnel to inspect the faulty insulator based on the target location.

[0042] The geographical location information of the target image is used to determine the geographical location information of the insulator in the target image. The geographical location information may be the longitude and latitude data of the location of the insulator in the target image.

[0043] Optionally, the geographic location information of the target image can be attached to the target image by the image acquisition device during the image acquisition process. It is obtained by parsing the attribute information of the target image. Optionally, if the image acquisition device is a camera configured on an unmanned aircraft, the cloud server will successfully match the to-be-matched feature with the fault feature, mark the insulator in the target image as a faulty insulator, and use 5G communication technology to feed back the detection result that the target insulator is a faulty insulator to the unmanned aircraft in real time. In response to receiving the detection result that the target insulator is a faulty insulator, the unmanned aircraft automatically records the geographic location of the target insulator through its own GPS positioning module and feeds it back to the cloud server.

[0044] The cloud server obtains the geographic location information of the target image as the target location; the cloud server generates insulator fault alarm information based on the target location, and sends the insulator fault alarm information to the terminal device, so that the power operation and maintenance personnel can inspect the faulty insulator according to the target location.

[0045] The technical solution provided in the embodiment of the present application improves the detection speed of insulator faults by pre-building a faulty insulator sample database and matching the feature data of the insulators in the target image with the fault features in the faulty insulator sample database. The faulty insulator sample database constructed based on the convolutional neural network model provides accurate and reliable data reference for insulator fault detection, ensuring the accuracy of insulator fault detection. The embodiment of the present application realizes the automation of insulator fault detection, reduces labor costs, effectively improves the work efficiency of power inspections, and improves the quality of power supply services.

[0046] The technical solution provided by the present application can be applied to the situation of daily power line inspection, or to the situation of troubleshooting the cause of a power failure after it has occurred. In an optional embodiment, an unmanned aircraft equipped with a high-definition camera is used to take images of insulators in service on the transmission line, and the images of the insulators taken are transmitted to a cloud server in real time through a 5G communication module. Through the cloud server, the insulators are detected for faults based on a sample database of faulty insulators, and the cloud server feeds back the insulator detection results to the unmanned aircraft and the user terminal in real time.

[0047] The faulty insulator sample database in the embodiment of the present application can provide fault detection support for hundreds of insulator images collected by unmanned aerial vehicles. The technical solution provided by the present application can quickly identify insulators with flashover and discharge traces, and the inspection efficiency is high. Generally, one insulator can be inspected in 1-2 seconds, which is more than 100 times the efficiency of traditional manual inspections. It effectively improves the efficiency of insulator fault detection, discovers faulty insulators in time, realizes timely fault elimination, reduces the duration of high-voltage line power outages, and improves the service quality of power supply.

[0048] Embodiment 2

[0049] Figure 2 It is a flow chart of another insulator fault detection method provided in the second embodiment of the present application. This embodiment is further optimized on the basis of the above embodiment. Specifically, before the operation of "matching the feature to be matched with the fault feature in the fault insulator sample database", the following operations are added: obtaining an alternative image, and extracting the feature data of the alternative image as an alternative feature; inputting the alternative feature into the convolutional neural network model, and outputting the fault type of the alternative feature through the convolutional neural network model; wherein the convolutional neural network model is pre-trained; if the fault type is the expected type, the alternative feature is used as the fault feature and added to the fault insulator sample database; it is determined whether the number of the fault features in the fault insulator sample database reaches the set number, if not, then continue to add fault features to the fault insulator sample database to further expand the fault insulator sample database.

[0050] like Figure 2 As shown, the insulator fault detection method includes:

[0051] S210: Acquire a candidate image, and extract feature data of the candidate image as candidate features.

[0052] The candidate image is an image used to construct a faulty insulator sample database, and the candidate image includes at least one insulator body. The candidate image may be a historical insulator image. The number of candidate images is at least two. Considering that the fault sample database serves as a data support for insulator fault detection, it should theoretically cover common fault types that affect the insulation performance of insulators. Therefore, within a certain range, the more candidate images, the better.

[0053] The cloud server obtains the candidate image and extracts feature data of the candidate image. Specifically, the cloud server uses a preset feature extraction algorithm to extract the feature data of the candidate image as the candidate feature.

[0054] S220. Input the candidate feature into the convolutional neural network model, and output the fault type of the candidate feature through the convolutional neural network model; wherein the convolutional neural network model is pre-trained.

[0055] The cloud server inputs the candidate features into the convolutional neural network, and outputs the fault type of the candidate image through the convolutional neural network. The fault type of the candidate image represents the fault type of the insulator in the candidate image. The convolutional neural network refers to a pre-trained convolutional neural network, and the trained convolutional neural network refers to a convolutional neural network with an insulator fault detection rate that reaches the upper limit. The specific structure of the convolutional neural network is not limited here. An exemplary convolutional neural network can be a YOLO detection network.

[0056] The convolutional neural network is trained based on sample features consistent with the candidate feature type and label data corresponding to the sample features. The label data is determined according to the fault condition of the insulator corresponding to the sample feature. The fault type refers to the label data corresponding to the candidate feature output by the convolutional neural network, and the fault type is related to the sample label used to train the convolutional neural network. Exemplarily, the fault type may include normal and specific faults such as flashover, breakdown, stains or cracks.

[0057] S230: If the fault type is an expected type, the selected feature is used as the fault feature and added to the fault insulator sample database.

[0058] Among them, the expected type is related to the actual business needs. The expected type is the insulator fault that the user predetermines to troubleshoot. Considering that the faulty insulator sample database serves as the data support for insulator fault detection, it should theoretically cover the common fault types that affect the insulation performance of insulators. Optionally, all fault types except the normal fault type are considered as expected types.

[0059] In an optional embodiment, the expected type includes at least one of flashover, breakdown, stains and cracks. Flashover refers to the phenomenon of discharge along the surface of a solid insulator when the gas or liquid dielectric surrounding the insulator is broken down. After flashover occurs, sparks or arcs in the flashover channel cause local overheating of the insulation surface, resulting in carbonization and leaving fine marks on the surface of the insulator. Breakdown refers to the actual voltage passing through the insulator far exceeding the working voltage of the insulator, forming a conductive channel in the insulator, and a strong current passes through the insulator instantaneously. After breakdown occurs, discharge burn marks will form on the surface of the insulator. Cracks refer to the situation where cracks or defects appear on the main body of the insulator. Stains refer to the phenomenon that a large amount of dirt exists on the surface of the main body of the insulator. The presence of a large amount of dirt on the insulation is an important cause of insulator flashover and insulator breakdown.

[0060] Flashover, breakdown, stains and cracks are all common faults that affect the insulation performance of insulators. In order to ensure stable power transmission, it is necessary to troubleshoot insulators in service during daily power line inspections and replace faulty insulators.

[0061] In order to facilitate data management, improve data management efficiency, and facilitate users to query data in the faulty insulator sample database. When constructing the faulty insulator sample data in the embodiment of the present application, a partitioned storage method according to the insulator type is adopted. In an optional embodiment, the alternative feature is used as the fault feature and added to the faulty insulator sample database, including: determining the type information of the insulator in the alternative image according to the alternative feature; establishing an association relationship between the type information and the alternative feature; and adding the alternative feature associated with the type information as the fault feature to the faulty insulator sample database.

[0062] Due to the differences in the materials and functional uses of insulators, different types of insulators differ in color, shape, volume, and installation position. For example, insulators include disc insulators, rod insulators, red insulators, gray insulators, glass insulators, composite insulators, or porcelain insulators.

[0063] The cloud server determines the type information of the insulator in the candidate image based on the candidate features. Specifically, the cloud server determines the type information of the insulator in the candidate image based on the color features and shape features in the candidate features. The cloud server establishes an association relationship between the type information and the candidate features based on the insulator type information, and associates the candidate features with the type of the insulator. The cloud server adds the candidate features associated with the type information as fault features to the fault insulator database. When the insulator type is determined, the corresponding candidate features can be queried. In order to facilitate data management and improve data management efficiency, it is convenient for users to query data in the fault insulator sample database.

[0064] S240: Determine whether the number of the fault features in the faulty insulator sample database reaches a set number; if not, continue to add fault features to the faulty insulator sample database to further expand the faulty insulator sample database.

[0065] The insulator sample database provides data support for insulator fault detection. In order to ensure the accuracy of insulator fault detection, the insulator sample database needs to reach a certain scale before it can be put into use. After the cloud server adds the candidate features as fault features to the insulator sample database, it counts the number of fault features in the faulty insulator sample data to determine whether the number of fault features in the faulty insulator sample database reaches the set number. The set number is determined by relevant technical personnel based on actual business needs, and the specific number of the set number is not limited here.

[0066] If the number of fault features does not reach the set number, it indicates that the scale of the faulty insulator sample data has not yet reached the conditions for being put into use, and the cloud server needs to continue to add fault features to the faulty insulator sample database to further expand the faulty insulator sample database.

[0067] If the number of fault features reaches the set number, it means that the scale of the fault insulator sample data has reached the conditions for being put into use, and the fault insulator sample database is built. It is worth noting that the built fault insulator sample database is not fixed, and users can update the fault insulator database according to actual business needs.

[0068] S250, extracting feature data of insulators in the target image using a preset feature extraction algorithm as features to be matched.

[0069] S260. Match the feature to be matched with the fault feature in the fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model.

[0070] It can be known that there is a one-to-one correspondence between the fault feature and the type of the feature to be matched. For example, when the type to be matched includes at least one of color feature, texture feature and shape feature, the fault feature also includes at least one of color feature, texture feature and shape feature. The cloud server matches the feature to be matched with the fault feature. Optionally, the fault feature is feature matched with the feature to be matched according to the feature type.

[0071] S270: If the match is successful, mark the insulator in the target image as a faulty insulator.

[0072] The embodiment of the present application screens the alternative features through a pre-trained convolutional neural network, thereby ensuring the quality of the faulty insulator sample database. The embodiment of the present application counts the number of fault features in the faulty insulator sample database, and when the number of fault features reaches a set number, the faulty insulator sample database is put into use, thereby ensuring the scale of the faulty insulator sample database. The embodiment of the present application provides strong data support for faulty insulator detection by constructing a faulty insulator sample database including a set number of fault features, thereby ensuring the accuracy and efficiency of insulator fault detection.

[0073] Embodiment 3

[0074] Figure 3 This is a flow chart of another insulator fault detection method provided in the third embodiment of the present application. This embodiment is further optimized on the basis of the above embodiment. Specifically, "matching the to-be-matched feature with the fault feature in the fault insulator sample database" is refined.

[0075] like Figure 3 As shown, the insulator fault detection method includes:

[0076] S310, extracting feature data of insulators in the target image using a preset feature extraction algorithm as features to be matched.

[0077] In an optional embodiment, the feature to be matched includes: at least one of a color feature, a texture feature and a shape feature.

[0078] After an insulator fails, its color, texture, and shape will change to a certain extent. For example, after an insulator flashes over, sparks or arcs in the flashover channel cause local overheating of the insulation surface, resulting in carbonization and leaving fine marks on the insulator surface; after an insulator breaks down, discharge burn marks will form on the insulator surface; the main body of the faulty insulator may be cracked or incomplete. Specifically, the cloud server uses a preset feature extraction algorithm to extract the color features, texture features, and shape features of the insulator in the target image as features to be matched.

[0079] S320: Determine the type of the insulator in the target image as the target type according to the color feature and the shape feature in the features to be matched.

[0080] Due to the differences in the materials and functional uses of insulators, different types of insulators differ in color, shape, volume, and installation position. For example, insulators include disc insulators, rod insulators, red insulators, gray insulators, glass insulators, composite insulators, or porcelain insulators.

[0081] The target type refers to the type of the insulator in the target image. The cloud server determines the insulator type to which the target insulator belongs based on the color features and shape features in the features to be matched, and obtains the target type.

[0082] S330: Select, from the faulty insulator sample database, a fault feature belonging to the target type as a reference feature.

[0083] Since the fault features in the faulty insulator sample database are classified and stored according to the type information of the insulator, the fault features belonging to the type of insulator can be queried according to the insulator type information. The cloud server uses the target type as an index and selects the fault features belonging to the target type in the faulty insulator sample database as reference features. The reference features are the fault features used for comparison with the features to be matched.

[0084] There is also a certain correlation between the type of insulator and the type of insulator failure. For example, the probability of anti-pollution insulator flashover and stain attachment is much lower than that of general insulators. The fault features belonging to the target type are selected as reference features to narrow the feature matching range. This can effectively improve the detection efficiency and accuracy of insulator faults.

[0085] S340: Determine the similarity between the feature to be matched and the reference feature, and determine whether the match is successful based on the similarity and a set threshold.

[0086] The cloud server uses a similarity calculation algorithm to calculate the similarity between the feature to be matched and the reference feature, and compares the calculated similarity with the set threshold to determine whether the feature to be matched matches the reference feature. The set threshold is determined by relevant technical personnel based on actual business needs, and the set threshold is related to the insulator fault detection accuracy required by actual business needs. The set threshold is not limited here and is determined based on actual conditions.

[0087] S350: If the matching is successful, mark the insulator in the target image as a faulty insulator.

[0088] The technical solution provided in the embodiment of the present application is to match the feature to be matched with the fault feature in the faulty insulator sample database, and determine the type of the insulator in the target image as the target type according to the color feature and the shape feature; in the faulty insulator sample database, select the fault feature belonging to the target type as the reference feature; determine the similarity between the feature to be matched and the reference feature, and determine whether the match is successful according to the similarity and the set threshold. If the match is successful, mark the insulator in the target image as a faulty insulator. The technical solution provided in the embodiment of the present application is to select the fault feature belonging to the target type as the reference feature and match it with the feature to be matched, taking into account the association between the insulator type and the fault type that the insulator is prone to, narrowing the feature matching range, and improving the efficiency and accuracy of insulator fault detection.

[0089] Embodiment 4

[0090] Figure 4 The fourth embodiment of the present application provides an insulator fault detection device, which can be applied to the case of optimizing the exhaust system of the engine. The device can be implemented by software and / or hardware, and can be integrated into electronic devices such as smart terminals.

[0091] like Figure 4 As shown, the device may include: a to-be-matched feature extraction module 410, a feature matching module 420 and a fault insulator marking module 430

[0092] The feature extraction module 410 is used to extract the feature data of the insulator in the target image as the feature to be matched by using a preset feature extraction algorithm;

[0093] A feature matching module 420 is used to match the to-be-matched feature with the fault feature in the fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model;

[0094] The faulty insulator marking module 430 is configured to mark the insulator in the target image as a faulty insulator if the matching is successful.

[0095] The technical solution provided in the embodiment of the present application improves the detection speed of insulator faults by pre-building a faulty insulator sample database and matching the feature data of the insulators in the target image with the fault features in the faulty insulator sample database. The faulty insulator sample database constructed based on the convolutional neural network model provides accurate and reliable data reference for insulator fault detection, ensuring the accuracy of insulator fault detection. The embodiment of the present application realizes the automation of insulator fault detection, reduces labor costs, effectively improves the work efficiency of power inspections, and improves the quality of power supply services.

[0096] Optionally, the device further comprises: an alternative feature extraction module, used to obtain an alternative image and extract feature data of the alternative image as an alternative feature before matching the feature to be matched with the fault feature in the fault insulator sample database;

[0097] A fault type determination module, used to input the candidate feature into the convolutional neural network model, and output the fault type of the candidate feature through the convolutional neural network model; wherein the convolutional neural network model is pre-trained;

[0098] A fault feature adding module, configured to add the selected feature as the fault feature to the fault insulator sample database if the fault type is an expected type;

[0099] The fault feature quantity judgment module is used to judge whether the number of the fault features in the fault insulator sample database reaches a set number. If not, continue to add fault features to the fault insulator sample database to further expand the fault insulator sample database.

[0100] Optionally, the fault feature adding module includes: an insulator type determination submodule, used to determine the type information of the insulator in the alternative image based on the alternative features; an association relationship establishment submodule, used to establish an association relationship between the type information and the alternative features; and a fault feature adding submodule, used to add the alternative features associated with the type information as the fault features to the fault insulator sample database.

[0101] Optionally, the feature matching module 420 includes: a target type determination submodule, used to determine the type of insulator in the target image as the target type based on the color features and shape features in the features to be matched; a reference feature determination submodule, used to select a fault feature belonging to the target type in the faulty insulator sample database as a reference feature; and a similarity determination submodule, used to determine the similarity between the features to be matched and the reference features, and determine whether the match is successful based on the similarity and a set threshold.

[0102] Optionally, the device also includes: a target position determination module, used to obtain the geographic location information of the target image as the target position after marking the insulator in the target image as a faulty insulator; an insulator fault alarm information generation module, used to generate insulator fault alarm information according to the target position so as to enable power operation and maintenance personnel to inspect the faulty insulator according to the target position.

[0103] Optionally, the expected type includes at least one of flashover, breakdown, stain and crack.

[0104] Optionally, the target image is collected by a camera configured on an unmanned aircraft.

[0105] An insulator fault detection device provided by an embodiment of the present invention can execute an insulator fault detection method provided by any embodiment of the present invention, and has a corresponding performance module and beneficial effects for executing an insulator fault detection method.

[0106] Embodiment 5

[0107] Embodiment 5 of the present application further provides a storage medium including computer executable instructions, wherein the computer executable instructions are used to execute an insulator fault detection method when executed by a computer processor, the method comprising:

[0108] The feature data of the insulator in the target image is extracted using a preset feature extraction algorithm as the feature to be matched;

[0109] Matching the to-be-matched feature with the fault feature in a fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model;

[0110] If the matching is successful, the insulator in the target image is marked as a faulty insulator.

[0111] Storage media refers to any of various types of memory electronic devices or storage electronic devices. The term "storage medium" is intended to include: installation media, such as CD-ROM, floppy disk or tape device; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (such as hard disk or optical storage); registers or other similar types of memory elements, etc. The storage medium may also include other types of memory or combinations thereof. In addition, the storage medium may be located in the computer system in which the program is executed, or may be located in a different second computer system, which is connected to the computer system via a network (such as the Internet). The second computer system can provide program instructions to the computer for execution. The term "storage medium" may include two or more storage media that may reside in different unknowns (for example, in different computer systems connected via a network). The storage medium may store program instructions (for example, embodied as a computer program) that can be executed by one or more processors.

[0112] Of course, the storage medium containing computer executable instructions provided in an embodiment of the present application is not limited to the engine exhaust system optimization insulator fault detection operation as described above, and can also execute related operations in the insulator fault detection method provided in any embodiment of the present application.

[0113] Embodiment 6

[0114] Embodiment 6 of the present application provides an electronic device, in which the insulator fault detection device provided in the embodiment of the present application can be integrated. The electronic device can be configured in a system, or can be a device that executes part or all of the functions in the system. Figure 5 Schematic diagram of the structure of an electronic device provided in Example 6 of the present application. Figure 5 As shown, this embodiment provides an electronic device 500, which includes: one or more processors 520; a storage device 510, which is used to store one or more programs. When the one or more programs are executed by the one or more processors 520, the one or more processors 520 implement the insulator fault detection method provided in the embodiment of the present application. The method includes:

[0115] The feature data of the insulator in the target image is extracted using a preset feature extraction algorithm as the feature to be matched;

[0116] Matching the to-be-matched feature with the fault feature in a fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model;

[0117] If the matching is successful, the insulator in the target image is marked as a faulty insulator.

[0118] Of course, those skilled in the art can understand that the processor 520 also implements the technical solution of the insulator fault detection method provided in any embodiment of the present application.

[0119] Figure 5 The electronic device 500 shown is merely an example and should not bring any limitation to the performance and scope of use of the embodiments of the present application.

[0120] like Figure 5 As shown, the electronic device 500 includes a processor 520, a storage device 510, an input device 530, and an output device 540; the number of the processor 520 in the electronic device can be one or more. Figure 5 A processor 520 is taken as an example; the processor 520, the storage device 510, the input device 530 and the output device 540 in the electronic device can be connected via a bus or other means. Figure 5 The connection via bus 550 is taken as an example.

[0121] The storage device 510 is a computer-readable storage medium that can be used to store software programs, computer executable programs, and module units, such as program instructions corresponding to the insulator fault detection method in the embodiment of the present application.

[0122] The storage device 510 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for performance; the data storage area may store data created according to the use of the terminal, etc. In addition, the storage device 510 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the storage device 510 may further include a memory remotely arranged relative to the processor 520, and these remote memories may be connected via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0123] The input device 530 may be used to receive input numbers, character information or voice information, and generate key signal input related to user settings and performance control of the electronic device. The output device 540 may include electronic devices such as a display screen and a speaker.

[0124] The insulator fault detection device, medium and electronic device provided in the above embodiments can execute the insulator fault detection method provided in any embodiment of the present application, and have the corresponding performance modules and beneficial effects of executing the method. For technical details not described in detail in the above embodiments, please refer to the insulator fault detection method provided in any embodiment of the present application.

[0125] Note that the above are only preferred embodiments of the present application and the technical principles used. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the appended claims.

Claims

1. An insulator fault detection method, characterized in that: The method comprises: The feature data of the insulator in the target image is extracted using a preset feature extraction algorithm as the feature to be matched; Matching the to-be-matched feature with the fault feature in a fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model; If the match is successful, marking the insulator in the target image as a faulty insulator; The matching of the to-be-matched feature with the fault feature in the fault insulator sample database includes: According to the color features and shape features in the features to be matched, determining the type of the insulator in the target image as the target type; In the faulty insulator sample database, a fault feature belonging to the target type is selected as a reference feature; wherein the fault features in the faulty insulator sample database are classified and stored according to the type information of the insulator; Determine the similarity between the feature to be matched and the reference feature, and determine whether the match is successful based on the similarity and a set threshold; Before matching the feature to be matched with the fault feature in the fault insulator sample database, the method further includes: Acquire a candidate image, and extract feature data of the candidate image as a candidate feature; Inputting the candidate feature into the convolutional neural network model, and outputting the fault type of the candidate feature through the convolutional neural network model; wherein the convolutional neural network model is pre-trained; If the fault type is an expected type, the selected feature is used as the fault feature and added to the fault insulator sample database; It is determined whether the number of the fault features in the faulty insulator sample database reaches a set number. If not, the fault features are continuously added to the faulty insulator sample database to further expand the faulty insulator sample database.

2. The method according to claim 1, characterized in that The method further comprises: taking the selected feature as the fault feature and adding it to the fault insulator sample database; Determining type information of insulators in the candidate image according to the candidate features; Establishing an association relationship between the type information and the candidate features; The candidate feature associated with the type information is used as the fault feature and added to the fault insulator sample database.

3. The method according to claim 1, characterized in that After marking the insulator in the target image as a faulty insulator, the method further includes: Acquire geographic location information of the target image as the target location; According to the target position, insulator fault alarm information is generated to enable power operation and maintenance personnel to repair the faulty insulator according to the target position.

4. The method according to claim 1, characterized in that: in, The expected types include at least one of flashover, breakdown, stain and crack.

5. The method according to claim 1, characterized in that in, The target image is collected by a camera configured on the unmanned aircraft.

6. An insulator fault detection device, characterized in that: The device comprises: A feature extraction module to be matched is used to extract feature data of insulators in the target image using a preset feature extraction algorithm as features to be matched; A feature matching module, used to match the to-be-matched feature with the fault feature in the fault insulator sample database; wherein the fault insulator sample database is pre-constructed based on a convolutional neural network model; A faulty insulator marking module, used for marking the insulator in the target image as a faulty insulator if the matching is successful; The feature matching module comprises: A target type determination submodule, used to determine the type of the insulator in the target image as the target type according to the color feature and the shape feature in the feature to be matched; A reference feature determination submodule, configured to select a fault feature belonging to the target type from the faulty insulator sample database as a reference feature; wherein the fault features in the faulty insulator sample database are classified and stored according to the type information of the insulator; A similarity determination submodule is used to determine the similarity between the feature to be matched and the reference feature, and determine whether the match is successful based on the similarity and a set threshold; The insulator fault detection device also includes: A candidate feature extraction module, used for obtaining a candidate image and extracting feature data of the candidate image as a candidate feature before matching the feature to be matched with the fault feature in the fault insulator sample database; A fault type determination module, used to input the candidate feature into the convolutional neural network model, and output the fault type of the candidate feature through the convolutional neural network model; wherein the convolutional neural network model is pre-trained; A fault feature adding module, configured to add the selected feature as the fault feature to the fault insulator sample database if the fault type is an expected type; The fault feature quantity judgment module is used to judge whether the number of the fault features in the fault insulator sample database reaches a set number. If not, continue to add fault features to the fault insulator sample database to further expand the fault insulator sample database.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the insulator fault detection method according to any one of claims 1 to 5 is implemented.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the insulator fault detection method according to any one of claims 1 to 5 is implemented.

Images